diff --git a/.gitattributes b/.gitattributes index ecd66c5b66dd7f19184bae01ab683ec4a06565d1..6faf1b2949ca16386d040d66c52328668cca5502 100644 --- a/.gitattributes +++ b/.gitattributes @@ -294,3 +294,16 @@ checkpoints/Llama-3.2-1B/babylm_reverse_partial_10M_seed0/artifacts/models--meta checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-900/tokenizer.json filter=lfs diff=lfs merge=lfs -text checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-100/tokenizer.json filter=lfs diff=lfs merge=lfs -text checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1700/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/GPT-2/babylm_reverse_full_10M_seed0/artifacts/models--gpt2/blobs/248dfc3911869ec493c76e65bf2fcf7f615828b0254c12b473182f0f81d3a707 filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2100/model.safetensors b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2100/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..56d0736fedfb4bc4764813c1573eecfe864edc1c --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2100/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d3621a4e42cc2c5cad279b3450fd9ef09ec5b4546548562775d917cab8173c52 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2200/model.safetensors b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2200/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0152b4b2fdeef814105a6049cfca1f70f328b67c --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2200/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3e1b7fbfd29153849db4684c989fd9536f3ec8949a70db40ec7a646b540d8320 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2300/model.safetensors b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2300/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..49a0a79c9ad2eb87b483ab6c428afdcfe729e54e --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2300/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fa77bada3cf328889026531cde3aced0ca7348eaf7dfe4d63fdb5a5453dae863 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2400/model.safetensors b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2400/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d89c7aede5781517128545de4c8bf3d7b355fa99 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2400/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:febd1939ba5ec58e6aee28d517d528577d1029323762745f64034e8a94ceda7e +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2500/model.safetensors b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2500/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..662be6c3400b43a4cce6f689c9806270bfc90520 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2500/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2fc4646004037925fcefb11d3282b67c94a692c05be616969767b9b8b7f9732a +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2600/model.safetensors b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2600/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..51ca4bc35fb1d7d6b61e427a480eb3a4a8dad212 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-2600/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dbc180db5bbccea247d1aa54689010da96cf250136eb68add7c3eb2b5aebda6 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-300/model.safetensors b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-300/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bab73a90ff30f246cc0d89296133db58c22aca1f --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_control_10M_seed0/runs/checkpoint-300/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a3c9aa6b8aff52d50b47aee227a6dd86316334acc17165d36d1da41059471549 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_full_10M_seed0/artifacts/models--gpt2/blobs/248dfc3911869ec493c76e65bf2fcf7f615828b0254c12b473182f0f81d3a707 b/checkpoints/GPT-2/babylm_reverse_full_10M_seed0/artifacts/models--gpt2/blobs/248dfc3911869ec493c76e65bf2fcf7f615828b0254c12b473182f0f81d3a707 new file mode 100644 index 0000000000000000000000000000000000000000..44b36d6e32d13c8fb28b0feab0ac8bfefa7efeda --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_full_10M_seed0/artifacts/models--gpt2/blobs/248dfc3911869ec493c76e65bf2fcf7f615828b0254c12b473182f0f81d3a707 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:248dfc3911869ec493c76e65bf2fcf7f615828b0254c12b473182f0f81d3a707 +size 548105171 diff --git a/checkpoints/GPT-2/babylm_reverse_full_10M_seed0/artifacts/models--gpt2/snapshots/607a30d783dfa663caf39e06633721c8d4cfcd7e/model.safetensors b/checkpoints/GPT-2/babylm_reverse_full_10M_seed0/artifacts/models--gpt2/snapshots/607a30d783dfa663caf39e06633721c8d4cfcd7e/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..44b36d6e32d13c8fb28b0feab0ac8bfefa7efeda --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_full_10M_seed0/artifacts/models--gpt2/snapshots/607a30d783dfa663caf39e06633721c8d4cfcd7e/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:248dfc3911869ec493c76e65bf2fcf7f615828b0254c12b473182f0f81d3a707 +size 548105171 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-100/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-100/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7e8d1e3e1bbad3bd29a9f9ffc55961e5d1dfc8cf --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-100/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbce2dc19c88327ccdd638378cf309339267b94959339b032dfd46918ad5f15e +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1100/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1100/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6e1c1c77dea606946df25408ec1cc63ea3f0346d --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1100/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3e58a65c3d4eb283fe014174a890cfbbecc025f3a137b375dc09ea0372d0e7da +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1200/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1200/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6d43fd326b98d2a8cde0e97a4253264285e46dc7 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1200/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:28b9d6469233b6f517cffe52eef1034c5aae0d6cafa930fb7e59f1feefdf4ac4 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1300/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1300/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fa5c09065524238b4171c2002246902c46f35afc --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1300/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e47d5e2ac22b27408d78568cbf135ee5dcbccfddd4c98f85a0715f294ca8236 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1400/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1400/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d729c0701cf25d9c4ac0a223079f3151660de867 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1400/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:777ec5eeb0591dd995361a0caf4686be57b66ceb13635213f7068a3e019e093c +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1500/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1500/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0affe3ad0f2b7ad9732e6f44e19829de7c050b37 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1500/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7821603eb8e686f0804cc0300c8319d32845730696a41690e42094dfb96b2a3d +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1600/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1600/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c90e5b9bd140ccb041dadfa4f2cb3cb5a7bf3191 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1600/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc275143ad3bf5c42fadd6fde0e0ed95ef73bab086244257694ee0f1f20a4b85 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1700/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1700/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9cd55cbd15e4f92d6902512c2a66a9110c67ea6e --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1700/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d32b436b597203c3bb9a72bc9db598262e90264454ae0c922146743e81a47bb +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1800/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1800/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3611dd443384bddf8358a45b717a628efe2d45df --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1800/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:719b6e0326e50aa26ad4f9e839be1d9c355596499cd156c99cd2ca72f12eea91 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1900/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1900/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d86bce9129602bbda48007639d59d92d75b3c6c6 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-1900/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b93da9b1f889ff06893a0272c9789e61719115c2cd31c59b75fc45b7690b1ef4 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-200/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-200/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..402a0ed677d10825c14a26c6d9667126f1322bf0 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-200/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be3096f2cd2008ababe2056da753b61f62ce6711d69c3629a0a6b9ca6cc171ed +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2000/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4643c6ae22d1a5b1c06cc533f9976ea03c7c42a4 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2eaab632ae2ef5e08af8b8e61edf8b8dd19262ae37a3a07ca3ee49085ecf2293 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2700/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2700/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d1536a58d8bbb4e7ed0ae38f8bdd6f6b7f30d0aa --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2700/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fda7ac84cde4f9f75fc3d58bd11610d90df061735584d346d8eb0a3c5e755cae +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2919/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2919/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e1eed8bcdcce1a311b6950a7dd950bf7f64db3f6 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-2919/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7593aafc3cf49f93f2f49c76c8652f670db002ec4244b7173e5776af928476ab +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-400/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-400/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a645f57fc62a07287c95c96ca97e6d445c8c8eba --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-400/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70e695632c2b4ea1232d5e8ff06eef4707b8c1a92389243721f1bba7a878ef00 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-600/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-600/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1d7aa79ef99caaa8adac481fd357e3574f7879bc --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-600/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e1b863cbbce509888f1fc064504a60412b84fa240b550e43ed394f28bd8d492 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-700/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-700/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cc9a6be1da1fc00779c4d1fea39080c01f03004c --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-700/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dca9ac06808d48d1c8224c697c45d4744b449b4e8032e6a12fdd7853f6a3c57 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-800/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-800/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..eaa77d0f4280612eeb6c74441f96f350b1dc2220 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-800/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6db5b63776b0a238d47e9a64800764dd780e4f3b85c1cf88a6c0c8a56940528d +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/model.safetensors b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..708f0743d66d3d79b0a6681a65ffd453b398c191 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f741e772b467b67e4d2598373dc9efc8372a84b24a6ca00e4f31c85642d18e0 +size 326089504 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/rng_state_1.pth b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..303cd1b38aea56cea8c592b6f3568a6219a57073 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1adf48ea3a4dfba5404637be6f82c9a15ebd900e796e29b6ffeb0fdc3b1d475c +size 16567 diff --git a/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/training_args.bin b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..b9c6310f232ce991a14691700385427742356e43 --- /dev/null +++ b/checkpoints/GPT-2/babylm_reverse_partial_10M_seed0/runs/checkpoint-900/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8eb0a6f24460a35a5849475791340c8d7dfb54225c5749b791b10e833071abfa +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..fad444f8143e1badd2a8d15a75e619abd485ceab --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:186b472c4073bfc086d5b4584aeaefc23f9c40f1a8199963bda08775617e1e12 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..39b6620f3f8d750109fe44053f80f2e221c6e290 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d8957d748b592f49a56723f1936535313c244cee56956dbc893c99d40f6ef84b +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..04d30c4b88bb88b6860f56816d3db2c6c7742594 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be2cd13083af0074c65198c01c90c0a3c599fbe6e0ea705a8835ca51f393d0c0 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..508856775b7a4bcca9f243e316cf313f5885d75b --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:31b0601a70e7904c3660f5346767114aa9fa6d4a2cba625fac4b4b9ea3ada066 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1100/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..9f4723485ff30930b59224e8afe0cb0b2b3c51bc --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f34c8972483cae48cde438f54c14179a5393e31298d634393f2e166c6a9bfb3e +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..6f245457df6ae96d9dae4e4269752c2f036e6ee3 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d0068d515416f5010564895d620c337c347a04bf8d5382015b5833708d52850 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2f6895ce66cfa6f18386f6edde947b44dd0255fa --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1597fdb72beb896bfa337f48e1d75c0339960cd7266fb14e0c9cea974436401f +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1200/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..3429b5bbf5e65631c22532fed743338bede54d6b --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7020c2ba12cf7959bf7033d7ad83dcf637e59ddbd3ff5b17fc4b9a1ebed3202 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..cbba424cb2109da26bffbb01960eb52b2dc47b0d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f29ab0036e394c39267ad86bf247bf1558e2213add95383fce34a980f109b42e +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..b47aeb757c57446497f8f5d1181a50c3faf2dd79 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:14413139ea5b777791c35297ed8a0a717bb3d6a3553268cc92b15fa61be1ae39 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..7808373865c368d93cf2d7dc92803c8297862fca --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3288dac9d8ea621655c7c1e03f2306d032714a31bcf52c99a781b6eb9a2538c3 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1300/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/generation_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/latest b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/latest new file mode 100644 index 0000000000000000000000000000000000000000..c2a0b5b17b996514a9bea0debe9403a641844bce --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/latest @@ -0,0 +1 @@ +global_step1400 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..466c494b53a5cf3030ac585b29d1563c49b789d4 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5075060d7b0d18a80b592cad740c3383144e4b5ba0bc7e7b9ff8ebb2afad8e5 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..262d37671147896b5dfa4ab77f0e9578ea0d968c --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46766e2820279a98976091f452887a7ad876d6c74fa04b8df6451b77352f4c07 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..b31f6e0c1e981c4c1dee9a4626e2a6c6ae5abb8e --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:deab7068d9411eb964839c7901679767e090930bd8dfb3faa9e266fdf4d5aed9 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..ca9d1750dc48e5b39af290118340b268d0d3dd0a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b25c616033492cff2f6d83011c726280ec581bdc9542e5bfc9175deb14d2286 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/special_tokens_map.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/tokenizer_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0b782d4602b361f7577abb52b710a8c0e088588d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/tokenizer_config.json @@ -0,0 +1,2078 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128257": { + "content": "🅁", + "lstrip": true, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/trainer_state.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..8d275d6cdb373bb06a8c26a6e9a523562c3ead96 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/trainer_state.json @@ -0,0 +1,10953 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.4381099126861838, + "eval_steps": 10, + "global_step": 1400, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7525, + "step": 1 + }, + { + "epoch": 0.002054442732408834, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7605, + "step": 2 + }, + { + "epoch": 0.0030816640986132513, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8008, + "step": 3 + }, + { + "epoch": 0.004108885464817668, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.809, + "step": 4 + }, + { + "epoch": 0.005136106831022085, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7841, + "step": 5 + }, + { + "epoch": 0.0061633281972265025, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7924, + "step": 6 + }, + { + "epoch": 0.007190549563430919, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7732, + "step": 7 + }, + { + "epoch": 0.008217770929635337, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7953, + "step": 8 + }, + { + "epoch": 0.009244992295839754, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7747, + "step": 9 + }, + { + "epoch": 0.01027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7909, + "step": 10 + }, + { + "epoch": 0.01027221366204417, + "eval_loss": 2.813098192214966, + "eval_runtime": 38.1187, + "eval_samples_per_second": 26.234, + "eval_steps_per_second": 1.102, + "step": 10 + }, + { + "epoch": 0.011299435028248588, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7592, + "step": 11 + }, + { + "epoch": 0.012326656394453005, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7394, + "step": 12 + }, + { + "epoch": 0.01335387776065742, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8111, + "step": 13 + }, + { + "epoch": 0.014381099126861838, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7786, + "step": 14 + }, + { + "epoch": 0.015408320493066256, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7711, + "step": 15 + }, + { + "epoch": 0.016435541859270673, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7622, + "step": 16 + }, + { + "epoch": 0.01746276322547509, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7291, + "step": 17 + }, + { + "epoch": 0.01848998459167951, + "grad_norm": 6.8249125480651855, + "learning_rate": 3.424657534246575e-08, + "loss": 2.7995, + "step": 18 + }, + { + "epoch": 0.019517205957883924, + "grad_norm": 5.581949710845947, + "learning_rate": 5.136986301369863e-08, + "loss": 2.5998, + "step": 19 + }, + { + "epoch": 0.02054442732408834, + "grad_norm": 3.0888617038726807, + "learning_rate": 6.84931506849315e-08, + "loss": 2.6031, + "step": 20 + }, + { + "epoch": 0.02054442732408834, + "eval_loss": 2.60642147064209, + "eval_runtime": 38.2226, + "eval_samples_per_second": 26.163, + "eval_steps_per_second": 1.099, + "step": 20 + }, + { + "epoch": 0.02157164869029276, + "grad_norm": 4.42941427230835, + "learning_rate": 8.561643835616439e-08, + "loss": 2.5816, + "step": 21 + }, + { + "epoch": 0.022598870056497175, + "grad_norm": 2.359813690185547, + "learning_rate": 1.0273972602739726e-07, + "loss": 2.5352, + "step": 22 + }, + { + "epoch": 0.02362609142270159, + "grad_norm": 2.9022927284240723, + "learning_rate": 1.1986301369863014e-07, + "loss": 2.4641, + "step": 23 + }, + { + "epoch": 0.02465331278890601, + "grad_norm": 1.6895997524261475, + "learning_rate": 1.36986301369863e-07, + "loss": 2.4582, + "step": 24 + }, + { + "epoch": 0.025680534155110426, + "grad_norm": 2.4555776119232178, + "learning_rate": 1.541095890410959e-07, + "loss": 2.4171, + "step": 25 + }, + { + "epoch": 0.02670775552131484, + "grad_norm": 1.9735779762268066, + "learning_rate": 1.7123287671232878e-07, + "loss": 2.4249, + "step": 26 + }, + { + "epoch": 0.02773497688751926, + "grad_norm": 2.71161150932312, + "learning_rate": 1.8835616438356165e-07, + "loss": 2.3791, + "step": 27 + }, + { + "epoch": 0.028762198253723677, + "grad_norm": 1.8482733964920044, + "learning_rate": 2.0547945205479452e-07, + "loss": 2.374, + "step": 28 + }, + { + "epoch": 0.029789419619928096, + "grad_norm": 2.191246747970581, + "learning_rate": 2.226027397260274e-07, + "loss": 2.3557, + "step": 29 + }, + { + "epoch": 0.030816640986132512, + "grad_norm": 2.0575971603393555, + "learning_rate": 2.397260273972603e-07, + "loss": 2.3315, + "step": 30 + }, + { + "epoch": 0.030816640986132512, + "eval_loss": 2.399639844894409, + "eval_runtime": 37.737, + "eval_samples_per_second": 26.499, + "eval_steps_per_second": 1.113, + "step": 30 + }, + { + "epoch": 0.03184386235233693, + "grad_norm": 2.0346758365631104, + "learning_rate": 2.568493150684932e-07, + "loss": 2.2791, + "step": 31 + }, + { + "epoch": 0.03287108371854135, + "grad_norm": 1.8710602521896362, + "learning_rate": 2.73972602739726e-07, + "loss": 2.3219, + "step": 32 + }, + { + "epoch": 0.03389830508474576, + "grad_norm": 1.9982675313949585, + "learning_rate": 2.910958904109589e-07, + "loss": 2.273, + "step": 33 + }, + { + "epoch": 0.03492552645095018, + "grad_norm": 2.092076301574707, + "learning_rate": 3.082191780821918e-07, + "loss": 2.2555, + "step": 34 + }, + { + "epoch": 0.035952747817154594, + "grad_norm": 1.870373249053955, + "learning_rate": 3.2534246575342466e-07, + "loss": 2.2314, + "step": 35 + }, + { + "epoch": 0.03697996918335902, + "grad_norm": 1.8378052711486816, + "learning_rate": 3.4246575342465755e-07, + "loss": 2.1244, + "step": 36 + }, + { + "epoch": 0.03800719054956343, + "grad_norm": 2.0156054496765137, + "learning_rate": 3.595890410958904e-07, + "loss": 2.2502, + "step": 37 + }, + { + "epoch": 0.03903441191576785, + "grad_norm": 1.6175312995910645, + "learning_rate": 3.767123287671233e-07, + "loss": 2.1566, + "step": 38 + }, + { + "epoch": 0.040061633281972264, + "grad_norm": 1.9051450490951538, + "learning_rate": 3.938356164383562e-07, + "loss": 2.167, + "step": 39 + }, + { + "epoch": 0.04108885464817668, + "grad_norm": 2.383063793182373, + "learning_rate": 4.1095890410958903e-07, + "loss": 2.2099, + "step": 40 + }, + { + "epoch": 0.04108885464817668, + "eval_loss": 2.2763354778289795, + "eval_runtime": 37.8085, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 40 + }, + { + "epoch": 0.042116076014381096, + "grad_norm": 2.1310720443725586, + "learning_rate": 4.2808219178082193e-07, + "loss": 2.1714, + "step": 41 + }, + { + "epoch": 0.04314329738058552, + "grad_norm": 1.771941900253296, + "learning_rate": 4.452054794520548e-07, + "loss": 2.0895, + "step": 42 + }, + { + "epoch": 0.044170518746789934, + "grad_norm": 2.3784172534942627, + "learning_rate": 4.6232876712328767e-07, + "loss": 2.1555, + "step": 43 + }, + { + "epoch": 0.04519774011299435, + "grad_norm": 1.6692813634872437, + "learning_rate": 4.794520547945206e-07, + "loss": 2.11, + "step": 44 + }, + { + "epoch": 0.046224961479198766, + "grad_norm": 2.1780643463134766, + "learning_rate": 4.965753424657534e-07, + "loss": 2.1256, + "step": 45 + }, + { + "epoch": 0.04725218284540318, + "grad_norm": 1.9390568733215332, + "learning_rate": 5.136986301369864e-07, + "loss": 2.0971, + "step": 46 + }, + { + "epoch": 0.048279404211607604, + "grad_norm": 2.276216506958008, + "learning_rate": 5.308219178082192e-07, + "loss": 2.1479, + "step": 47 + }, + { + "epoch": 0.04930662557781202, + "grad_norm": 2.148573160171509, + "learning_rate": 5.47945205479452e-07, + "loss": 2.0808, + "step": 48 + }, + { + "epoch": 0.050333846944016436, + "grad_norm": 2.259422540664673, + "learning_rate": 5.65068493150685e-07, + "loss": 2.0823, + "step": 49 + }, + { + "epoch": 0.05136106831022085, + "grad_norm": 1.7703359127044678, + "learning_rate": 5.821917808219178e-07, + "loss": 2.0236, + "step": 50 + }, + { + "epoch": 0.05136106831022085, + "eval_loss": 2.2037274837493896, + "eval_runtime": 37.8385, + "eval_samples_per_second": 26.428, + "eval_steps_per_second": 1.11, + "step": 50 + }, + { + "epoch": 0.05238828967642527, + "grad_norm": 2.56268310546875, + "learning_rate": 5.993150684931507e-07, + "loss": 2.0228, + "step": 51 + }, + { + "epoch": 0.05341551104262968, + "grad_norm": 2.5256848335266113, + "learning_rate": 6.164383561643836e-07, + "loss": 2.1401, + "step": 52 + }, + { + "epoch": 0.054442732408834106, + "grad_norm": 2.6329638957977295, + "learning_rate": 6.335616438356165e-07, + "loss": 2.0107, + "step": 53 + }, + { + "epoch": 0.05546995377503852, + "grad_norm": 2.477762222290039, + "learning_rate": 6.506849315068493e-07, + "loss": 2.0648, + "step": 54 + }, + { + "epoch": 0.05649717514124294, + "grad_norm": 1.8869198560714722, + "learning_rate": 6.678082191780823e-07, + "loss": 2.0214, + "step": 55 + }, + { + "epoch": 0.05752439650744735, + "grad_norm": 2.707345485687256, + "learning_rate": 6.849315068493151e-07, + "loss": 1.9893, + "step": 56 + }, + { + "epoch": 0.05855161787365177, + "grad_norm": 2.067230701446533, + "learning_rate": 7.020547945205481e-07, + "loss": 2.0308, + "step": 57 + }, + { + "epoch": 0.05957883923985619, + "grad_norm": 2.1961982250213623, + "learning_rate": 7.191780821917808e-07, + "loss": 1.948, + "step": 58 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 2.1032280921936035, + "learning_rate": 7.363013698630137e-07, + "loss": 2.0098, + "step": 59 + }, + { + "epoch": 0.061633281972265024, + "grad_norm": 2.5191144943237305, + "learning_rate": 7.534246575342466e-07, + "loss": 2.0183, + "step": 60 + }, + { + "epoch": 0.061633281972265024, + "eval_loss": 2.1474344730377197, + "eval_runtime": 37.8631, + "eval_samples_per_second": 26.411, + "eval_steps_per_second": 1.109, + "step": 60 + }, + { + "epoch": 0.06266050333846944, + "grad_norm": 2.177739143371582, + "learning_rate": 7.705479452054795e-07, + "loss": 1.9702, + "step": 61 + }, + { + "epoch": 0.06368772470467386, + "grad_norm": 2.499784231185913, + "learning_rate": 7.876712328767124e-07, + "loss": 1.9825, + "step": 62 + }, + { + "epoch": 0.06471494607087827, + "grad_norm": 2.181591749191284, + "learning_rate": 8.047945205479453e-07, + "loss": 2.0414, + "step": 63 + }, + { + "epoch": 0.0657421674370827, + "grad_norm": 1.9575999975204468, + "learning_rate": 8.219178082191781e-07, + "loss": 2.0115, + "step": 64 + }, + { + "epoch": 0.0667693888032871, + "grad_norm": 2.3431336879730225, + "learning_rate": 8.39041095890411e-07, + "loss": 1.8781, + "step": 65 + }, + { + "epoch": 0.06779661016949153, + "grad_norm": 2.042747974395752, + "learning_rate": 8.561643835616439e-07, + "loss": 1.9798, + "step": 66 + }, + { + "epoch": 0.06882383153569595, + "grad_norm": 2.784104585647583, + "learning_rate": 8.732876712328768e-07, + "loss": 1.9274, + "step": 67 + }, + { + "epoch": 0.06985105290190036, + "grad_norm": 1.9781445264816284, + "learning_rate": 8.904109589041097e-07, + "loss": 2.0036, + "step": 68 + }, + { + "epoch": 0.07087827426810478, + "grad_norm": 2.631537437438965, + "learning_rate": 9.075342465753426e-07, + "loss": 1.938, + "step": 69 + }, + { + "epoch": 0.07190549563430919, + "grad_norm": 1.902701497077942, + "learning_rate": 9.246575342465753e-07, + "loss": 2.0545, + "step": 70 + }, + { + "epoch": 0.07190549563430919, + "eval_loss": 2.115269422531128, + "eval_runtime": 37.887, + "eval_samples_per_second": 26.394, + "eval_steps_per_second": 1.109, + "step": 70 + }, + { + "epoch": 0.07293271700051361, + "grad_norm": 1.9323904514312744, + "learning_rate": 9.417808219178083e-07, + "loss": 1.9463, + "step": 71 + }, + { + "epoch": 0.07395993836671803, + "grad_norm": 2.1631956100463867, + "learning_rate": 9.589041095890411e-07, + "loss": 1.8982, + "step": 72 + }, + { + "epoch": 0.07498715973292244, + "grad_norm": 1.959518313407898, + "learning_rate": 9.76027397260274e-07, + "loss": 1.9538, + "step": 73 + }, + { + "epoch": 0.07601438109912687, + "grad_norm": 1.9834569692611694, + "learning_rate": 9.931506849315068e-07, + "loss": 1.9257, + "step": 74 + }, + { + "epoch": 0.07704160246533127, + "grad_norm": 2.2818732261657715, + "learning_rate": 1.0102739726027399e-06, + "loss": 1.9318, + "step": 75 + }, + { + "epoch": 0.0780688238315357, + "grad_norm": 1.7696406841278076, + "learning_rate": 1.0273972602739727e-06, + "loss": 1.912, + "step": 76 + }, + { + "epoch": 0.07909604519774012, + "grad_norm": 2.396303176879883, + "learning_rate": 1.0445205479452056e-06, + "loss": 1.9482, + "step": 77 + }, + { + "epoch": 0.08012326656394453, + "grad_norm": 2.2371881008148193, + "learning_rate": 1.0616438356164384e-06, + "loss": 1.9448, + "step": 78 + }, + { + "epoch": 0.08115048793014895, + "grad_norm": 2.40077543258667, + "learning_rate": 1.0787671232876712e-06, + "loss": 1.8764, + "step": 79 + }, + { + "epoch": 0.08217770929635336, + "grad_norm": 2.267916679382324, + "learning_rate": 1.095890410958904e-06, + "loss": 1.9478, + "step": 80 + }, + { + "epoch": 0.08217770929635336, + "eval_loss": 2.0851893424987793, + "eval_runtime": 37.9176, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 80 + }, + { + "epoch": 0.08320493066255778, + "grad_norm": 2.286259651184082, + "learning_rate": 1.1130136986301371e-06, + "loss": 1.8856, + "step": 81 + }, + { + "epoch": 0.08423215202876219, + "grad_norm": 2.160848379135132, + "learning_rate": 1.13013698630137e-06, + "loss": 1.9196, + "step": 82 + }, + { + "epoch": 0.08525937339496661, + "grad_norm": 2.3618931770324707, + "learning_rate": 1.1472602739726028e-06, + "loss": 1.9229, + "step": 83 + }, + { + "epoch": 0.08628659476117104, + "grad_norm": 2.1963961124420166, + "learning_rate": 1.1643835616438357e-06, + "loss": 1.8998, + "step": 84 + }, + { + "epoch": 0.08731381612737545, + "grad_norm": 2.197981357574463, + "learning_rate": 1.1815068493150685e-06, + "loss": 1.8716, + "step": 85 + }, + { + "epoch": 0.08834103749357987, + "grad_norm": 2.3562612533569336, + "learning_rate": 1.1986301369863014e-06, + "loss": 1.9336, + "step": 86 + }, + { + "epoch": 0.08936825885978428, + "grad_norm": 2.49712872505188, + "learning_rate": 1.2157534246575344e-06, + "loss": 1.8827, + "step": 87 + }, + { + "epoch": 0.0903954802259887, + "grad_norm": 2.315244436264038, + "learning_rate": 1.2328767123287673e-06, + "loss": 1.8801, + "step": 88 + }, + { + "epoch": 0.09142270159219312, + "grad_norm": 2.774587869644165, + "learning_rate": 1.25e-06, + "loss": 1.8972, + "step": 89 + }, + { + "epoch": 0.09244992295839753, + "grad_norm": 2.262364387512207, + "learning_rate": 1.267123287671233e-06, + "loss": 1.8426, + "step": 90 + }, + { + "epoch": 0.09244992295839753, + "eval_loss": 2.069462299346924, + "eval_runtime": 37.9319, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 90 + }, + { + "epoch": 0.09347714432460195, + "grad_norm": 2.6195597648620605, + "learning_rate": 1.284246575342466e-06, + "loss": 1.871, + "step": 91 + }, + { + "epoch": 0.09450436569080636, + "grad_norm": 2.051847219467163, + "learning_rate": 1.3013698630136986e-06, + "loss": 1.8922, + "step": 92 + }, + { + "epoch": 0.09553158705701079, + "grad_norm": 2.5930886268615723, + "learning_rate": 1.3184931506849317e-06, + "loss": 1.8818, + "step": 93 + }, + { + "epoch": 0.09655880842321521, + "grad_norm": 2.304164171218872, + "learning_rate": 1.3356164383561645e-06, + "loss": 1.8779, + "step": 94 + }, + { + "epoch": 0.09758602978941962, + "grad_norm": 2.187354564666748, + "learning_rate": 1.3527397260273976e-06, + "loss": 1.9272, + "step": 95 + }, + { + "epoch": 0.09861325115562404, + "grad_norm": 2.2492458820343018, + "learning_rate": 1.3698630136986302e-06, + "loss": 1.805, + "step": 96 + }, + { + "epoch": 0.09964047252182845, + "grad_norm": 2.3325352668762207, + "learning_rate": 1.386986301369863e-06, + "loss": 1.8817, + "step": 97 + }, + { + "epoch": 0.10066769388803287, + "grad_norm": 2.1708617210388184, + "learning_rate": 1.4041095890410961e-06, + "loss": 1.7956, + "step": 98 + }, + { + "epoch": 0.1016949152542373, + "grad_norm": 2.4267706871032715, + "learning_rate": 1.421232876712329e-06, + "loss": 1.8116, + "step": 99 + }, + { + "epoch": 0.1027221366204417, + "grad_norm": 2.3286805152893066, + "learning_rate": 1.4383561643835616e-06, + "loss": 1.8299, + "step": 100 + }, + { + "epoch": 0.1027221366204417, + "eval_loss": 2.0549135208129883, + "eval_runtime": 37.935, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 100 + }, + { + "epoch": 0.10374935798664613, + "grad_norm": 2.5391643047332764, + "learning_rate": 1.4554794520547946e-06, + "loss": 1.8521, + "step": 101 + }, + { + "epoch": 0.10477657935285054, + "grad_norm": 2.018380880355835, + "learning_rate": 1.4726027397260275e-06, + "loss": 1.8209, + "step": 102 + }, + { + "epoch": 0.10580380071905496, + "grad_norm": 2.3706746101379395, + "learning_rate": 1.4897260273972605e-06, + "loss": 1.79, + "step": 103 + }, + { + "epoch": 0.10683102208525937, + "grad_norm": 2.1290764808654785, + "learning_rate": 1.5068493150684932e-06, + "loss": 1.8141, + "step": 104 + }, + { + "epoch": 0.10785824345146379, + "grad_norm": 2.2768120765686035, + "learning_rate": 1.5239726027397262e-06, + "loss": 1.8826, + "step": 105 + }, + { + "epoch": 0.10888546481766821, + "grad_norm": 2.316584348678589, + "learning_rate": 1.541095890410959e-06, + "loss": 1.8604, + "step": 106 + }, + { + "epoch": 0.10991268618387262, + "grad_norm": 2.14679217338562, + "learning_rate": 1.5582191780821921e-06, + "loss": 1.865, + "step": 107 + }, + { + "epoch": 0.11093990755007704, + "grad_norm": 2.3288021087646484, + "learning_rate": 1.5753424657534248e-06, + "loss": 1.8512, + "step": 108 + }, + { + "epoch": 0.11196712891628145, + "grad_norm": 2.300482988357544, + "learning_rate": 1.5924657534246576e-06, + "loss": 1.7455, + "step": 109 + }, + { + "epoch": 0.11299435028248588, + "grad_norm": 2.4226229190826416, + "learning_rate": 1.6095890410958907e-06, + "loss": 1.8058, + "step": 110 + }, + { + "epoch": 0.11299435028248588, + "eval_loss": 2.025261640548706, + "eval_runtime": 37.6779, + "eval_samples_per_second": 26.541, + "eval_steps_per_second": 1.115, + "step": 110 + }, + { + "epoch": 0.1140215716486903, + "grad_norm": 2.0930111408233643, + "learning_rate": 1.6267123287671235e-06, + "loss": 1.7375, + "step": 111 + }, + { + "epoch": 0.1150487930148947, + "grad_norm": 2.4973645210266113, + "learning_rate": 1.6438356164383561e-06, + "loss": 1.8268, + "step": 112 + }, + { + "epoch": 0.11607601438109913, + "grad_norm": 2.3295114040374756, + "learning_rate": 1.6609589041095892e-06, + "loss": 1.7882, + "step": 113 + }, + { + "epoch": 0.11710323574730354, + "grad_norm": 2.3693318367004395, + "learning_rate": 1.678082191780822e-06, + "loss": 1.8192, + "step": 114 + }, + { + "epoch": 0.11813045711350796, + "grad_norm": 2.553637742996216, + "learning_rate": 1.695205479452055e-06, + "loss": 1.7491, + "step": 115 + }, + { + "epoch": 0.11915767847971238, + "grad_norm": 2.303945541381836, + "learning_rate": 1.7123287671232877e-06, + "loss": 1.8737, + "step": 116 + }, + { + "epoch": 0.12018489984591679, + "grad_norm": 2.250356435775757, + "learning_rate": 1.7294520547945206e-06, + "loss": 1.8039, + "step": 117 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 2.3854403495788574, + "learning_rate": 1.7465753424657536e-06, + "loss": 1.8357, + "step": 118 + }, + { + "epoch": 0.12223934257832562, + "grad_norm": 2.4049572944641113, + "learning_rate": 1.7636986301369865e-06, + "loss": 1.8209, + "step": 119 + }, + { + "epoch": 0.12326656394453005, + "grad_norm": 2.313551425933838, + "learning_rate": 1.7808219178082193e-06, + "loss": 1.9026, + "step": 120 + }, + { + "epoch": 0.12326656394453005, + "eval_loss": 2.012171745300293, + "eval_runtime": 37.7908, + "eval_samples_per_second": 26.461, + "eval_steps_per_second": 1.111, + "step": 120 + }, + { + "epoch": 0.12429378531073447, + "grad_norm": 2.518320322036743, + "learning_rate": 1.7979452054794521e-06, + "loss": 1.8241, + "step": 121 + }, + { + "epoch": 0.12532100667693888, + "grad_norm": 2.6504592895507812, + "learning_rate": 1.8150684931506852e-06, + "loss": 1.8091, + "step": 122 + }, + { + "epoch": 0.1263482280431433, + "grad_norm": 2.556610107421875, + "learning_rate": 1.832191780821918e-06, + "loss": 1.7547, + "step": 123 + }, + { + "epoch": 0.12737544940934772, + "grad_norm": 2.247633695602417, + "learning_rate": 1.8493150684931507e-06, + "loss": 1.7923, + "step": 124 + }, + { + "epoch": 0.12840267077555212, + "grad_norm": 2.195596933364868, + "learning_rate": 1.8664383561643837e-06, + "loss": 1.747, + "step": 125 + }, + { + "epoch": 0.12942989214175654, + "grad_norm": 2.121396780014038, + "learning_rate": 1.8835616438356166e-06, + "loss": 1.8088, + "step": 126 + }, + { + "epoch": 0.13045711350796096, + "grad_norm": 2.4572501182556152, + "learning_rate": 1.9006849315068496e-06, + "loss": 1.7285, + "step": 127 + }, + { + "epoch": 0.1314843348741654, + "grad_norm": 2.2590177059173584, + "learning_rate": 1.9178082191780823e-06, + "loss": 1.7929, + "step": 128 + }, + { + "epoch": 0.1325115562403698, + "grad_norm": 2.5099234580993652, + "learning_rate": 1.9349315068493153e-06, + "loss": 1.8365, + "step": 129 + }, + { + "epoch": 0.1335387776065742, + "grad_norm": 2.3064465522766113, + "learning_rate": 1.952054794520548e-06, + "loss": 1.8351, + "step": 130 + }, + { + "epoch": 0.1335387776065742, + "eval_loss": 2.0043866634368896, + "eval_runtime": 37.8661, + "eval_samples_per_second": 26.409, + "eval_steps_per_second": 1.109, + "step": 130 + }, + { + "epoch": 0.13456599897277863, + "grad_norm": 2.5977718830108643, + "learning_rate": 1.969178082191781e-06, + "loss": 1.7603, + "step": 131 + }, + { + "epoch": 0.13559322033898305, + "grad_norm": 2.23221755027771, + "learning_rate": 1.9863013698630136e-06, + "loss": 1.8237, + "step": 132 + }, + { + "epoch": 0.13662044170518747, + "grad_norm": 2.247044324874878, + "learning_rate": 2.0034246575342467e-06, + "loss": 1.8006, + "step": 133 + }, + { + "epoch": 0.1376476630713919, + "grad_norm": 2.1796107292175293, + "learning_rate": 2.0205479452054797e-06, + "loss": 1.7354, + "step": 134 + }, + { + "epoch": 0.1386748844375963, + "grad_norm": 2.3269693851470947, + "learning_rate": 2.037671232876713e-06, + "loss": 1.783, + "step": 135 + }, + { + "epoch": 0.1397021058038007, + "grad_norm": 2.2903053760528564, + "learning_rate": 2.0547945205479454e-06, + "loss": 1.8072, + "step": 136 + }, + { + "epoch": 0.14072932717000514, + "grad_norm": 2.634150981903076, + "learning_rate": 2.071917808219178e-06, + "loss": 1.7251, + "step": 137 + }, + { + "epoch": 0.14175654853620956, + "grad_norm": 2.42579984664917, + "learning_rate": 2.089041095890411e-06, + "loss": 1.7997, + "step": 138 + }, + { + "epoch": 0.14278376990241398, + "grad_norm": 2.615828514099121, + "learning_rate": 2.106164383561644e-06, + "loss": 1.766, + "step": 139 + }, + { + "epoch": 0.14381099126861838, + "grad_norm": 2.312714099884033, + "learning_rate": 2.123287671232877e-06, + "loss": 1.7538, + "step": 140 + }, + { + "epoch": 0.14381099126861838, + "eval_loss": 1.9918214082717896, + "eval_runtime": 37.8698, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 140 + }, + { + "epoch": 0.1448382126348228, + "grad_norm": 2.7540030479431152, + "learning_rate": 2.14041095890411e-06, + "loss": 1.7856, + "step": 141 + }, + { + "epoch": 0.14586543400102722, + "grad_norm": 2.289191246032715, + "learning_rate": 2.1575342465753425e-06, + "loss": 1.7288, + "step": 142 + }, + { + "epoch": 0.14689265536723164, + "grad_norm": 2.788001775741577, + "learning_rate": 2.1746575342465755e-06, + "loss": 1.8703, + "step": 143 + }, + { + "epoch": 0.14791987673343607, + "grad_norm": 2.4128854274749756, + "learning_rate": 2.191780821917808e-06, + "loss": 1.7461, + "step": 144 + }, + { + "epoch": 0.14894709809964046, + "grad_norm": 2.4108099937438965, + "learning_rate": 2.2089041095890412e-06, + "loss": 1.73, + "step": 145 + }, + { + "epoch": 0.14997431946584489, + "grad_norm": 2.4315948486328125, + "learning_rate": 2.2260273972602743e-06, + "loss": 1.7303, + "step": 146 + }, + { + "epoch": 0.1510015408320493, + "grad_norm": 2.494056224822998, + "learning_rate": 2.243150684931507e-06, + "loss": 1.7616, + "step": 147 + }, + { + "epoch": 0.15202876219825373, + "grad_norm": 2.4557394981384277, + "learning_rate": 2.26027397260274e-06, + "loss": 1.7655, + "step": 148 + }, + { + "epoch": 0.15305598356445815, + "grad_norm": 2.4114091396331787, + "learning_rate": 2.2773972602739726e-06, + "loss": 1.8538, + "step": 149 + }, + { + "epoch": 0.15408320493066255, + "grad_norm": 2.609219789505005, + "learning_rate": 2.2945205479452057e-06, + "loss": 1.7767, + "step": 150 + }, + { + "epoch": 0.15408320493066255, + "eval_loss": 1.981367588043213, + "eval_runtime": 37.9562, + "eval_samples_per_second": 26.346, + "eval_steps_per_second": 1.107, + "step": 150 + }, + { + "epoch": 0.15511042629686697, + "grad_norm": 2.213622808456421, + "learning_rate": 2.3116438356164387e-06, + "loss": 1.687, + "step": 151 + }, + { + "epoch": 0.1561376476630714, + "grad_norm": 2.6517975330352783, + "learning_rate": 2.3287671232876713e-06, + "loss": 1.7311, + "step": 152 + }, + { + "epoch": 0.15716486902927582, + "grad_norm": 2.5174458026885986, + "learning_rate": 2.3458904109589044e-06, + "loss": 1.7835, + "step": 153 + }, + { + "epoch": 0.15819209039548024, + "grad_norm": 2.485480308532715, + "learning_rate": 2.363013698630137e-06, + "loss": 1.7749, + "step": 154 + }, + { + "epoch": 0.15921931176168463, + "grad_norm": 2.5386061668395996, + "learning_rate": 2.38013698630137e-06, + "loss": 1.7223, + "step": 155 + }, + { + "epoch": 0.16024653312788906, + "grad_norm": 2.2895941734313965, + "learning_rate": 2.3972602739726027e-06, + "loss": 1.6845, + "step": 156 + }, + { + "epoch": 0.16127375449409348, + "grad_norm": 2.5453789234161377, + "learning_rate": 2.4143835616438358e-06, + "loss": 1.7627, + "step": 157 + }, + { + "epoch": 0.1623009758602979, + "grad_norm": 2.5322303771972656, + "learning_rate": 2.431506849315069e-06, + "loss": 1.764, + "step": 158 + }, + { + "epoch": 0.1633281972265023, + "grad_norm": 2.2669899463653564, + "learning_rate": 2.4486301369863015e-06, + "loss": 1.7487, + "step": 159 + }, + { + "epoch": 0.16435541859270672, + "grad_norm": 2.4577853679656982, + "learning_rate": 2.4657534246575345e-06, + "loss": 1.7635, + "step": 160 + }, + { + "epoch": 0.16435541859270672, + "eval_loss": 1.9793355464935303, + "eval_runtime": 37.8576, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 160 + }, + { + "epoch": 0.16538263995891114, + "grad_norm": 2.354893922805786, + "learning_rate": 2.482876712328767e-06, + "loss": 1.7538, + "step": 161 + }, + { + "epoch": 0.16640986132511557, + "grad_norm": 2.3255856037139893, + "learning_rate": 2.5e-06, + "loss": 1.7277, + "step": 162 + }, + { + "epoch": 0.16743708269132, + "grad_norm": 2.3501994609832764, + "learning_rate": 2.5171232876712333e-06, + "loss": 1.6985, + "step": 163 + }, + { + "epoch": 0.16846430405752438, + "grad_norm": 2.386627674102783, + "learning_rate": 2.534246575342466e-06, + "loss": 1.7552, + "step": 164 + }, + { + "epoch": 0.1694915254237288, + "grad_norm": 2.2238457202911377, + "learning_rate": 2.551369863013699e-06, + "loss": 1.7805, + "step": 165 + }, + { + "epoch": 0.17051874678993323, + "grad_norm": 2.2653403282165527, + "learning_rate": 2.568493150684932e-06, + "loss": 1.7656, + "step": 166 + }, + { + "epoch": 0.17154596815613765, + "grad_norm": 2.233132839202881, + "learning_rate": 2.585616438356164e-06, + "loss": 1.6997, + "step": 167 + }, + { + "epoch": 0.17257318952234207, + "grad_norm": 2.1696043014526367, + "learning_rate": 2.6027397260273973e-06, + "loss": 1.7487, + "step": 168 + }, + { + "epoch": 0.17360041088854647, + "grad_norm": 2.4479899406433105, + "learning_rate": 2.6198630136986303e-06, + "loss": 1.7957, + "step": 169 + }, + { + "epoch": 0.1746276322547509, + "grad_norm": 2.3054702281951904, + "learning_rate": 2.6369863013698634e-06, + "loss": 1.766, + "step": 170 + }, + { + "epoch": 0.1746276322547509, + "eval_loss": 1.9543790817260742, + "eval_runtime": 37.8837, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 170 + }, + { + "epoch": 0.17565485362095531, + "grad_norm": 2.3591110706329346, + "learning_rate": 2.654109589041096e-06, + "loss": 1.7149, + "step": 171 + }, + { + "epoch": 0.17668207498715974, + "grad_norm": 2.450526237487793, + "learning_rate": 2.671232876712329e-06, + "loss": 1.7074, + "step": 172 + }, + { + "epoch": 0.17770929635336416, + "grad_norm": 2.1511754989624023, + "learning_rate": 2.688356164383562e-06, + "loss": 1.7036, + "step": 173 + }, + { + "epoch": 0.17873651771956856, + "grad_norm": 2.5944275856018066, + "learning_rate": 2.705479452054795e-06, + "loss": 1.6566, + "step": 174 + }, + { + "epoch": 0.17976373908577298, + "grad_norm": 2.4301369190216064, + "learning_rate": 2.7226027397260274e-06, + "loss": 1.7977, + "step": 175 + }, + { + "epoch": 0.1807909604519774, + "grad_norm": 2.287665367126465, + "learning_rate": 2.7397260273972604e-06, + "loss": 1.7526, + "step": 176 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.380032777786255, + "learning_rate": 2.7568493150684935e-06, + "loss": 1.6753, + "step": 177 + }, + { + "epoch": 0.18284540318438625, + "grad_norm": 2.4040510654449463, + "learning_rate": 2.773972602739726e-06, + "loss": 1.758, + "step": 178 + }, + { + "epoch": 0.18387262455059064, + "grad_norm": 2.5074045658111572, + "learning_rate": 2.791095890410959e-06, + "loss": 1.6498, + "step": 179 + }, + { + "epoch": 0.18489984591679506, + "grad_norm": 2.3926949501037598, + "learning_rate": 2.8082191780821922e-06, + "loss": 1.8006, + "step": 180 + }, + { + "epoch": 0.18489984591679506, + "eval_loss": 1.9446762800216675, + "eval_runtime": 37.8572, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 180 + }, + { + "epoch": 0.1859270672829995, + "grad_norm": 2.4991633892059326, + "learning_rate": 2.8253424657534253e-06, + "loss": 1.7649, + "step": 181 + }, + { + "epoch": 0.1869542886492039, + "grad_norm": 2.2558324337005615, + "learning_rate": 2.842465753424658e-06, + "loss": 1.7743, + "step": 182 + }, + { + "epoch": 0.18798151001540833, + "grad_norm": 2.219717025756836, + "learning_rate": 2.8595890410958905e-06, + "loss": 1.7807, + "step": 183 + }, + { + "epoch": 0.18900873138161273, + "grad_norm": 2.406562566757202, + "learning_rate": 2.876712328767123e-06, + "loss": 1.6289, + "step": 184 + }, + { + "epoch": 0.19003595274781715, + "grad_norm": 2.356579303741455, + "learning_rate": 2.8938356164383562e-06, + "loss": 1.6747, + "step": 185 + }, + { + "epoch": 0.19106317411402157, + "grad_norm": 2.4844322204589844, + "learning_rate": 2.9109589041095893e-06, + "loss": 1.6731, + "step": 186 + }, + { + "epoch": 0.192090395480226, + "grad_norm": 2.8179149627685547, + "learning_rate": 2.9280821917808223e-06, + "loss": 1.6786, + "step": 187 + }, + { + "epoch": 0.19311761684643042, + "grad_norm": 2.395416259765625, + "learning_rate": 2.945205479452055e-06, + "loss": 1.7313, + "step": 188 + }, + { + "epoch": 0.1941448382126348, + "grad_norm": 2.848740816116333, + "learning_rate": 2.962328767123288e-06, + "loss": 1.7258, + "step": 189 + }, + { + "epoch": 0.19517205957883924, + "grad_norm": 2.427395820617676, + "learning_rate": 2.979452054794521e-06, + "loss": 1.6633, + "step": 190 + }, + { + "epoch": 0.19517205957883924, + "eval_loss": 1.9443999528884888, + "eval_runtime": 37.8579, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 190 + }, + { + "epoch": 0.19619928094504366, + "grad_norm": 2.5347256660461426, + "learning_rate": 2.9965753424657533e-06, + "loss": 1.7342, + "step": 191 + }, + { + "epoch": 0.19722650231124808, + "grad_norm": 2.8696866035461426, + "learning_rate": 3.0136986301369864e-06, + "loss": 1.7037, + "step": 192 + }, + { + "epoch": 0.1982537236774525, + "grad_norm": 2.427468776702881, + "learning_rate": 3.0308219178082194e-06, + "loss": 1.672, + "step": 193 + }, + { + "epoch": 0.1992809450436569, + "grad_norm": 2.4253737926483154, + "learning_rate": 3.0479452054794525e-06, + "loss": 1.7252, + "step": 194 + }, + { + "epoch": 0.20030816640986132, + "grad_norm": 2.801668643951416, + "learning_rate": 3.065068493150685e-06, + "loss": 1.6694, + "step": 195 + }, + { + "epoch": 0.20133538777606574, + "grad_norm": 2.442263603210449, + "learning_rate": 3.082191780821918e-06, + "loss": 1.6866, + "step": 196 + }, + { + "epoch": 0.20236260914227017, + "grad_norm": 2.777463674545288, + "learning_rate": 3.099315068493151e-06, + "loss": 1.686, + "step": 197 + }, + { + "epoch": 0.2033898305084746, + "grad_norm": 2.491227388381958, + "learning_rate": 3.1164383561643843e-06, + "loss": 1.6414, + "step": 198 + }, + { + "epoch": 0.20441705187467898, + "grad_norm": 2.33585262298584, + "learning_rate": 3.1335616438356165e-06, + "loss": 1.7808, + "step": 199 + }, + { + "epoch": 0.2054442732408834, + "grad_norm": 2.518160581588745, + "learning_rate": 3.1506849315068495e-06, + "loss": 1.6592, + "step": 200 + }, + { + "epoch": 0.2054442732408834, + "eval_loss": 1.9371097087860107, + "eval_runtime": 37.8602, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 200 + }, + { + "epoch": 0.20647149460708783, + "grad_norm": 2.460204839706421, + "learning_rate": 3.167808219178082e-06, + "loss": 1.6668, + "step": 201 + }, + { + "epoch": 0.20749871597329225, + "grad_norm": 2.377824068069458, + "learning_rate": 3.184931506849315e-06, + "loss": 1.6462, + "step": 202 + }, + { + "epoch": 0.20852593733949668, + "grad_norm": 2.426177740097046, + "learning_rate": 3.2020547945205483e-06, + "loss": 1.6272, + "step": 203 + }, + { + "epoch": 0.20955315870570107, + "grad_norm": 2.4532034397125244, + "learning_rate": 3.2191780821917813e-06, + "loss": 1.6169, + "step": 204 + }, + { + "epoch": 0.2105803800719055, + "grad_norm": 2.517059564590454, + "learning_rate": 3.236301369863014e-06, + "loss": 1.5991, + "step": 205 + }, + { + "epoch": 0.21160760143810992, + "grad_norm": 2.22658109664917, + "learning_rate": 3.253424657534247e-06, + "loss": 1.6754, + "step": 206 + }, + { + "epoch": 0.21263482280431434, + "grad_norm": 2.344799518585205, + "learning_rate": 3.2705479452054796e-06, + "loss": 1.6648, + "step": 207 + }, + { + "epoch": 0.21366204417051873, + "grad_norm": 2.331875801086426, + "learning_rate": 3.2876712328767123e-06, + "loss": 1.6112, + "step": 208 + }, + { + "epoch": 0.21468926553672316, + "grad_norm": 2.262343645095825, + "learning_rate": 3.3047945205479453e-06, + "loss": 1.6479, + "step": 209 + }, + { + "epoch": 0.21571648690292758, + "grad_norm": 2.282278537750244, + "learning_rate": 3.3219178082191784e-06, + "loss": 1.6753, + "step": 210 + }, + { + "epoch": 0.21571648690292758, + "eval_loss": 1.9338418245315552, + "eval_runtime": 37.6083, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 210 + }, + { + "epoch": 0.216743708269132, + "grad_norm": 2.578476905822754, + "learning_rate": 3.3390410958904114e-06, + "loss": 1.7043, + "step": 211 + }, + { + "epoch": 0.21777092963533642, + "grad_norm": 2.3118813037872314, + "learning_rate": 3.356164383561644e-06, + "loss": 1.7541, + "step": 212 + }, + { + "epoch": 0.21879815100154082, + "grad_norm": 2.2418692111968994, + "learning_rate": 3.373287671232877e-06, + "loss": 1.6886, + "step": 213 + }, + { + "epoch": 0.21982537236774524, + "grad_norm": 2.3583438396453857, + "learning_rate": 3.39041095890411e-06, + "loss": 1.7241, + "step": 214 + }, + { + "epoch": 0.22085259373394966, + "grad_norm": 2.345090866088867, + "learning_rate": 3.4075342465753424e-06, + "loss": 1.6194, + "step": 215 + }, + { + "epoch": 0.2218798151001541, + "grad_norm": 2.3996200561523438, + "learning_rate": 3.4246575342465754e-06, + "loss": 1.6889, + "step": 216 + }, + { + "epoch": 0.2229070364663585, + "grad_norm": 2.3552374839782715, + "learning_rate": 3.4417808219178085e-06, + "loss": 1.6713, + "step": 217 + }, + { + "epoch": 0.2239342578325629, + "grad_norm": 2.2653744220733643, + "learning_rate": 3.458904109589041e-06, + "loss": 1.673, + "step": 218 + }, + { + "epoch": 0.22496147919876733, + "grad_norm": 2.263288974761963, + "learning_rate": 3.476027397260274e-06, + "loss": 1.673, + "step": 219 + }, + { + "epoch": 0.22598870056497175, + "grad_norm": 2.3173880577087402, + "learning_rate": 3.4931506849315072e-06, + "loss": 1.6803, + "step": 220 + }, + { + "epoch": 0.22598870056497175, + "eval_loss": 1.9166476726531982, + "eval_runtime": 37.7659, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 220 + }, + { + "epoch": 0.22701592193117617, + "grad_norm": 2.4005839824676514, + "learning_rate": 3.5102739726027403e-06, + "loss": 1.6809, + "step": 221 + }, + { + "epoch": 0.2280431432973806, + "grad_norm": 2.5032525062561035, + "learning_rate": 3.527397260273973e-06, + "loss": 1.6411, + "step": 222 + }, + { + "epoch": 0.229070364663585, + "grad_norm": 2.533583879470825, + "learning_rate": 3.5445205479452056e-06, + "loss": 1.6608, + "step": 223 + }, + { + "epoch": 0.2300975860297894, + "grad_norm": 2.3575572967529297, + "learning_rate": 3.5616438356164386e-06, + "loss": 1.7236, + "step": 224 + }, + { + "epoch": 0.23112480739599384, + "grad_norm": 2.3408942222595215, + "learning_rate": 3.5787671232876712e-06, + "loss": 1.7101, + "step": 225 + }, + { + "epoch": 0.23215202876219826, + "grad_norm": 2.3819596767425537, + "learning_rate": 3.5958904109589043e-06, + "loss": 1.6414, + "step": 226 + }, + { + "epoch": 0.23317925012840268, + "grad_norm": 2.2874338626861572, + "learning_rate": 3.6130136986301373e-06, + "loss": 1.6767, + "step": 227 + }, + { + "epoch": 0.23420647149460708, + "grad_norm": 2.3523948192596436, + "learning_rate": 3.6301369863013704e-06, + "loss": 1.6932, + "step": 228 + }, + { + "epoch": 0.2352336928608115, + "grad_norm": 2.476128578186035, + "learning_rate": 3.647260273972603e-06, + "loss": 1.7125, + "step": 229 + }, + { + "epoch": 0.23626091422701592, + "grad_norm": 2.3841593265533447, + "learning_rate": 3.664383561643836e-06, + "loss": 1.6469, + "step": 230 + }, + { + "epoch": 0.23626091422701592, + "eval_loss": 1.925522804260254, + "eval_runtime": 37.8581, + "eval_samples_per_second": 26.414, + "eval_steps_per_second": 1.109, + "step": 230 + }, + { + "epoch": 0.23728813559322035, + "grad_norm": 2.4401519298553467, + "learning_rate": 3.6815068493150687e-06, + "loss": 1.6726, + "step": 231 + }, + { + "epoch": 0.23831535695942477, + "grad_norm": 2.329674005508423, + "learning_rate": 3.6986301369863014e-06, + "loss": 1.6349, + "step": 232 + }, + { + "epoch": 0.23934257832562916, + "grad_norm": 2.654290199279785, + "learning_rate": 3.7157534246575344e-06, + "loss": 1.7309, + "step": 233 + }, + { + "epoch": 0.24036979969183359, + "grad_norm": 2.1599373817443848, + "learning_rate": 3.7328767123287675e-06, + "loss": 1.6437, + "step": 234 + }, + { + "epoch": 0.241397021058038, + "grad_norm": 2.6170263290405273, + "learning_rate": 3.7500000000000005e-06, + "loss": 1.6862, + "step": 235 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 2.2836415767669678, + "learning_rate": 3.767123287671233e-06, + "loss": 1.7073, + "step": 236 + }, + { + "epoch": 0.24345146379044685, + "grad_norm": 2.3001456260681152, + "learning_rate": 3.784246575342466e-06, + "loss": 1.6455, + "step": 237 + }, + { + "epoch": 0.24447868515665125, + "grad_norm": 2.4549131393432617, + "learning_rate": 3.8013698630136993e-06, + "loss": 1.6566, + "step": 238 + }, + { + "epoch": 0.24550590652285567, + "grad_norm": 2.4387831687927246, + "learning_rate": 3.818493150684932e-06, + "loss": 1.627, + "step": 239 + }, + { + "epoch": 0.2465331278890601, + "grad_norm": 2.136474847793579, + "learning_rate": 3.8356164383561645e-06, + "loss": 1.7286, + "step": 240 + }, + { + "epoch": 0.2465331278890601, + "eval_loss": 1.9188542366027832, + "eval_runtime": 37.9079, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 240 + }, + { + "epoch": 0.24756034925526452, + "grad_norm": 2.430297374725342, + "learning_rate": 3.852739726027397e-06, + "loss": 1.7014, + "step": 241 + }, + { + "epoch": 0.24858757062146894, + "grad_norm": 2.302480697631836, + "learning_rate": 3.869863013698631e-06, + "loss": 1.6412, + "step": 242 + }, + { + "epoch": 0.24961479198767333, + "grad_norm": 2.2905426025390625, + "learning_rate": 3.886986301369863e-06, + "loss": 1.6445, + "step": 243 + }, + { + "epoch": 0.25064201335387776, + "grad_norm": 2.55385684967041, + "learning_rate": 3.904109589041096e-06, + "loss": 1.6265, + "step": 244 + }, + { + "epoch": 0.2516692347200822, + "grad_norm": 2.3053576946258545, + "learning_rate": 3.921232876712329e-06, + "loss": 1.6338, + "step": 245 + }, + { + "epoch": 0.2526964560862866, + "grad_norm": 2.4894707202911377, + "learning_rate": 3.938356164383562e-06, + "loss": 1.5966, + "step": 246 + }, + { + "epoch": 0.253723677452491, + "grad_norm": 2.6198930740356445, + "learning_rate": 3.9554794520547955e-06, + "loss": 1.6935, + "step": 247 + }, + { + "epoch": 0.25475089881869545, + "grad_norm": 2.3548998832702637, + "learning_rate": 3.972602739726027e-06, + "loss": 1.6227, + "step": 248 + }, + { + "epoch": 0.25577812018489987, + "grad_norm": 2.535460948944092, + "learning_rate": 3.989726027397261e-06, + "loss": 1.6699, + "step": 249 + }, + { + "epoch": 0.25680534155110424, + "grad_norm": 2.3680567741394043, + "learning_rate": 4.006849315068493e-06, + "loss": 1.6666, + "step": 250 + }, + { + "epoch": 0.25680534155110424, + "eval_loss": 1.9149967432022095, + "eval_runtime": 37.9423, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 250 + }, + { + "epoch": 0.25783256291730866, + "grad_norm": 2.3914358615875244, + "learning_rate": 4.023972602739726e-06, + "loss": 1.6633, + "step": 251 + }, + { + "epoch": 0.2588597842835131, + "grad_norm": 2.4293038845062256, + "learning_rate": 4.0410958904109595e-06, + "loss": 1.6573, + "step": 252 + }, + { + "epoch": 0.2598870056497175, + "grad_norm": 2.3409900665283203, + "learning_rate": 4.058219178082192e-06, + "loss": 1.6815, + "step": 253 + }, + { + "epoch": 0.26091422701592193, + "grad_norm": 2.5807456970214844, + "learning_rate": 4.075342465753426e-06, + "loss": 1.6412, + "step": 254 + }, + { + "epoch": 0.26194144838212635, + "grad_norm": 2.317861795425415, + "learning_rate": 4.092465753424658e-06, + "loss": 1.6311, + "step": 255 + }, + { + "epoch": 0.2629686697483308, + "grad_norm": 2.231682777404785, + "learning_rate": 4.109589041095891e-06, + "loss": 1.7021, + "step": 256 + }, + { + "epoch": 0.2639958911145352, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.1267123287671235e-06, + "loss": 1.6446, + "step": 257 + }, + { + "epoch": 0.2650231124807396, + "grad_norm": 2.2619071006774902, + "learning_rate": 4.143835616438356e-06, + "loss": 1.7101, + "step": 258 + }, + { + "epoch": 0.26605033384694404, + "grad_norm": 2.461945056915283, + "learning_rate": 4.16095890410959e-06, + "loss": 1.6901, + "step": 259 + }, + { + "epoch": 0.2670775552131484, + "grad_norm": 2.516348123550415, + "learning_rate": 4.178082191780822e-06, + "loss": 1.644, + "step": 260 + }, + { + "epoch": 0.2670775552131484, + "eval_loss": 1.904897928237915, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 260 + }, + { + "epoch": 0.26810477657935283, + "grad_norm": 2.2311813831329346, + "learning_rate": 4.195205479452055e-06, + "loss": 1.6301, + "step": 261 + }, + { + "epoch": 0.26913199794555726, + "grad_norm": 2.605109214782715, + "learning_rate": 4.212328767123288e-06, + "loss": 1.6393, + "step": 262 + }, + { + "epoch": 0.2701592193117617, + "grad_norm": 2.6103134155273438, + "learning_rate": 4.229452054794521e-06, + "loss": 1.6468, + "step": 263 + }, + { + "epoch": 0.2711864406779661, + "grad_norm": 2.5538523197174072, + "learning_rate": 4.246575342465754e-06, + "loss": 1.6294, + "step": 264 + }, + { + "epoch": 0.2722136620441705, + "grad_norm": 2.5841965675354004, + "learning_rate": 4.263698630136986e-06, + "loss": 1.5837, + "step": 265 + }, + { + "epoch": 0.27324088341037495, + "grad_norm": 2.4604523181915283, + "learning_rate": 4.28082191780822e-06, + "loss": 1.7473, + "step": 266 + }, + { + "epoch": 0.27426810477657937, + "grad_norm": 2.6701765060424805, + "learning_rate": 4.297945205479452e-06, + "loss": 1.6539, + "step": 267 + }, + { + "epoch": 0.2752953261427838, + "grad_norm": 2.360069990158081, + "learning_rate": 4.315068493150685e-06, + "loss": 1.6385, + "step": 268 + }, + { + "epoch": 0.2763225475089882, + "grad_norm": 2.4922797679901123, + "learning_rate": 4.3321917808219185e-06, + "loss": 1.6586, + "step": 269 + }, + { + "epoch": 0.2773497688751926, + "grad_norm": 2.5443379878997803, + "learning_rate": 4.349315068493151e-06, + "loss": 1.6318, + "step": 270 + }, + { + "epoch": 0.2773497688751926, + "eval_loss": 1.9062609672546387, + "eval_runtime": 37.92, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 270 + }, + { + "epoch": 0.278376990241397, + "grad_norm": 2.4064481258392334, + "learning_rate": 4.3664383561643846e-06, + "loss": 1.6178, + "step": 271 + }, + { + "epoch": 0.2794042116076014, + "grad_norm": 2.681637763977051, + "learning_rate": 4.383561643835616e-06, + "loss": 1.7071, + "step": 272 + }, + { + "epoch": 0.28043143297380585, + "grad_norm": 2.5805184841156006, + "learning_rate": 4.40068493150685e-06, + "loss": 1.6553, + "step": 273 + }, + { + "epoch": 0.2814586543400103, + "grad_norm": 2.4169883728027344, + "learning_rate": 4.4178082191780825e-06, + "loss": 1.6249, + "step": 274 + }, + { + "epoch": 0.2824858757062147, + "grad_norm": 2.557570457458496, + "learning_rate": 4.434931506849315e-06, + "loss": 1.5932, + "step": 275 + }, + { + "epoch": 0.2835130970724191, + "grad_norm": 2.3808107376098633, + "learning_rate": 4.4520547945205486e-06, + "loss": 1.6377, + "step": 276 + }, + { + "epoch": 0.28454031843862354, + "grad_norm": 2.402404546737671, + "learning_rate": 4.469178082191781e-06, + "loss": 1.6285, + "step": 277 + }, + { + "epoch": 0.28556753980482796, + "grad_norm": 2.4676456451416016, + "learning_rate": 4.486301369863014e-06, + "loss": 1.6838, + "step": 278 + }, + { + "epoch": 0.28659476117103233, + "grad_norm": 2.659083366394043, + "learning_rate": 4.503424657534247e-06, + "loss": 1.5872, + "step": 279 + }, + { + "epoch": 0.28762198253723675, + "grad_norm": 2.326733350753784, + "learning_rate": 4.52054794520548e-06, + "loss": 1.6142, + "step": 280 + }, + { + "epoch": 0.28762198253723675, + "eval_loss": 1.904759407043457, + "eval_runtime": 37.9651, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 280 + }, + { + "epoch": 0.2886492039034412, + "grad_norm": 2.299504518508911, + "learning_rate": 4.537671232876713e-06, + "loss": 1.6673, + "step": 281 + }, + { + "epoch": 0.2896764252696456, + "grad_norm": 2.4932825565338135, + "learning_rate": 4.554794520547945e-06, + "loss": 1.6586, + "step": 282 + }, + { + "epoch": 0.29070364663585, + "grad_norm": 2.2599050998687744, + "learning_rate": 4.571917808219179e-06, + "loss": 1.6568, + "step": 283 + }, + { + "epoch": 0.29173086800205444, + "grad_norm": 2.435098886489868, + "learning_rate": 4.589041095890411e-06, + "loss": 1.6113, + "step": 284 + }, + { + "epoch": 0.29275808936825887, + "grad_norm": 2.539246082305908, + "learning_rate": 4.606164383561644e-06, + "loss": 1.6306, + "step": 285 + }, + { + "epoch": 0.2937853107344633, + "grad_norm": 2.596198320388794, + "learning_rate": 4.6232876712328774e-06, + "loss": 1.6473, + "step": 286 + }, + { + "epoch": 0.2948125321006677, + "grad_norm": 2.349133014678955, + "learning_rate": 4.64041095890411e-06, + "loss": 1.6042, + "step": 287 + }, + { + "epoch": 0.29583975346687214, + "grad_norm": 2.3572661876678467, + "learning_rate": 4.657534246575343e-06, + "loss": 1.6635, + "step": 288 + }, + { + "epoch": 0.2968669748330765, + "grad_norm": 2.4210307598114014, + "learning_rate": 4.674657534246575e-06, + "loss": 1.624, + "step": 289 + }, + { + "epoch": 0.2978941961992809, + "grad_norm": 2.427403688430786, + "learning_rate": 4.691780821917809e-06, + "loss": 1.6824, + "step": 290 + }, + { + "epoch": 0.2978941961992809, + "eval_loss": 1.8888994455337524, + "eval_runtime": 37.9777, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 290 + }, + { + "epoch": 0.29892141756548535, + "grad_norm": 2.534029722213745, + "learning_rate": 4.7089041095890414e-06, + "loss": 1.6456, + "step": 291 + }, + { + "epoch": 0.29994863893168977, + "grad_norm": 2.329305648803711, + "learning_rate": 4.726027397260274e-06, + "loss": 1.646, + "step": 292 + }, + { + "epoch": 0.3009758602978942, + "grad_norm": 2.561548948287964, + "learning_rate": 4.7431506849315075e-06, + "loss": 1.6511, + "step": 293 + }, + { + "epoch": 0.3020030816640986, + "grad_norm": 2.707712411880493, + "learning_rate": 4.76027397260274e-06, + "loss": 1.5984, + "step": 294 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 2.403089761734009, + "learning_rate": 4.777397260273973e-06, + "loss": 1.6164, + "step": 295 + }, + { + "epoch": 0.30405752439650746, + "grad_norm": 2.424910306930542, + "learning_rate": 4.7945205479452054e-06, + "loss": 1.6291, + "step": 296 + }, + { + "epoch": 0.3050847457627119, + "grad_norm": 2.3474302291870117, + "learning_rate": 4.811643835616439e-06, + "loss": 1.553, + "step": 297 + }, + { + "epoch": 0.3061119671289163, + "grad_norm": 2.2211227416992188, + "learning_rate": 4.8287671232876716e-06, + "loss": 1.6108, + "step": 298 + }, + { + "epoch": 0.3071391884951207, + "grad_norm": 2.495983600616455, + "learning_rate": 4.845890410958904e-06, + "loss": 1.6135, + "step": 299 + }, + { + "epoch": 0.3081664098613251, + "grad_norm": 2.525495767593384, + "learning_rate": 4.863013698630138e-06, + "loss": 1.6061, + "step": 300 + }, + { + "epoch": 0.3081664098613251, + "eval_loss": 1.8982168436050415, + "eval_runtime": 37.9848, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 300 + }, + { + "epoch": 0.3091936312275295, + "grad_norm": 2.449064254760742, + "learning_rate": 4.88013698630137e-06, + "loss": 1.5827, + "step": 301 + }, + { + "epoch": 0.31022085259373394, + "grad_norm": 2.288095712661743, + "learning_rate": 4.897260273972603e-06, + "loss": 1.704, + "step": 302 + }, + { + "epoch": 0.31124807395993837, + "grad_norm": 2.5301458835601807, + "learning_rate": 4.914383561643836e-06, + "loss": 1.6342, + "step": 303 + }, + { + "epoch": 0.3122752953261428, + "grad_norm": 2.2793986797332764, + "learning_rate": 4.931506849315069e-06, + "loss": 1.6688, + "step": 304 + }, + { + "epoch": 0.3133025166923472, + "grad_norm": 2.38783597946167, + "learning_rate": 4.948630136986302e-06, + "loss": 1.7014, + "step": 305 + }, + { + "epoch": 0.31432973805855163, + "grad_norm": 2.5224838256835938, + "learning_rate": 4.965753424657534e-06, + "loss": 1.5812, + "step": 306 + }, + { + "epoch": 0.31535695942475606, + "grad_norm": 2.434432029724121, + "learning_rate": 4.982876712328768e-06, + "loss": 1.6186, + "step": 307 + }, + { + "epoch": 0.3163841807909605, + "grad_norm": 2.3708460330963135, + "learning_rate": 5e-06, + "loss": 1.5669, + "step": 308 + }, + { + "epoch": 0.31741140215716485, + "grad_norm": 2.5524990558624268, + "learning_rate": 4.998096688237534e-06, + "loss": 1.6688, + "step": 309 + }, + { + "epoch": 0.31843862352336927, + "grad_norm": 2.322441577911377, + "learning_rate": 4.996193376475067e-06, + "loss": 1.5571, + "step": 310 + }, + { + "epoch": 0.31843862352336927, + "eval_loss": 1.8914265632629395, + "eval_runtime": 37.6505, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 310 + }, + { + "epoch": 0.3194658448895737, + "grad_norm": 2.5248687267303467, + "learning_rate": 4.9942900647126005e-06, + "loss": 1.5585, + "step": 311 + }, + { + "epoch": 0.3204930662557781, + "grad_norm": 2.2464914321899414, + "learning_rate": 4.992386752950134e-06, + "loss": 1.5921, + "step": 312 + }, + { + "epoch": 0.32152028762198254, + "grad_norm": 2.5355331897735596, + "learning_rate": 4.990483441187667e-06, + "loss": 1.6263, + "step": 313 + }, + { + "epoch": 0.32254750898818696, + "grad_norm": 2.5611143112182617, + "learning_rate": 4.9885801294252e-06, + "loss": 1.6267, + "step": 314 + }, + { + "epoch": 0.3235747303543914, + "grad_norm": 2.364356517791748, + "learning_rate": 4.986676817662733e-06, + "loss": 1.5847, + "step": 315 + }, + { + "epoch": 0.3246019517205958, + "grad_norm": 2.455165147781372, + "learning_rate": 4.9847735059002665e-06, + "loss": 1.661, + "step": 316 + }, + { + "epoch": 0.3256291730868002, + "grad_norm": 2.4312195777893066, + "learning_rate": 4.9828701941378e-06, + "loss": 1.5995, + "step": 317 + }, + { + "epoch": 0.3266563944530046, + "grad_norm": 2.4081499576568604, + "learning_rate": 4.980966882375333e-06, + "loss": 1.5756, + "step": 318 + }, + { + "epoch": 0.327683615819209, + "grad_norm": 2.2608511447906494, + "learning_rate": 4.979063570612867e-06, + "loss": 1.543, + "step": 319 + }, + { + "epoch": 0.32871083718541344, + "grad_norm": 2.6551144123077393, + "learning_rate": 4.9771602588504e-06, + "loss": 1.6467, + "step": 320 + }, + { + "epoch": 0.32871083718541344, + "eval_loss": 1.89058518409729, + "eval_runtime": 37.8192, + "eval_samples_per_second": 26.442, + "eval_steps_per_second": 1.111, + "step": 320 + }, + { + "epoch": 0.32973805855161786, + "grad_norm": 2.3897273540496826, + "learning_rate": 4.975256947087933e-06, + "loss": 1.5607, + "step": 321 + }, + { + "epoch": 0.3307652799178223, + "grad_norm": 2.3869469165802, + "learning_rate": 4.973353635325467e-06, + "loss": 1.6079, + "step": 322 + }, + { + "epoch": 0.3317925012840267, + "grad_norm": 2.159193515777588, + "learning_rate": 4.971450323563e-06, + "loss": 1.5475, + "step": 323 + }, + { + "epoch": 0.33281972265023113, + "grad_norm": 2.3069653511047363, + "learning_rate": 4.969547011800533e-06, + "loss": 1.616, + "step": 324 + }, + { + "epoch": 0.33384694401643555, + "grad_norm": 2.3131422996520996, + "learning_rate": 4.967643700038067e-06, + "loss": 1.6106, + "step": 325 + }, + { + "epoch": 0.33487416538264, + "grad_norm": 2.4398200511932373, + "learning_rate": 4.9657403882756e-06, + "loss": 1.5779, + "step": 326 + }, + { + "epoch": 0.3359013867488444, + "grad_norm": 2.278024435043335, + "learning_rate": 4.9638370765131335e-06, + "loss": 1.5724, + "step": 327 + }, + { + "epoch": 0.33692860811504877, + "grad_norm": 2.4989173412323, + "learning_rate": 4.961933764750667e-06, + "loss": 1.6337, + "step": 328 + }, + { + "epoch": 0.3379558294812532, + "grad_norm": 2.46319580078125, + "learning_rate": 4.9600304529882e-06, + "loss": 1.6519, + "step": 329 + }, + { + "epoch": 0.3389830508474576, + "grad_norm": 2.2703049182891846, + "learning_rate": 4.9581271412257335e-06, + "loss": 1.6081, + "step": 330 + }, + { + "epoch": 0.3389830508474576, + "eval_loss": 1.8755545616149902, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 330 + }, + { + "epoch": 0.34001027221366203, + "grad_norm": 2.502196788787842, + "learning_rate": 4.956223829463266e-06, + "loss": 1.5221, + "step": 331 + }, + { + "epoch": 0.34103749357986646, + "grad_norm": 2.284280776977539, + "learning_rate": 4.954320517700799e-06, + "loss": 1.6268, + "step": 332 + }, + { + "epoch": 0.3420647149460709, + "grad_norm": 2.3551881313323975, + "learning_rate": 4.952417205938333e-06, + "loss": 1.6078, + "step": 333 + }, + { + "epoch": 0.3430919363122753, + "grad_norm": 2.343780755996704, + "learning_rate": 4.950513894175866e-06, + "loss": 1.591, + "step": 334 + }, + { + "epoch": 0.3441191576784797, + "grad_norm": 2.3524727821350098, + "learning_rate": 4.9486105824133995e-06, + "loss": 1.605, + "step": 335 + }, + { + "epoch": 0.34514637904468415, + "grad_norm": 2.349896192550659, + "learning_rate": 4.946707270650933e-06, + "loss": 1.6106, + "step": 336 + }, + { + "epoch": 0.34617360041088857, + "grad_norm": 2.5154407024383545, + "learning_rate": 4.944803958888466e-06, + "loss": 1.6429, + "step": 337 + }, + { + "epoch": 0.34720082177709294, + "grad_norm": 2.2692947387695312, + "learning_rate": 4.942900647126e-06, + "loss": 1.5865, + "step": 338 + }, + { + "epoch": 0.34822804314329736, + "grad_norm": 2.2672805786132812, + "learning_rate": 4.940997335363533e-06, + "loss": 1.5464, + "step": 339 + }, + { + "epoch": 0.3492552645095018, + "grad_norm": 2.5147039890289307, + "learning_rate": 4.939094023601066e-06, + "loss": 1.5837, + "step": 340 + }, + { + "epoch": 0.3492552645095018, + "eval_loss": 1.8672455549240112, + "eval_runtime": 37.9644, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 340 + }, + { + "epoch": 0.3502824858757062, + "grad_norm": 2.3416433334350586, + "learning_rate": 4.9371907118386e-06, + "loss": 1.5937, + "step": 341 + }, + { + "epoch": 0.35130970724191063, + "grad_norm": 2.3044960498809814, + "learning_rate": 4.935287400076133e-06, + "loss": 1.6132, + "step": 342 + }, + { + "epoch": 0.35233692860811505, + "grad_norm": 2.4360430240631104, + "learning_rate": 4.933384088313666e-06, + "loss": 1.5731, + "step": 343 + }, + { + "epoch": 0.3533641499743195, + "grad_norm": 2.392050266265869, + "learning_rate": 4.9314807765512e-06, + "loss": 1.5098, + "step": 344 + }, + { + "epoch": 0.3543913713405239, + "grad_norm": 2.3815808296203613, + "learning_rate": 4.929577464788733e-06, + "loss": 1.5827, + "step": 345 + }, + { + "epoch": 0.3554185927067283, + "grad_norm": 2.423077344894409, + "learning_rate": 4.9276741530262665e-06, + "loss": 1.5857, + "step": 346 + }, + { + "epoch": 0.35644581407293274, + "grad_norm": 2.5535898208618164, + "learning_rate": 4.9257708412638e-06, + "loss": 1.5884, + "step": 347 + }, + { + "epoch": 0.3574730354391371, + "grad_norm": 2.2665445804595947, + "learning_rate": 4.923867529501332e-06, + "loss": 1.5561, + "step": 348 + }, + { + "epoch": 0.35850025680534153, + "grad_norm": 2.3357481956481934, + "learning_rate": 4.921964217738866e-06, + "loss": 1.6444, + "step": 349 + }, + { + "epoch": 0.35952747817154596, + "grad_norm": 2.4537885189056396, + "learning_rate": 4.920060905976399e-06, + "loss": 1.6262, + "step": 350 + }, + { + "epoch": 0.35952747817154596, + "eval_loss": 1.8621584177017212, + "eval_runtime": 38.0027, + "eval_samples_per_second": 26.314, + "eval_steps_per_second": 1.105, + "step": 350 + }, + { + "epoch": 0.3605546995377504, + "grad_norm": 2.379754066467285, + "learning_rate": 4.9181575942139325e-06, + "loss": 1.6227, + "step": 351 + }, + { + "epoch": 0.3615819209039548, + "grad_norm": 2.432905435562134, + "learning_rate": 4.916254282451466e-06, + "loss": 1.6554, + "step": 352 + }, + { + "epoch": 0.3626091422701592, + "grad_norm": 2.34236216545105, + "learning_rate": 4.914350970688999e-06, + "loss": 1.5613, + "step": 353 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.3019444942474365, + "learning_rate": 4.9124476589265325e-06, + "loss": 1.5535, + "step": 354 + }, + { + "epoch": 0.36466358500256807, + "grad_norm": 2.3354644775390625, + "learning_rate": 4.910544347164066e-06, + "loss": 1.6384, + "step": 355 + }, + { + "epoch": 0.3656908063687725, + "grad_norm": 2.5154788494110107, + "learning_rate": 4.908641035401599e-06, + "loss": 1.6986, + "step": 356 + }, + { + "epoch": 0.3667180277349769, + "grad_norm": 2.2707178592681885, + "learning_rate": 4.906737723639133e-06, + "loss": 1.5705, + "step": 357 + }, + { + "epoch": 0.3677452491011813, + "grad_norm": 2.3058297634124756, + "learning_rate": 4.904834411876665e-06, + "loss": 1.6033, + "step": 358 + }, + { + "epoch": 0.3687724704673857, + "grad_norm": 2.3923370838165283, + "learning_rate": 4.902931100114199e-06, + "loss": 1.5333, + "step": 359 + }, + { + "epoch": 0.3697996918335901, + "grad_norm": 2.4023196697235107, + "learning_rate": 4.901027788351733e-06, + "loss": 1.5775, + "step": 360 + }, + { + "epoch": 0.3697996918335901, + "eval_loss": 1.867856740951538, + "eval_runtime": 38.0125, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 360 + }, + { + "epoch": 0.37082691319979455, + "grad_norm": 2.4903581142425537, + "learning_rate": 4.899124476589266e-06, + "loss": 1.5491, + "step": 361 + }, + { + "epoch": 0.371854134565999, + "grad_norm": 2.556683301925659, + "learning_rate": 4.8972211648267994e-06, + "loss": 1.5967, + "step": 362 + }, + { + "epoch": 0.3728813559322034, + "grad_norm": 2.3017377853393555, + "learning_rate": 4.895317853064333e-06, + "loss": 1.5661, + "step": 363 + }, + { + "epoch": 0.3739085772984078, + "grad_norm": 2.286191940307617, + "learning_rate": 4.893414541301866e-06, + "loss": 1.5679, + "step": 364 + }, + { + "epoch": 0.37493579866461224, + "grad_norm": 2.260812520980835, + "learning_rate": 4.891511229539399e-06, + "loss": 1.5965, + "step": 365 + }, + { + "epoch": 0.37596302003081666, + "grad_norm": 2.4569408893585205, + "learning_rate": 4.889607917776932e-06, + "loss": 1.5923, + "step": 366 + }, + { + "epoch": 0.37699024139702103, + "grad_norm": 2.571552038192749, + "learning_rate": 4.887704606014465e-06, + "loss": 1.6021, + "step": 367 + }, + { + "epoch": 0.37801746276322545, + "grad_norm": 2.287754774093628, + "learning_rate": 4.885801294251999e-06, + "loss": 1.5304, + "step": 368 + }, + { + "epoch": 0.3790446841294299, + "grad_norm": 2.317784070968628, + "learning_rate": 4.883897982489532e-06, + "loss": 1.6233, + "step": 369 + }, + { + "epoch": 0.3800719054956343, + "grad_norm": 2.290601968765259, + "learning_rate": 4.8819946707270655e-06, + "loss": 1.624, + "step": 370 + }, + { + "epoch": 0.3800719054956343, + "eval_loss": 1.8604528903961182, + "eval_runtime": 38.0176, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 370 + }, + { + "epoch": 0.3810991268618387, + "grad_norm": 2.28519868850708, + "learning_rate": 4.880091358964599e-06, + "loss": 1.5286, + "step": 371 + }, + { + "epoch": 0.38212634822804314, + "grad_norm": 2.271810293197632, + "learning_rate": 4.878188047202132e-06, + "loss": 1.6879, + "step": 372 + }, + { + "epoch": 0.38315356959424757, + "grad_norm": 2.391848564147949, + "learning_rate": 4.876284735439666e-06, + "loss": 1.6458, + "step": 373 + }, + { + "epoch": 0.384180790960452, + "grad_norm": 2.3787031173706055, + "learning_rate": 4.874381423677199e-06, + "loss": 1.5738, + "step": 374 + }, + { + "epoch": 0.3852080123266564, + "grad_norm": 2.404219150543213, + "learning_rate": 4.8724781119147315e-06, + "loss": 1.6184, + "step": 375 + }, + { + "epoch": 0.38623523369286084, + "grad_norm": 2.372661590576172, + "learning_rate": 4.870574800152265e-06, + "loss": 1.5316, + "step": 376 + }, + { + "epoch": 0.3872624550590652, + "grad_norm": 2.4397103786468506, + "learning_rate": 4.868671488389798e-06, + "loss": 1.5605, + "step": 377 + }, + { + "epoch": 0.3882896764252696, + "grad_norm": 2.371006727218628, + "learning_rate": 4.8667681766273316e-06, + "loss": 1.5184, + "step": 378 + }, + { + "epoch": 0.38931689779147405, + "grad_norm": 2.2621910572052, + "learning_rate": 4.864864864864866e-06, + "loss": 1.5788, + "step": 379 + }, + { + "epoch": 0.39034411915767847, + "grad_norm": 2.2888615131378174, + "learning_rate": 4.862961553102399e-06, + "loss": 1.6244, + "step": 380 + }, + { + "epoch": 0.39034411915767847, + "eval_loss": 1.8586912155151367, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 380 + }, + { + "epoch": 0.3913713405238829, + "grad_norm": 2.4439220428466797, + "learning_rate": 4.8610582413399325e-06, + "loss": 1.5749, + "step": 381 + }, + { + "epoch": 0.3923985618900873, + "grad_norm": 2.2423763275146484, + "learning_rate": 4.859154929577465e-06, + "loss": 1.5738, + "step": 382 + }, + { + "epoch": 0.39342578325629174, + "grad_norm": 2.267467737197876, + "learning_rate": 4.857251617814998e-06, + "loss": 1.6212, + "step": 383 + }, + { + "epoch": 0.39445300462249616, + "grad_norm": 2.3680620193481445, + "learning_rate": 4.855348306052532e-06, + "loss": 1.6129, + "step": 384 + }, + { + "epoch": 0.3954802259887006, + "grad_norm": 2.3782172203063965, + "learning_rate": 4.853444994290065e-06, + "loss": 1.5605, + "step": 385 + }, + { + "epoch": 0.396507447354905, + "grad_norm": 2.5745720863342285, + "learning_rate": 4.8515416825275985e-06, + "loss": 1.634, + "step": 386 + }, + { + "epoch": 0.3975346687211094, + "grad_norm": 2.38212513923645, + "learning_rate": 4.849638370765132e-06, + "loss": 1.5671, + "step": 387 + }, + { + "epoch": 0.3985618900873138, + "grad_norm": 2.3395450115203857, + "learning_rate": 4.847735059002665e-06, + "loss": 1.5088, + "step": 388 + }, + { + "epoch": 0.3995891114535182, + "grad_norm": 2.474440574645996, + "learning_rate": 4.8458317472401985e-06, + "loss": 1.5997, + "step": 389 + }, + { + "epoch": 0.40061633281972264, + "grad_norm": 2.429133176803589, + "learning_rate": 4.843928435477732e-06, + "loss": 1.5611, + "step": 390 + }, + { + "epoch": 0.40061633281972264, + "eval_loss": 1.8613603115081787, + "eval_runtime": 37.9862, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 390 + }, + { + "epoch": 0.40164355418592707, + "grad_norm": 2.403207778930664, + "learning_rate": 4.842025123715265e-06, + "loss": 1.5976, + "step": 391 + }, + { + "epoch": 0.4026707755521315, + "grad_norm": 2.383018732070923, + "learning_rate": 4.840121811952798e-06, + "loss": 1.5791, + "step": 392 + }, + { + "epoch": 0.4036979969183359, + "grad_norm": 2.5303306579589844, + "learning_rate": 4.838218500190331e-06, + "loss": 1.6035, + "step": 393 + }, + { + "epoch": 0.40472521828454033, + "grad_norm": 2.391075372695923, + "learning_rate": 4.8363151884278645e-06, + "loss": 1.5795, + "step": 394 + }, + { + "epoch": 0.40575243965074476, + "grad_norm": 2.396934747695923, + "learning_rate": 4.834411876665398e-06, + "loss": 1.5965, + "step": 395 + }, + { + "epoch": 0.4067796610169492, + "grad_norm": 2.4631423950195312, + "learning_rate": 4.832508564902931e-06, + "loss": 1.5372, + "step": 396 + }, + { + "epoch": 0.40780688238315355, + "grad_norm": 2.3457326889038086, + "learning_rate": 4.830605253140465e-06, + "loss": 1.5729, + "step": 397 + }, + { + "epoch": 0.40883410374935797, + "grad_norm": 2.3763277530670166, + "learning_rate": 4.828701941377998e-06, + "loss": 1.5309, + "step": 398 + }, + { + "epoch": 0.4098613251155624, + "grad_norm": 2.442439556121826, + "learning_rate": 4.826798629615531e-06, + "loss": 1.5465, + "step": 399 + }, + { + "epoch": 0.4108885464817668, + "grad_norm": 2.3464581966400146, + "learning_rate": 4.824895317853065e-06, + "loss": 1.5665, + "step": 400 + }, + { + "epoch": 0.4108885464817668, + "eval_loss": 1.8671027421951294, + "eval_runtime": 38.0166, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 400 + }, + { + "epoch": 0.41191576784797124, + "grad_norm": 2.5441887378692627, + "learning_rate": 4.822992006090598e-06, + "loss": 1.5232, + "step": 401 + }, + { + "epoch": 0.41294298921417566, + "grad_norm": 2.4234976768493652, + "learning_rate": 4.821088694328131e-06, + "loss": 1.5569, + "step": 402 + }, + { + "epoch": 0.4139702105803801, + "grad_norm": 2.3603928089141846, + "learning_rate": 4.819185382565665e-06, + "loss": 1.612, + "step": 403 + }, + { + "epoch": 0.4149974319465845, + "grad_norm": 2.337862491607666, + "learning_rate": 4.817282070803198e-06, + "loss": 1.5882, + "step": 404 + }, + { + "epoch": 0.41602465331278893, + "grad_norm": 2.3805742263793945, + "learning_rate": 4.8153787590407315e-06, + "loss": 1.5204, + "step": 405 + }, + { + "epoch": 0.41705187467899335, + "grad_norm": 2.3335204124450684, + "learning_rate": 4.813475447278265e-06, + "loss": 1.6041, + "step": 406 + }, + { + "epoch": 0.4180790960451977, + "grad_norm": 2.406137466430664, + "learning_rate": 4.811572135515798e-06, + "loss": 1.5348, + "step": 407 + }, + { + "epoch": 0.41910631741140214, + "grad_norm": 2.2823917865753174, + "learning_rate": 4.809668823753332e-06, + "loss": 1.5351, + "step": 408 + }, + { + "epoch": 0.42013353877760656, + "grad_norm": 2.3577091693878174, + "learning_rate": 4.807765511990864e-06, + "loss": 1.5733, + "step": 409 + }, + { + "epoch": 0.421160760143811, + "grad_norm": 2.536691665649414, + "learning_rate": 4.8058622002283975e-06, + "loss": 1.6153, + "step": 410 + }, + { + "epoch": 0.421160760143811, + "eval_loss": 1.8613901138305664, + "eval_runtime": 37.6689, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 410 + }, + { + "epoch": 0.4221879815100154, + "grad_norm": 2.4020118713378906, + "learning_rate": 4.803958888465931e-06, + "loss": 1.5722, + "step": 411 + }, + { + "epoch": 0.42321520287621983, + "grad_norm": 2.5284810066223145, + "learning_rate": 4.802055576703464e-06, + "loss": 1.6546, + "step": 412 + }, + { + "epoch": 0.42424242424242425, + "grad_norm": 2.270094156265259, + "learning_rate": 4.8001522649409976e-06, + "loss": 1.5345, + "step": 413 + }, + { + "epoch": 0.4252696456086287, + "grad_norm": 2.241417407989502, + "learning_rate": 4.798248953178531e-06, + "loss": 1.5167, + "step": 414 + }, + { + "epoch": 0.4262968669748331, + "grad_norm": 2.3733017444610596, + "learning_rate": 4.796345641416064e-06, + "loss": 1.5532, + "step": 415 + }, + { + "epoch": 0.42732408834103747, + "grad_norm": 2.452587842941284, + "learning_rate": 4.794442329653598e-06, + "loss": 1.6136, + "step": 416 + }, + { + "epoch": 0.4283513097072419, + "grad_norm": 2.417217969894409, + "learning_rate": 4.792539017891131e-06, + "loss": 1.5398, + "step": 417 + }, + { + "epoch": 0.4293785310734463, + "grad_norm": 2.4502453804016113, + "learning_rate": 4.790635706128664e-06, + "loss": 1.5741, + "step": 418 + }, + { + "epoch": 0.43040575243965074, + "grad_norm": 2.2996811866760254, + "learning_rate": 4.788732394366197e-06, + "loss": 1.511, + "step": 419 + }, + { + "epoch": 0.43143297380585516, + "grad_norm": 2.3573555946350098, + "learning_rate": 4.786829082603731e-06, + "loss": 1.5117, + "step": 420 + }, + { + "epoch": 0.43143297380585516, + "eval_loss": 1.8690649271011353, + "eval_runtime": 37.7813, + "eval_samples_per_second": 26.468, + "eval_steps_per_second": 1.112, + "step": 420 + }, + { + "epoch": 0.4324601951720596, + "grad_norm": 2.37233567237854, + "learning_rate": 4.7849257708412645e-06, + "loss": 1.5873, + "step": 421 + }, + { + "epoch": 0.433487416538264, + "grad_norm": 2.146909713745117, + "learning_rate": 4.783022459078798e-06, + "loss": 1.5291, + "step": 422 + }, + { + "epoch": 0.4345146379044684, + "grad_norm": 2.5237205028533936, + "learning_rate": 4.781119147316331e-06, + "loss": 1.5545, + "step": 423 + }, + { + "epoch": 0.43554185927067285, + "grad_norm": 2.356842279434204, + "learning_rate": 4.7792158355538645e-06, + "loss": 1.5635, + "step": 424 + }, + { + "epoch": 0.43656908063687727, + "grad_norm": 2.1963610649108887, + "learning_rate": 4.777312523791398e-06, + "loss": 1.5355, + "step": 425 + }, + { + "epoch": 0.43759630200308164, + "grad_norm": 2.6425886154174805, + "learning_rate": 4.77540921202893e-06, + "loss": 1.5803, + "step": 426 + }, + { + "epoch": 0.43862352336928606, + "grad_norm": 2.4072272777557373, + "learning_rate": 4.773505900266464e-06, + "loss": 1.5649, + "step": 427 + }, + { + "epoch": 0.4396507447354905, + "grad_norm": 2.592374801635742, + "learning_rate": 4.771602588503997e-06, + "loss": 1.5693, + "step": 428 + }, + { + "epoch": 0.4406779661016949, + "grad_norm": 2.4534225463867188, + "learning_rate": 4.7696992767415305e-06, + "loss": 1.591, + "step": 429 + }, + { + "epoch": 0.44170518746789933, + "grad_norm": 2.434457778930664, + "learning_rate": 4.767795964979064e-06, + "loss": 1.5991, + "step": 430 + }, + { + "epoch": 0.44170518746789933, + "eval_loss": 1.8584965467453003, + "eval_runtime": 37.8854, + "eval_samples_per_second": 26.395, + "eval_steps_per_second": 1.109, + "step": 430 + }, + { + "epoch": 0.44273240883410375, + "grad_norm": 2.3150746822357178, + "learning_rate": 4.765892653216597e-06, + "loss": 1.6032, + "step": 431 + }, + { + "epoch": 0.4437596302003082, + "grad_norm": 2.3040223121643066, + "learning_rate": 4.763989341454131e-06, + "loss": 1.5115, + "step": 432 + }, + { + "epoch": 0.4447868515665126, + "grad_norm": 2.413007974624634, + "learning_rate": 4.762086029691664e-06, + "loss": 1.5305, + "step": 433 + }, + { + "epoch": 0.445814072932717, + "grad_norm": 2.526289463043213, + "learning_rate": 4.760182717929197e-06, + "loss": 1.564, + "step": 434 + }, + { + "epoch": 0.44684129429892144, + "grad_norm": 2.4122776985168457, + "learning_rate": 4.758279406166731e-06, + "loss": 1.4918, + "step": 435 + }, + { + "epoch": 0.4478685156651258, + "grad_norm": 2.474274158477783, + "learning_rate": 4.756376094404263e-06, + "loss": 1.5583, + "step": 436 + }, + { + "epoch": 0.44889573703133023, + "grad_norm": 2.5322892665863037, + "learning_rate": 4.7544727826417966e-06, + "loss": 1.6081, + "step": 437 + }, + { + "epoch": 0.44992295839753466, + "grad_norm": 2.391660690307617, + "learning_rate": 4.75256947087933e-06, + "loss": 1.5628, + "step": 438 + }, + { + "epoch": 0.4509501797637391, + "grad_norm": 2.426671028137207, + "learning_rate": 4.750666159116863e-06, + "loss": 1.531, + "step": 439 + }, + { + "epoch": 0.4519774011299435, + "grad_norm": 2.7095694541931152, + "learning_rate": 4.7487628473543975e-06, + "loss": 1.4811, + "step": 440 + }, + { + "epoch": 0.4519774011299435, + "eval_loss": 1.8348888158798218, + "eval_runtime": 37.9114, + "eval_samples_per_second": 26.377, + "eval_steps_per_second": 1.108, + "step": 440 + }, + { + "epoch": 0.4530046224961479, + "grad_norm": 2.5425240993499756, + "learning_rate": 4.746859535591931e-06, + "loss": 1.5453, + "step": 441 + }, + { + "epoch": 0.45403184386235235, + "grad_norm": 2.6848669052124023, + "learning_rate": 4.744956223829464e-06, + "loss": 1.526, + "step": 442 + }, + { + "epoch": 0.45505906522855677, + "grad_norm": 2.634068489074707, + "learning_rate": 4.743052912066997e-06, + "loss": 1.5748, + "step": 443 + }, + { + "epoch": 0.4560862865947612, + "grad_norm": 2.492030382156372, + "learning_rate": 4.74114960030453e-06, + "loss": 1.5858, + "step": 444 + }, + { + "epoch": 0.4571135079609656, + "grad_norm": 2.787740707397461, + "learning_rate": 4.7392462885420635e-06, + "loss": 1.4774, + "step": 445 + }, + { + "epoch": 0.45814072932717, + "grad_norm": 2.395914077758789, + "learning_rate": 4.737342976779597e-06, + "loss": 1.558, + "step": 446 + }, + { + "epoch": 0.4591679506933744, + "grad_norm": 2.6642563343048096, + "learning_rate": 4.73543966501713e-06, + "loss": 1.5558, + "step": 447 + }, + { + "epoch": 0.4601951720595788, + "grad_norm": 2.619943857192993, + "learning_rate": 4.7335363532546636e-06, + "loss": 1.5442, + "step": 448 + }, + { + "epoch": 0.46122239342578325, + "grad_norm": 2.355133056640625, + "learning_rate": 4.731633041492197e-06, + "loss": 1.5613, + "step": 449 + }, + { + "epoch": 0.4622496147919877, + "grad_norm": 2.6572225093841553, + "learning_rate": 4.72972972972973e-06, + "loss": 1.5951, + "step": 450 + }, + { + "epoch": 0.4622496147919877, + "eval_loss": 1.83199143409729, + "eval_runtime": 37.8996, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 450 + }, + { + "epoch": 0.4632768361581921, + "grad_norm": 2.692873001098633, + "learning_rate": 4.727826417967264e-06, + "loss": 1.5531, + "step": 451 + }, + { + "epoch": 0.4643040575243965, + "grad_norm": 2.3903489112854004, + "learning_rate": 4.725923106204797e-06, + "loss": 1.5411, + "step": 452 + }, + { + "epoch": 0.46533127889060094, + "grad_norm": 2.453263521194458, + "learning_rate": 4.72401979444233e-06, + "loss": 1.5632, + "step": 453 + }, + { + "epoch": 0.46635850025680536, + "grad_norm": 2.5637760162353516, + "learning_rate": 4.722116482679863e-06, + "loss": 1.5629, + "step": 454 + }, + { + "epoch": 0.46738572162300973, + "grad_norm": 2.387033700942993, + "learning_rate": 4.720213170917396e-06, + "loss": 1.5786, + "step": 455 + }, + { + "epoch": 0.46841294298921415, + "grad_norm": 2.4563071727752686, + "learning_rate": 4.71830985915493e-06, + "loss": 1.5321, + "step": 456 + }, + { + "epoch": 0.4694401643554186, + "grad_norm": 2.607451915740967, + "learning_rate": 4.716406547392463e-06, + "loss": 1.5958, + "step": 457 + }, + { + "epoch": 0.470467385721623, + "grad_norm": 2.4846854209899902, + "learning_rate": 4.714503235629996e-06, + "loss": 1.5205, + "step": 458 + }, + { + "epoch": 0.4714946070878274, + "grad_norm": 2.377629041671753, + "learning_rate": 4.71259992386753e-06, + "loss": 1.5403, + "step": 459 + }, + { + "epoch": 0.47252182845403184, + "grad_norm": 2.4470160007476807, + "learning_rate": 4.710696612105063e-06, + "loss": 1.5456, + "step": 460 + }, + { + "epoch": 0.47252182845403184, + "eval_loss": 1.83836829662323, + "eval_runtime": 37.88, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 460 + }, + { + "epoch": 0.47354904982023627, + "grad_norm": 2.383418560028076, + "learning_rate": 4.708793300342596e-06, + "loss": 1.5468, + "step": 461 + }, + { + "epoch": 0.4745762711864407, + "grad_norm": 2.3672895431518555, + "learning_rate": 4.70688998858013e-06, + "loss": 1.5313, + "step": 462 + }, + { + "epoch": 0.4756034925526451, + "grad_norm": 2.415717124938965, + "learning_rate": 4.704986676817663e-06, + "loss": 1.5645, + "step": 463 + }, + { + "epoch": 0.47663071391884954, + "grad_norm": 2.4690067768096924, + "learning_rate": 4.7030833650551965e-06, + "loss": 1.5285, + "step": 464 + }, + { + "epoch": 0.4776579352850539, + "grad_norm": 2.3845672607421875, + "learning_rate": 4.70118005329273e-06, + "loss": 1.596, + "step": 465 + }, + { + "epoch": 0.4786851566512583, + "grad_norm": 2.2801802158355713, + "learning_rate": 4.699276741530263e-06, + "loss": 1.5777, + "step": 466 + }, + { + "epoch": 0.47971237801746275, + "grad_norm": 2.4668991565704346, + "learning_rate": 4.697373429767797e-06, + "loss": 1.5566, + "step": 467 + }, + { + "epoch": 0.48073959938366717, + "grad_norm": 2.446078062057495, + "learning_rate": 4.69547011800533e-06, + "loss": 1.5664, + "step": 468 + }, + { + "epoch": 0.4817668207498716, + "grad_norm": 2.2310681343078613, + "learning_rate": 4.693566806242863e-06, + "loss": 1.5425, + "step": 469 + }, + { + "epoch": 0.482794042116076, + "grad_norm": 2.3938229084014893, + "learning_rate": 4.691663494480397e-06, + "loss": 1.576, + "step": 470 + }, + { + "epoch": 0.482794042116076, + "eval_loss": 1.83778977394104, + "eval_runtime": 37.9276, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 470 + }, + { + "epoch": 0.48382126348228044, + "grad_norm": 2.23827862739563, + "learning_rate": 4.689760182717929e-06, + "loss": 1.4763, + "step": 471 + }, + { + "epoch": 0.48484848484848486, + "grad_norm": 2.3148930072784424, + "learning_rate": 4.6878568709554626e-06, + "loss": 1.5386, + "step": 472 + }, + { + "epoch": 0.4858757062146893, + "grad_norm": 2.407036781311035, + "learning_rate": 4.685953559192996e-06, + "loss": 1.5576, + "step": 473 + }, + { + "epoch": 0.4869029275808937, + "grad_norm": 2.4675374031066895, + "learning_rate": 4.684050247430529e-06, + "loss": 1.4931, + "step": 474 + }, + { + "epoch": 0.4879301489470981, + "grad_norm": 2.4593775272369385, + "learning_rate": 4.682146935668063e-06, + "loss": 1.5348, + "step": 475 + }, + { + "epoch": 0.4889573703133025, + "grad_norm": 2.4169375896453857, + "learning_rate": 4.680243623905596e-06, + "loss": 1.5614, + "step": 476 + }, + { + "epoch": 0.4899845916795069, + "grad_norm": 2.6177303791046143, + "learning_rate": 4.678340312143129e-06, + "loss": 1.5138, + "step": 477 + }, + { + "epoch": 0.49101181304571134, + "grad_norm": 2.6418113708496094, + "learning_rate": 4.676437000380663e-06, + "loss": 1.6377, + "step": 478 + }, + { + "epoch": 0.49203903441191577, + "grad_norm": 2.338491439819336, + "learning_rate": 4.674533688618196e-06, + "loss": 1.5682, + "step": 479 + }, + { + "epoch": 0.4930662557781202, + "grad_norm": 2.5901622772216797, + "learning_rate": 4.6726303768557295e-06, + "loss": 1.4899, + "step": 480 + }, + { + "epoch": 0.4930662557781202, + "eval_loss": 1.8238886594772339, + "eval_runtime": 37.9605, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 480 + }, + { + "epoch": 0.4940934771443246, + "grad_norm": 2.470827102661133, + "learning_rate": 4.670727065093263e-06, + "loss": 1.5016, + "step": 481 + }, + { + "epoch": 0.49512069851052903, + "grad_norm": 2.371201992034912, + "learning_rate": 4.668823753330796e-06, + "loss": 1.5399, + "step": 482 + }, + { + "epoch": 0.49614791987673346, + "grad_norm": 2.532802104949951, + "learning_rate": 4.6669204415683295e-06, + "loss": 1.5985, + "step": 483 + }, + { + "epoch": 0.4971751412429379, + "grad_norm": 2.6294727325439453, + "learning_rate": 4.665017129805863e-06, + "loss": 1.5802, + "step": 484 + }, + { + "epoch": 0.49820236260914225, + "grad_norm": 2.2164878845214844, + "learning_rate": 4.663113818043396e-06, + "loss": 1.5384, + "step": 485 + }, + { + "epoch": 0.49922958397534667, + "grad_norm": 2.311232805252075, + "learning_rate": 4.66121050628093e-06, + "loss": 1.5785, + "step": 486 + }, + { + "epoch": 0.5002568053415511, + "grad_norm": 2.2916736602783203, + "learning_rate": 4.659307194518463e-06, + "loss": 1.5401, + "step": 487 + }, + { + "epoch": 0.5012840267077555, + "grad_norm": 2.3978800773620605, + "learning_rate": 4.6574038827559955e-06, + "loss": 1.5619, + "step": 488 + }, + { + "epoch": 0.50231124807396, + "grad_norm": 2.438917398452759, + "learning_rate": 4.655500570993529e-06, + "loss": 1.5824, + "step": 489 + }, + { + "epoch": 0.5033384694401644, + "grad_norm": 2.528862237930298, + "learning_rate": 4.653597259231062e-06, + "loss": 1.4802, + "step": 490 + }, + { + "epoch": 0.5033384694401644, + "eval_loss": 1.83504056930542, + "eval_runtime": 37.9357, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 490 + }, + { + "epoch": 0.5043656908063687, + "grad_norm": 2.54522442817688, + "learning_rate": 4.651693947468596e-06, + "loss": 1.6186, + "step": 491 + }, + { + "epoch": 0.5053929121725732, + "grad_norm": 2.5037031173706055, + "learning_rate": 4.649790635706129e-06, + "loss": 1.5159, + "step": 492 + }, + { + "epoch": 0.5064201335387776, + "grad_norm": 2.4600465297698975, + "learning_rate": 4.647887323943662e-06, + "loss": 1.5378, + "step": 493 + }, + { + "epoch": 0.507447354904982, + "grad_norm": 2.393695831298828, + "learning_rate": 4.645984012181196e-06, + "loss": 1.4485, + "step": 494 + }, + { + "epoch": 0.5084745762711864, + "grad_norm": 2.488626718521118, + "learning_rate": 4.644080700418729e-06, + "loss": 1.57, + "step": 495 + }, + { + "epoch": 0.5095017976373909, + "grad_norm": 2.409858465194702, + "learning_rate": 4.642177388656262e-06, + "loss": 1.5343, + "step": 496 + }, + { + "epoch": 0.5105290190035953, + "grad_norm": 2.5607573986053467, + "learning_rate": 4.640274076893796e-06, + "loss": 1.5825, + "step": 497 + }, + { + "epoch": 0.5115562403697997, + "grad_norm": 2.5953354835510254, + "learning_rate": 4.638370765131328e-06, + "loss": 1.5129, + "step": 498 + }, + { + "epoch": 0.5125834617360041, + "grad_norm": 2.3321897983551025, + "learning_rate": 4.636467453368862e-06, + "loss": 1.4972, + "step": 499 + }, + { + "epoch": 0.5136106831022085, + "grad_norm": 2.332052707672119, + "learning_rate": 4.634564141606396e-06, + "loss": 1.567, + "step": 500 + }, + { + "epoch": 0.5136106831022085, + "eval_loss": 1.824558138847351, + "eval_runtime": 37.9157, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 500 + }, + { + "epoch": 0.514637904468413, + "grad_norm": 2.431809902191162, + "learning_rate": 4.632660829843929e-06, + "loss": 1.494, + "step": 501 + }, + { + "epoch": 0.5156651258346173, + "grad_norm": 2.7118983268737793, + "learning_rate": 4.630757518081463e-06, + "loss": 1.6289, + "step": 502 + }, + { + "epoch": 0.5166923472008218, + "grad_norm": 2.3549370765686035, + "learning_rate": 4.628854206318996e-06, + "loss": 1.5363, + "step": 503 + }, + { + "epoch": 0.5177195685670262, + "grad_norm": 2.3343000411987305, + "learning_rate": 4.626950894556529e-06, + "loss": 1.5749, + "step": 504 + }, + { + "epoch": 0.5187467899332306, + "grad_norm": 2.4281466007232666, + "learning_rate": 4.625047582794062e-06, + "loss": 1.4958, + "step": 505 + }, + { + "epoch": 0.519774011299435, + "grad_norm": 2.3858845233917236, + "learning_rate": 4.623144271031595e-06, + "loss": 1.5031, + "step": 506 + }, + { + "epoch": 0.5208012326656395, + "grad_norm": 2.3714635372161865, + "learning_rate": 4.6212409592691286e-06, + "loss": 1.5808, + "step": 507 + }, + { + "epoch": 0.5218284540318439, + "grad_norm": 2.5290520191192627, + "learning_rate": 4.619337647506662e-06, + "loss": 1.541, + "step": 508 + }, + { + "epoch": 0.5228556753980482, + "grad_norm": 2.5521368980407715, + "learning_rate": 4.617434335744195e-06, + "loss": 1.4992, + "step": 509 + }, + { + "epoch": 0.5238828967642527, + "grad_norm": 2.436588764190674, + "learning_rate": 4.615531023981729e-06, + "loss": 1.5283, + "step": 510 + }, + { + "epoch": 0.5238828967642527, + "eval_loss": 1.8220165967941284, + "eval_runtime": 37.6123, + "eval_samples_per_second": 26.587, + "eval_steps_per_second": 1.117, + "step": 510 + }, + { + "epoch": 0.5249101181304571, + "grad_norm": 2.5240023136138916, + "learning_rate": 4.613627712219262e-06, + "loss": 1.5426, + "step": 511 + }, + { + "epoch": 0.5259373394966615, + "grad_norm": 2.5292775630950928, + "learning_rate": 4.611724400456795e-06, + "loss": 1.5411, + "step": 512 + }, + { + "epoch": 0.5269645608628659, + "grad_norm": 2.3622210025787354, + "learning_rate": 4.609821088694329e-06, + "loss": 1.5083, + "step": 513 + }, + { + "epoch": 0.5279917822290704, + "grad_norm": 2.545267105102539, + "learning_rate": 4.607917776931862e-06, + "loss": 1.5541, + "step": 514 + }, + { + "epoch": 0.5290190035952748, + "grad_norm": 2.3417961597442627, + "learning_rate": 4.606014465169395e-06, + "loss": 1.5686, + "step": 515 + }, + { + "epoch": 0.5300462249614792, + "grad_norm": 2.2782580852508545, + "learning_rate": 4.604111153406928e-06, + "loss": 1.4761, + "step": 516 + }, + { + "epoch": 0.5310734463276836, + "grad_norm": 2.4437527656555176, + "learning_rate": 4.602207841644461e-06, + "loss": 1.5818, + "step": 517 + }, + { + "epoch": 0.5321006676938881, + "grad_norm": 2.3594319820404053, + "learning_rate": 4.600304529881995e-06, + "loss": 1.5556, + "step": 518 + }, + { + "epoch": 0.5331278890600925, + "grad_norm": 2.325979471206665, + "learning_rate": 4.598401218119528e-06, + "loss": 1.5726, + "step": 519 + }, + { + "epoch": 0.5341551104262968, + "grad_norm": 2.2765676975250244, + "learning_rate": 4.5964979063570614e-06, + "loss": 1.5151, + "step": 520 + }, + { + "epoch": 0.5341551104262968, + "eval_loss": 1.825669288635254, + "eval_runtime": 37.7491, + "eval_samples_per_second": 26.491, + "eval_steps_per_second": 1.113, + "step": 520 + }, + { + "epoch": 0.5351823317925013, + "grad_norm": 2.357647657394409, + "learning_rate": 4.594594594594596e-06, + "loss": 1.5505, + "step": 521 + }, + { + "epoch": 0.5362095531587057, + "grad_norm": 2.3652896881103516, + "learning_rate": 4.592691282832128e-06, + "loss": 1.5621, + "step": 522 + }, + { + "epoch": 0.5372367745249101, + "grad_norm": 2.234096050262451, + "learning_rate": 4.5907879710696615e-06, + "loss": 1.4851, + "step": 523 + }, + { + "epoch": 0.5382639958911145, + "grad_norm": 2.3414828777313232, + "learning_rate": 4.588884659307195e-06, + "loss": 1.5594, + "step": 524 + }, + { + "epoch": 0.539291217257319, + "grad_norm": 2.3934929370880127, + "learning_rate": 4.586981347544728e-06, + "loss": 1.4994, + "step": 525 + }, + { + "epoch": 0.5403184386235234, + "grad_norm": 2.400479555130005, + "learning_rate": 4.585078035782262e-06, + "loss": 1.5346, + "step": 526 + }, + { + "epoch": 0.5413456599897278, + "grad_norm": 2.3817873001098633, + "learning_rate": 4.583174724019795e-06, + "loss": 1.4803, + "step": 527 + }, + { + "epoch": 0.5423728813559322, + "grad_norm": 2.4503185749053955, + "learning_rate": 4.581271412257328e-06, + "loss": 1.4711, + "step": 528 + }, + { + "epoch": 0.5434001027221366, + "grad_norm": 2.452578067779541, + "learning_rate": 4.579368100494862e-06, + "loss": 1.5645, + "step": 529 + }, + { + "epoch": 0.544427324088341, + "grad_norm": 2.3278186321258545, + "learning_rate": 4.577464788732395e-06, + "loss": 1.5428, + "step": 530 + }, + { + "epoch": 0.544427324088341, + "eval_loss": 1.8242785930633545, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 530 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 2.5192365646362305, + "learning_rate": 4.575561476969928e-06, + "loss": 1.5896, + "step": 531 + }, + { + "epoch": 0.5464817668207499, + "grad_norm": 2.308701276779175, + "learning_rate": 4.573658165207461e-06, + "loss": 1.5518, + "step": 532 + }, + { + "epoch": 0.5475089881869543, + "grad_norm": 2.4221107959747314, + "learning_rate": 4.571754853444994e-06, + "loss": 1.5387, + "step": 533 + }, + { + "epoch": 0.5485362095531587, + "grad_norm": 2.457606077194214, + "learning_rate": 4.569851541682528e-06, + "loss": 1.4997, + "step": 534 + }, + { + "epoch": 0.5495634309193631, + "grad_norm": 2.3406503200531006, + "learning_rate": 4.567948229920061e-06, + "loss": 1.4761, + "step": 535 + }, + { + "epoch": 0.5505906522855676, + "grad_norm": 2.555712938308716, + "learning_rate": 4.566044918157594e-06, + "loss": 1.5114, + "step": 536 + }, + { + "epoch": 0.551617873651772, + "grad_norm": 2.373013734817505, + "learning_rate": 4.564141606395128e-06, + "loss": 1.5516, + "step": 537 + }, + { + "epoch": 0.5526450950179764, + "grad_norm": 2.3915255069732666, + "learning_rate": 4.562238294632661e-06, + "loss": 1.5145, + "step": 538 + }, + { + "epoch": 0.5536723163841808, + "grad_norm": 2.3272323608398438, + "learning_rate": 4.5603349828701945e-06, + "loss": 1.5472, + "step": 539 + }, + { + "epoch": 0.5546995377503852, + "grad_norm": 2.416072130203247, + "learning_rate": 4.558431671107728e-06, + "loss": 1.5842, + "step": 540 + }, + { + "epoch": 0.5546995377503852, + "eval_loss": 1.825427532196045, + "eval_runtime": 37.9905, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 540 + }, + { + "epoch": 0.5557267591165896, + "grad_norm": 2.3225271701812744, + "learning_rate": 4.556528359345261e-06, + "loss": 1.5188, + "step": 541 + }, + { + "epoch": 0.556753980482794, + "grad_norm": 2.2387030124664307, + "learning_rate": 4.5546250475827946e-06, + "loss": 1.4778, + "step": 542 + }, + { + "epoch": 0.5577812018489985, + "grad_norm": 2.3054494857788086, + "learning_rate": 4.552721735820328e-06, + "loss": 1.5229, + "step": 543 + }, + { + "epoch": 0.5588084232152029, + "grad_norm": 2.4956626892089844, + "learning_rate": 4.550818424057861e-06, + "loss": 1.5367, + "step": 544 + }, + { + "epoch": 0.5598356445814073, + "grad_norm": 2.454986810684204, + "learning_rate": 4.548915112295395e-06, + "loss": 1.5444, + "step": 545 + }, + { + "epoch": 0.5608628659476117, + "grad_norm": 2.466810703277588, + "learning_rate": 4.547011800532928e-06, + "loss": 1.4914, + "step": 546 + }, + { + "epoch": 0.5618900873138162, + "grad_norm": 2.51777982711792, + "learning_rate": 4.545108488770461e-06, + "loss": 1.5775, + "step": 547 + }, + { + "epoch": 0.5629173086800205, + "grad_norm": 2.340421199798584, + "learning_rate": 4.543205177007995e-06, + "loss": 1.5185, + "step": 548 + }, + { + "epoch": 0.5639445300462249, + "grad_norm": 2.376772880554199, + "learning_rate": 4.541301865245527e-06, + "loss": 1.5588, + "step": 549 + }, + { + "epoch": 0.5649717514124294, + "grad_norm": 2.4004335403442383, + "learning_rate": 4.539398553483061e-06, + "loss": 1.5283, + "step": 550 + }, + { + "epoch": 0.5649717514124294, + "eval_loss": 1.8294963836669922, + "eval_runtime": 37.9391, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 550 + }, + { + "epoch": 0.5659989727786338, + "grad_norm": 2.424459934234619, + "learning_rate": 4.537495241720594e-06, + "loss": 1.5576, + "step": 551 + }, + { + "epoch": 0.5670261941448382, + "grad_norm": 2.4413998126983643, + "learning_rate": 4.535591929958127e-06, + "loss": 1.5444, + "step": 552 + }, + { + "epoch": 0.5680534155110426, + "grad_norm": 2.408897876739502, + "learning_rate": 4.533688618195661e-06, + "loss": 1.5269, + "step": 553 + }, + { + "epoch": 0.5690806368772471, + "grad_norm": 2.471024990081787, + "learning_rate": 4.531785306433194e-06, + "loss": 1.5109, + "step": 554 + }, + { + "epoch": 0.5701078582434514, + "grad_norm": 2.374882459640503, + "learning_rate": 4.529881994670727e-06, + "loss": 1.5205, + "step": 555 + }, + { + "epoch": 0.5711350796096559, + "grad_norm": 2.4166030883789062, + "learning_rate": 4.527978682908261e-06, + "loss": 1.5276, + "step": 556 + }, + { + "epoch": 0.5721623009758603, + "grad_norm": 2.2681620121002197, + "learning_rate": 4.526075371145794e-06, + "loss": 1.5374, + "step": 557 + }, + { + "epoch": 0.5731895223420647, + "grad_norm": 2.3933286666870117, + "learning_rate": 4.5241720593833275e-06, + "loss": 1.5349, + "step": 558 + }, + { + "epoch": 0.5742167437082691, + "grad_norm": 2.4086451530456543, + "learning_rate": 4.52226874762086e-06, + "loss": 1.5512, + "step": 559 + }, + { + "epoch": 0.5752439650744735, + "grad_norm": 2.559021472930908, + "learning_rate": 4.520365435858393e-06, + "loss": 1.5789, + "step": 560 + }, + { + "epoch": 0.5752439650744735, + "eval_loss": 1.8243234157562256, + "eval_runtime": 37.934, + "eval_samples_per_second": 26.362, + "eval_steps_per_second": 1.107, + "step": 560 + }, + { + "epoch": 0.576271186440678, + "grad_norm": 2.483494520187378, + "learning_rate": 4.518462124095928e-06, + "loss": 1.5348, + "step": 561 + }, + { + "epoch": 0.5772984078068824, + "grad_norm": 2.4039993286132812, + "learning_rate": 4.516558812333461e-06, + "loss": 1.5201, + "step": 562 + }, + { + "epoch": 0.5783256291730868, + "grad_norm": 2.5080959796905518, + "learning_rate": 4.514655500570994e-06, + "loss": 1.5864, + "step": 563 + }, + { + "epoch": 0.5793528505392912, + "grad_norm": 2.620887279510498, + "learning_rate": 4.512752188808528e-06, + "loss": 1.5901, + "step": 564 + }, + { + "epoch": 0.5803800719054957, + "grad_norm": 2.5691096782684326, + "learning_rate": 4.510848877046061e-06, + "loss": 1.5506, + "step": 565 + }, + { + "epoch": 0.5814072932717, + "grad_norm": 2.3777835369110107, + "learning_rate": 4.5089455652835936e-06, + "loss": 1.5149, + "step": 566 + }, + { + "epoch": 0.5824345146379045, + "grad_norm": 2.2791666984558105, + "learning_rate": 4.507042253521127e-06, + "loss": 1.4771, + "step": 567 + }, + { + "epoch": 0.5834617360041089, + "grad_norm": 2.4953627586364746, + "learning_rate": 4.50513894175866e-06, + "loss": 1.556, + "step": 568 + }, + { + "epoch": 0.5844889573703133, + "grad_norm": 2.3563215732574463, + "learning_rate": 4.503235629996194e-06, + "loss": 1.549, + "step": 569 + }, + { + "epoch": 0.5855161787365177, + "grad_norm": 2.49676513671875, + "learning_rate": 4.501332318233727e-06, + "loss": 1.513, + "step": 570 + }, + { + "epoch": 0.5855161787365177, + "eval_loss": 1.8353043794631958, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 570 + }, + { + "epoch": 0.5865434001027221, + "grad_norm": 2.385275363922119, + "learning_rate": 4.49942900647126e-06, + "loss": 1.5622, + "step": 571 + }, + { + "epoch": 0.5875706214689266, + "grad_norm": 2.3089914321899414, + "learning_rate": 4.497525694708794e-06, + "loss": 1.5426, + "step": 572 + }, + { + "epoch": 0.588597842835131, + "grad_norm": 2.307323455810547, + "learning_rate": 4.495622382946327e-06, + "loss": 1.5088, + "step": 573 + }, + { + "epoch": 0.5896250642013354, + "grad_norm": 2.597288131713867, + "learning_rate": 4.4937190711838605e-06, + "loss": 1.5788, + "step": 574 + }, + { + "epoch": 0.5906522855675398, + "grad_norm": 2.3923256397247314, + "learning_rate": 4.491815759421394e-06, + "loss": 1.5038, + "step": 575 + }, + { + "epoch": 0.5916795069337443, + "grad_norm": 2.4199156761169434, + "learning_rate": 4.489912447658926e-06, + "loss": 1.5057, + "step": 576 + }, + { + "epoch": 0.5927067282999486, + "grad_norm": 2.5643393993377686, + "learning_rate": 4.48800913589646e-06, + "loss": 1.5274, + "step": 577 + }, + { + "epoch": 0.593733949666153, + "grad_norm": 2.5854485034942627, + "learning_rate": 4.486105824133993e-06, + "loss": 1.5341, + "step": 578 + }, + { + "epoch": 0.5947611710323575, + "grad_norm": 2.5363597869873047, + "learning_rate": 4.4842025123715264e-06, + "loss": 1.4953, + "step": 579 + }, + { + "epoch": 0.5957883923985619, + "grad_norm": 2.3880693912506104, + "learning_rate": 4.48229920060906e-06, + "loss": 1.4868, + "step": 580 + }, + { + "epoch": 0.5957883923985619, + "eval_loss": 1.8251410722732544, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 580 + }, + { + "epoch": 0.5968156137647663, + "grad_norm": 2.5272583961486816, + "learning_rate": 4.480395888846594e-06, + "loss": 1.5851, + "step": 581 + }, + { + "epoch": 0.5978428351309707, + "grad_norm": 2.4496943950653076, + "learning_rate": 4.478492577084127e-06, + "loss": 1.4896, + "step": 582 + }, + { + "epoch": 0.5988700564971752, + "grad_norm": 2.3996055126190186, + "learning_rate": 4.47658926532166e-06, + "loss": 1.5253, + "step": 583 + }, + { + "epoch": 0.5998972778633795, + "grad_norm": 2.5964202880859375, + "learning_rate": 4.474685953559193e-06, + "loss": 1.6294, + "step": 584 + }, + { + "epoch": 0.600924499229584, + "grad_norm": 2.5567739009857178, + "learning_rate": 4.472782641796727e-06, + "loss": 1.4834, + "step": 585 + }, + { + "epoch": 0.6019517205957884, + "grad_norm": 2.440291166305542, + "learning_rate": 4.47087933003426e-06, + "loss": 1.4966, + "step": 586 + }, + { + "epoch": 0.6029789419619929, + "grad_norm": 2.525763988494873, + "learning_rate": 4.468976018271793e-06, + "loss": 1.5005, + "step": 587 + }, + { + "epoch": 0.6040061633281972, + "grad_norm": 2.4822962284088135, + "learning_rate": 4.467072706509327e-06, + "loss": 1.5358, + "step": 588 + }, + { + "epoch": 0.6050333846944016, + "grad_norm": 2.591149091720581, + "learning_rate": 4.46516939474686e-06, + "loss": 1.4975, + "step": 589 + }, + { + "epoch": 0.6060606060606061, + "grad_norm": 2.4884238243103027, + "learning_rate": 4.463266082984393e-06, + "loss": 1.6107, + "step": 590 + }, + { + "epoch": 0.6060606060606061, + "eval_loss": 1.8383466005325317, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 590 + }, + { + "epoch": 0.6070878274268104, + "grad_norm": 2.4965732097625732, + "learning_rate": 4.461362771221927e-06, + "loss": 1.4986, + "step": 591 + }, + { + "epoch": 0.6081150487930149, + "grad_norm": 2.544792413711548, + "learning_rate": 4.45945945945946e-06, + "loss": 1.5067, + "step": 592 + }, + { + "epoch": 0.6091422701592193, + "grad_norm": 2.5726170539855957, + "learning_rate": 4.457556147696993e-06, + "loss": 1.5861, + "step": 593 + }, + { + "epoch": 0.6101694915254238, + "grad_norm": 2.493027448654175, + "learning_rate": 4.455652835934526e-06, + "loss": 1.5857, + "step": 594 + }, + { + "epoch": 0.6111967128916281, + "grad_norm": 2.470771074295044, + "learning_rate": 4.453749524172059e-06, + "loss": 1.5419, + "step": 595 + }, + { + "epoch": 0.6122239342578326, + "grad_norm": 2.568052053451538, + "learning_rate": 4.451846212409593e-06, + "loss": 1.5391, + "step": 596 + }, + { + "epoch": 0.613251155624037, + "grad_norm": 2.4346213340759277, + "learning_rate": 4.449942900647126e-06, + "loss": 1.4761, + "step": 597 + }, + { + "epoch": 0.6142783769902413, + "grad_norm": 2.418133497238159, + "learning_rate": 4.4480395888846595e-06, + "loss": 1.5383, + "step": 598 + }, + { + "epoch": 0.6153055983564458, + "grad_norm": 2.4795567989349365, + "learning_rate": 4.446136277122193e-06, + "loss": 1.4777, + "step": 599 + }, + { + "epoch": 0.6163328197226502, + "grad_norm": 2.5122742652893066, + "learning_rate": 4.444232965359726e-06, + "loss": 1.5353, + "step": 600 + }, + { + "epoch": 0.6163328197226502, + "eval_loss": 1.8325241804122925, + "eval_runtime": 37.9546, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 600 + }, + { + "epoch": 0.6173600410888547, + "grad_norm": 2.604067325592041, + "learning_rate": 4.4423296535972596e-06, + "loss": 1.489, + "step": 601 + }, + { + "epoch": 0.618387262455059, + "grad_norm": 2.3741753101348877, + "learning_rate": 4.440426341834793e-06, + "loss": 1.532, + "step": 602 + }, + { + "epoch": 0.6194144838212635, + "grad_norm": 2.5373036861419678, + "learning_rate": 4.438523030072326e-06, + "loss": 1.561, + "step": 603 + }, + { + "epoch": 0.6204417051874679, + "grad_norm": 2.2958099842071533, + "learning_rate": 4.43661971830986e-06, + "loss": 1.5025, + "step": 604 + }, + { + "epoch": 0.6214689265536724, + "grad_norm": 2.417006015777588, + "learning_rate": 4.434716406547393e-06, + "loss": 1.4977, + "step": 605 + }, + { + "epoch": 0.6224961479198767, + "grad_norm": 2.335296392440796, + "learning_rate": 4.432813094784926e-06, + "loss": 1.5504, + "step": 606 + }, + { + "epoch": 0.6235233692860811, + "grad_norm": 2.4086828231811523, + "learning_rate": 4.43090978302246e-06, + "loss": 1.5217, + "step": 607 + }, + { + "epoch": 0.6245505906522856, + "grad_norm": 2.3816819190979004, + "learning_rate": 4.429006471259993e-06, + "loss": 1.5126, + "step": 608 + }, + { + "epoch": 0.6255778120184899, + "grad_norm": 2.307671546936035, + "learning_rate": 4.4271031594975265e-06, + "loss": 1.4876, + "step": 609 + }, + { + "epoch": 0.6266050333846944, + "grad_norm": 2.355837821960449, + "learning_rate": 4.425199847735059e-06, + "loss": 1.4827, + "step": 610 + }, + { + "epoch": 0.6266050333846944, + "eval_loss": 1.8269455432891846, + "eval_runtime": 37.6459, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 610 + }, + { + "epoch": 0.6276322547508988, + "grad_norm": 2.558631658554077, + "learning_rate": 4.423296535972592e-06, + "loss": 1.5172, + "step": 611 + }, + { + "epoch": 0.6286594761171033, + "grad_norm": 2.3170175552368164, + "learning_rate": 4.421393224210126e-06, + "loss": 1.4784, + "step": 612 + }, + { + "epoch": 0.6296866974833076, + "grad_norm": 2.4064524173736572, + "learning_rate": 4.419489912447659e-06, + "loss": 1.5616, + "step": 613 + }, + { + "epoch": 0.6307139188495121, + "grad_norm": 2.3719160556793213, + "learning_rate": 4.4175866006851924e-06, + "loss": 1.5292, + "step": 614 + }, + { + "epoch": 0.6317411402157165, + "grad_norm": 2.3808202743530273, + "learning_rate": 4.415683288922726e-06, + "loss": 1.5122, + "step": 615 + }, + { + "epoch": 0.632768361581921, + "grad_norm": 2.4208037853240967, + "learning_rate": 4.413779977160259e-06, + "loss": 1.5088, + "step": 616 + }, + { + "epoch": 0.6337955829481253, + "grad_norm": 2.5558979511260986, + "learning_rate": 4.4118766653977925e-06, + "loss": 1.5678, + "step": 617 + }, + { + "epoch": 0.6348228043143297, + "grad_norm": 2.443908452987671, + "learning_rate": 4.409973353635326e-06, + "loss": 1.5234, + "step": 618 + }, + { + "epoch": 0.6358500256805342, + "grad_norm": 2.420408010482788, + "learning_rate": 4.408070041872859e-06, + "loss": 1.4708, + "step": 619 + }, + { + "epoch": 0.6368772470467385, + "grad_norm": 2.3884904384613037, + "learning_rate": 4.406166730110393e-06, + "loss": 1.4821, + "step": 620 + }, + { + "epoch": 0.6368772470467385, + "eval_loss": 1.8217289447784424, + "eval_runtime": 37.7936, + "eval_samples_per_second": 26.459, + "eval_steps_per_second": 1.111, + "step": 620 + }, + { + "epoch": 0.637904468412943, + "grad_norm": 2.313720703125, + "learning_rate": 4.404263418347925e-06, + "loss": 1.4962, + "step": 621 + }, + { + "epoch": 0.6389316897791474, + "grad_norm": 2.3173866271972656, + "learning_rate": 4.402360106585459e-06, + "loss": 1.4857, + "step": 622 + }, + { + "epoch": 0.6399589111453519, + "grad_norm": 2.541757822036743, + "learning_rate": 4.400456794822993e-06, + "loss": 1.5177, + "step": 623 + }, + { + "epoch": 0.6409861325115562, + "grad_norm": 2.4382357597351074, + "learning_rate": 4.398553483060526e-06, + "loss": 1.4718, + "step": 624 + }, + { + "epoch": 0.6420133538777607, + "grad_norm": 2.2988059520721436, + "learning_rate": 4.396650171298059e-06, + "loss": 1.4883, + "step": 625 + }, + { + "epoch": 0.6430405752439651, + "grad_norm": 2.4817495346069336, + "learning_rate": 4.394746859535593e-06, + "loss": 1.4823, + "step": 626 + }, + { + "epoch": 0.6440677966101694, + "grad_norm": 2.4578585624694824, + "learning_rate": 4.392843547773125e-06, + "loss": 1.4882, + "step": 627 + }, + { + "epoch": 0.6450950179763739, + "grad_norm": 2.184972047805786, + "learning_rate": 4.390940236010659e-06, + "loss": 1.514, + "step": 628 + }, + { + "epoch": 0.6461222393425783, + "grad_norm": 2.4040279388427734, + "learning_rate": 4.389036924248192e-06, + "loss": 1.5343, + "step": 629 + }, + { + "epoch": 0.6471494607087828, + "grad_norm": 2.3676459789276123, + "learning_rate": 4.387133612485725e-06, + "loss": 1.5007, + "step": 630 + }, + { + "epoch": 0.6471494607087828, + "eval_loss": 1.8106164932250977, + "eval_runtime": 37.9383, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 630 + }, + { + "epoch": 0.6481766820749871, + "grad_norm": 2.234689235687256, + "learning_rate": 4.385230300723259e-06, + "loss": 1.5064, + "step": 631 + }, + { + "epoch": 0.6492039034411916, + "grad_norm": 2.281742572784424, + "learning_rate": 4.383326988960792e-06, + "loss": 1.4777, + "step": 632 + }, + { + "epoch": 0.650231124807396, + "grad_norm": 2.4314322471618652, + "learning_rate": 4.3814236771983255e-06, + "loss": 1.5331, + "step": 633 + }, + { + "epoch": 0.6512583461736005, + "grad_norm": 2.4092588424682617, + "learning_rate": 4.379520365435859e-06, + "loss": 1.5122, + "step": 634 + }, + { + "epoch": 0.6522855675398048, + "grad_norm": 2.517362117767334, + "learning_rate": 4.377617053673392e-06, + "loss": 1.52, + "step": 635 + }, + { + "epoch": 0.6533127889060092, + "grad_norm": 2.4784109592437744, + "learning_rate": 4.3757137419109256e-06, + "loss": 1.5771, + "step": 636 + }, + { + "epoch": 0.6543400102722137, + "grad_norm": 2.32181715965271, + "learning_rate": 4.373810430148459e-06, + "loss": 1.4718, + "step": 637 + }, + { + "epoch": 0.655367231638418, + "grad_norm": 2.5146260261535645, + "learning_rate": 4.3719071183859914e-06, + "loss": 1.5307, + "step": 638 + }, + { + "epoch": 0.6563944530046225, + "grad_norm": 2.2364494800567627, + "learning_rate": 4.370003806623525e-06, + "loss": 1.4803, + "step": 639 + }, + { + "epoch": 0.6574216743708269, + "grad_norm": 2.3735806941986084, + "learning_rate": 4.368100494861058e-06, + "loss": 1.5249, + "step": 640 + }, + { + "epoch": 0.6574216743708269, + "eval_loss": 1.8134880065917969, + "eval_runtime": 38.0376, + "eval_samples_per_second": 26.29, + "eval_steps_per_second": 1.104, + "step": 640 + }, + { + "epoch": 0.6584488957370314, + "grad_norm": 2.4600815773010254, + "learning_rate": 4.3661971830985915e-06, + "loss": 1.4893, + "step": 641 + }, + { + "epoch": 0.6594761171032357, + "grad_norm": 2.4106783866882324, + "learning_rate": 4.364293871336126e-06, + "loss": 1.5163, + "step": 642 + }, + { + "epoch": 0.6605033384694402, + "grad_norm": 2.276624917984009, + "learning_rate": 4.362390559573659e-06, + "loss": 1.482, + "step": 643 + }, + { + "epoch": 0.6615305598356446, + "grad_norm": 2.336341381072998, + "learning_rate": 4.3604872478111925e-06, + "loss": 1.4152, + "step": 644 + }, + { + "epoch": 0.662557781201849, + "grad_norm": 2.3329105377197266, + "learning_rate": 4.358583936048725e-06, + "loss": 1.5411, + "step": 645 + }, + { + "epoch": 0.6635850025680534, + "grad_norm": 2.2877161502838135, + "learning_rate": 4.356680624286258e-06, + "loss": 1.4997, + "step": 646 + }, + { + "epoch": 0.6646122239342578, + "grad_norm": 2.410820484161377, + "learning_rate": 4.354777312523792e-06, + "loss": 1.5446, + "step": 647 + }, + { + "epoch": 0.6656394453004623, + "grad_norm": 2.403038501739502, + "learning_rate": 4.352874000761325e-06, + "loss": 1.4916, + "step": 648 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 2.3261778354644775, + "learning_rate": 4.3509706889988584e-06, + "loss": 1.5561, + "step": 649 + }, + { + "epoch": 0.6676938880328711, + "grad_norm": 2.4512174129486084, + "learning_rate": 4.349067377236392e-06, + "loss": 1.556, + "step": 650 + }, + { + "epoch": 0.6676938880328711, + "eval_loss": 1.8238279819488525, + "eval_runtime": 37.9988, + "eval_samples_per_second": 26.317, + "eval_steps_per_second": 1.105, + "step": 650 + }, + { + "epoch": 0.6687211093990755, + "grad_norm": 2.4523823261260986, + "learning_rate": 4.347164065473925e-06, + "loss": 1.5676, + "step": 651 + }, + { + "epoch": 0.66974833076528, + "grad_norm": 2.3770365715026855, + "learning_rate": 4.3452607537114585e-06, + "loss": 1.4815, + "step": 652 + }, + { + "epoch": 0.6707755521314843, + "grad_norm": 2.2440760135650635, + "learning_rate": 4.343357441948992e-06, + "loss": 1.5141, + "step": 653 + }, + { + "epoch": 0.6718027734976888, + "grad_norm": 2.3341948986053467, + "learning_rate": 4.341454130186525e-06, + "loss": 1.4859, + "step": 654 + }, + { + "epoch": 0.6728299948638932, + "grad_norm": 2.507354497909546, + "learning_rate": 4.339550818424058e-06, + "loss": 1.4734, + "step": 655 + }, + { + "epoch": 0.6738572162300975, + "grad_norm": 2.395954132080078, + "learning_rate": 4.337647506661591e-06, + "loss": 1.5069, + "step": 656 + }, + { + "epoch": 0.674884437596302, + "grad_norm": 2.362211227416992, + "learning_rate": 4.3357441948991245e-06, + "loss": 1.5299, + "step": 657 + }, + { + "epoch": 0.6759116589625064, + "grad_norm": 2.460008144378662, + "learning_rate": 4.333840883136658e-06, + "loss": 1.4795, + "step": 658 + }, + { + "epoch": 0.6769388803287109, + "grad_norm": 2.3578217029571533, + "learning_rate": 4.331937571374191e-06, + "loss": 1.4906, + "step": 659 + }, + { + "epoch": 0.6779661016949152, + "grad_norm": 2.398843288421631, + "learning_rate": 4.3300342596117246e-06, + "loss": 1.4986, + "step": 660 + }, + { + "epoch": 0.6779661016949152, + "eval_loss": 1.8206881284713745, + "eval_runtime": 38.044, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 660 + }, + { + "epoch": 0.6789933230611197, + "grad_norm": 2.3587260246276855, + "learning_rate": 4.328130947849258e-06, + "loss": 1.461, + "step": 661 + }, + { + "epoch": 0.6800205444273241, + "grad_norm": 2.311952590942383, + "learning_rate": 4.326227636086791e-06, + "loss": 1.5041, + "step": 662 + }, + { + "epoch": 0.6810477657935285, + "grad_norm": 2.376188278198242, + "learning_rate": 4.324324324324325e-06, + "loss": 1.5042, + "step": 663 + }, + { + "epoch": 0.6820749871597329, + "grad_norm": 2.5807230472564697, + "learning_rate": 4.322421012561858e-06, + "loss": 1.4935, + "step": 664 + }, + { + "epoch": 0.6831022085259374, + "grad_norm": 2.368128538131714, + "learning_rate": 4.320517700799391e-06, + "loss": 1.5851, + "step": 665 + }, + { + "epoch": 0.6841294298921418, + "grad_norm": 2.4517524242401123, + "learning_rate": 4.318614389036925e-06, + "loss": 1.5528, + "step": 666 + }, + { + "epoch": 0.6851566512583461, + "grad_norm": 2.5436811447143555, + "learning_rate": 4.316711077274458e-06, + "loss": 1.4633, + "step": 667 + }, + { + "epoch": 0.6861838726245506, + "grad_norm": 2.4287915229797363, + "learning_rate": 4.3148077655119915e-06, + "loss": 1.5454, + "step": 668 + }, + { + "epoch": 0.687211093990755, + "grad_norm": 2.386528730392456, + "learning_rate": 4.312904453749525e-06, + "loss": 1.5116, + "step": 669 + }, + { + "epoch": 0.6882383153569595, + "grad_norm": 2.6602251529693604, + "learning_rate": 4.311001141987058e-06, + "loss": 1.5199, + "step": 670 + }, + { + "epoch": 0.6882383153569595, + "eval_loss": 1.8252663612365723, + "eval_runtime": 38.0791, + "eval_samples_per_second": 26.261, + "eval_steps_per_second": 1.103, + "step": 670 + }, + { + "epoch": 0.6892655367231638, + "grad_norm": 2.351128339767456, + "learning_rate": 4.3090978302245916e-06, + "loss": 1.507, + "step": 671 + }, + { + "epoch": 0.6902927580893683, + "grad_norm": 2.393441677093506, + "learning_rate": 4.307194518462124e-06, + "loss": 1.579, + "step": 672 + }, + { + "epoch": 0.6913199794555727, + "grad_norm": 2.3730578422546387, + "learning_rate": 4.3052912066996574e-06, + "loss": 1.5446, + "step": 673 + }, + { + "epoch": 0.6923472008217771, + "grad_norm": 2.2522945404052734, + "learning_rate": 4.303387894937191e-06, + "loss": 1.4705, + "step": 674 + }, + { + "epoch": 0.6933744221879815, + "grad_norm": 2.384563446044922, + "learning_rate": 4.301484583174724e-06, + "loss": 1.4636, + "step": 675 + }, + { + "epoch": 0.6944016435541859, + "grad_norm": 2.2879116535186768, + "learning_rate": 4.2995812714122575e-06, + "loss": 1.4914, + "step": 676 + }, + { + "epoch": 0.6954288649203904, + "grad_norm": 2.494788646697998, + "learning_rate": 4.297677959649791e-06, + "loss": 1.5126, + "step": 677 + }, + { + "epoch": 0.6964560862865947, + "grad_norm": 2.700012683868408, + "learning_rate": 4.295774647887324e-06, + "loss": 1.5362, + "step": 678 + }, + { + "epoch": 0.6974833076527992, + "grad_norm": 2.5130302906036377, + "learning_rate": 4.293871336124858e-06, + "loss": 1.5075, + "step": 679 + }, + { + "epoch": 0.6985105290190036, + "grad_norm": 2.486236572265625, + "learning_rate": 4.291968024362391e-06, + "loss": 1.5338, + "step": 680 + }, + { + "epoch": 0.6985105290190036, + "eval_loss": 1.8229657411575317, + "eval_runtime": 38.0642, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 680 + }, + { + "epoch": 0.699537750385208, + "grad_norm": 2.441683530807495, + "learning_rate": 4.290064712599924e-06, + "loss": 1.537, + "step": 681 + }, + { + "epoch": 0.7005649717514124, + "grad_norm": 2.396798610687256, + "learning_rate": 4.288161400837458e-06, + "loss": 1.4553, + "step": 682 + }, + { + "epoch": 0.7015921931176169, + "grad_norm": 2.3636221885681152, + "learning_rate": 4.286258089074991e-06, + "loss": 1.5138, + "step": 683 + }, + { + "epoch": 0.7026194144838213, + "grad_norm": 2.4584708213806152, + "learning_rate": 4.284354777312524e-06, + "loss": 1.4725, + "step": 684 + }, + { + "epoch": 0.7036466358500256, + "grad_norm": 2.850332498550415, + "learning_rate": 4.282451465550058e-06, + "loss": 1.5075, + "step": 685 + }, + { + "epoch": 0.7046738572162301, + "grad_norm": 2.6345887184143066, + "learning_rate": 4.280548153787591e-06, + "loss": 1.5054, + "step": 686 + }, + { + "epoch": 0.7057010785824345, + "grad_norm": 2.3535311222076416, + "learning_rate": 4.2786448420251245e-06, + "loss": 1.4607, + "step": 687 + }, + { + "epoch": 0.706728299948639, + "grad_norm": 2.490431547164917, + "learning_rate": 4.276741530262658e-06, + "loss": 1.4852, + "step": 688 + }, + { + "epoch": 0.7077555213148433, + "grad_norm": 2.4353036880493164, + "learning_rate": 4.27483821850019e-06, + "loss": 1.5503, + "step": 689 + }, + { + "epoch": 0.7087827426810478, + "grad_norm": 2.3356940746307373, + "learning_rate": 4.272934906737724e-06, + "loss": 1.5416, + "step": 690 + }, + { + "epoch": 0.7087827426810478, + "eval_loss": 1.8164218664169312, + "eval_runtime": 38.0592, + "eval_samples_per_second": 26.275, + "eval_steps_per_second": 1.104, + "step": 690 + }, + { + "epoch": 0.7098099640472522, + "grad_norm": 2.4772775173187256, + "learning_rate": 4.271031594975257e-06, + "loss": 1.5392, + "step": 691 + }, + { + "epoch": 0.7108371854134566, + "grad_norm": 2.2921438217163086, + "learning_rate": 4.2691282832127905e-06, + "loss": 1.4935, + "step": 692 + }, + { + "epoch": 0.711864406779661, + "grad_norm": 2.3176305294036865, + "learning_rate": 4.267224971450324e-06, + "loss": 1.4451, + "step": 693 + }, + { + "epoch": 0.7128916281458655, + "grad_norm": 2.6825685501098633, + "learning_rate": 4.265321659687857e-06, + "loss": 1.4878, + "step": 694 + }, + { + "epoch": 0.7139188495120699, + "grad_norm": 2.6141414642333984, + "learning_rate": 4.2634183479253906e-06, + "loss": 1.4695, + "step": 695 + }, + { + "epoch": 0.7149460708782742, + "grad_norm": 2.5179429054260254, + "learning_rate": 4.261515036162924e-06, + "loss": 1.492, + "step": 696 + }, + { + "epoch": 0.7159732922444787, + "grad_norm": 2.4744884967803955, + "learning_rate": 4.259611724400457e-06, + "loss": 1.4752, + "step": 697 + }, + { + "epoch": 0.7170005136106831, + "grad_norm": 2.4911048412323, + "learning_rate": 4.257708412637991e-06, + "loss": 1.423, + "step": 698 + }, + { + "epoch": 0.7180277349768875, + "grad_norm": 2.5116214752197266, + "learning_rate": 4.255805100875523e-06, + "loss": 1.4535, + "step": 699 + }, + { + "epoch": 0.7190549563430919, + "grad_norm": 2.476651430130005, + "learning_rate": 4.2539017891130565e-06, + "loss": 1.5349, + "step": 700 + }, + { + "epoch": 0.7190549563430919, + "eval_loss": 1.8069179058074951, + "eval_runtime": 38.0292, + "eval_samples_per_second": 26.296, + "eval_steps_per_second": 1.104, + "step": 700 + }, + { + "epoch": 0.7200821777092964, + "grad_norm": 2.5234575271606445, + "learning_rate": 4.25199847735059e-06, + "loss": 1.4596, + "step": 701 + }, + { + "epoch": 0.7211093990755008, + "grad_norm": 2.33290433883667, + "learning_rate": 4.250095165588123e-06, + "loss": 1.5147, + "step": 702 + }, + { + "epoch": 0.7221366204417052, + "grad_norm": 2.479883909225464, + "learning_rate": 4.2481918538256575e-06, + "loss": 1.47, + "step": 703 + }, + { + "epoch": 0.7231638418079096, + "grad_norm": 2.432253837585449, + "learning_rate": 4.246288542063191e-06, + "loss": 1.5152, + "step": 704 + }, + { + "epoch": 0.724191063174114, + "grad_norm": 2.358304738998413, + "learning_rate": 4.244385230300724e-06, + "loss": 1.4562, + "step": 705 + }, + { + "epoch": 0.7252182845403184, + "grad_norm": 2.480877161026001, + "learning_rate": 4.242481918538257e-06, + "loss": 1.4906, + "step": 706 + }, + { + "epoch": 0.7262455059065228, + "grad_norm": 2.4287192821502686, + "learning_rate": 4.24057860677579e-06, + "loss": 1.5294, + "step": 707 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 2.4782888889312744, + "learning_rate": 4.2386752950133234e-06, + "loss": 1.4767, + "step": 708 + }, + { + "epoch": 0.7282999486389317, + "grad_norm": 2.4948253631591797, + "learning_rate": 4.236771983250857e-06, + "loss": 1.4664, + "step": 709 + }, + { + "epoch": 0.7293271700051361, + "grad_norm": 2.4625205993652344, + "learning_rate": 4.23486867148839e-06, + "loss": 1.495, + "step": 710 + }, + { + "epoch": 0.7293271700051361, + "eval_loss": 1.8140079975128174, + "eval_runtime": 37.6964, + "eval_samples_per_second": 26.528, + "eval_steps_per_second": 1.114, + "step": 710 + }, + { + "epoch": 0.7303543913713405, + "grad_norm": 2.669027805328369, + "learning_rate": 4.2329653597259235e-06, + "loss": 1.5204, + "step": 711 + }, + { + "epoch": 0.731381612737545, + "grad_norm": 2.5024707317352295, + "learning_rate": 4.231062047963457e-06, + "loss": 1.5283, + "step": 712 + }, + { + "epoch": 0.7324088341037494, + "grad_norm": 2.498258590698242, + "learning_rate": 4.22915873620099e-06, + "loss": 1.5139, + "step": 713 + }, + { + "epoch": 0.7334360554699538, + "grad_norm": 2.4305667877197266, + "learning_rate": 4.227255424438524e-06, + "loss": 1.5337, + "step": 714 + }, + { + "epoch": 0.7344632768361582, + "grad_norm": 2.3153419494628906, + "learning_rate": 4.225352112676057e-06, + "loss": 1.4646, + "step": 715 + }, + { + "epoch": 0.7354904982023626, + "grad_norm": 2.2835583686828613, + "learning_rate": 4.2234488009135895e-06, + "loss": 1.4328, + "step": 716 + }, + { + "epoch": 0.736517719568567, + "grad_norm": 2.401418924331665, + "learning_rate": 4.221545489151123e-06, + "loss": 1.4685, + "step": 717 + }, + { + "epoch": 0.7375449409347714, + "grad_norm": 2.454864978790283, + "learning_rate": 4.219642177388656e-06, + "loss": 1.4925, + "step": 718 + }, + { + "epoch": 0.7385721623009759, + "grad_norm": 2.2332873344421387, + "learning_rate": 4.2177388656261896e-06, + "loss": 1.4481, + "step": 719 + }, + { + "epoch": 0.7395993836671803, + "grad_norm": 2.46628475189209, + "learning_rate": 4.215835553863723e-06, + "loss": 1.4173, + "step": 720 + }, + { + "epoch": 0.7395993836671803, + "eval_loss": 1.817592740058899, + "eval_runtime": 37.8086, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 720 + }, + { + "epoch": 0.7406266050333847, + "grad_norm": 2.346245765686035, + "learning_rate": 4.213932242101256e-06, + "loss": 1.4688, + "step": 721 + }, + { + "epoch": 0.7416538263995891, + "grad_norm": 2.2708640098571777, + "learning_rate": 4.21202893033879e-06, + "loss": 1.4937, + "step": 722 + }, + { + "epoch": 0.7426810477657936, + "grad_norm": 2.3716516494750977, + "learning_rate": 4.210125618576323e-06, + "loss": 1.4869, + "step": 723 + }, + { + "epoch": 0.743708269131998, + "grad_norm": 2.443058490753174, + "learning_rate": 4.208222306813856e-06, + "loss": 1.542, + "step": 724 + }, + { + "epoch": 0.7447354904982023, + "grad_norm": 2.327270984649658, + "learning_rate": 4.20631899505139e-06, + "loss": 1.3933, + "step": 725 + }, + { + "epoch": 0.7457627118644068, + "grad_norm": 2.3734512329101562, + "learning_rate": 4.204415683288923e-06, + "loss": 1.5109, + "step": 726 + }, + { + "epoch": 0.7467899332306112, + "grad_norm": 2.3631174564361572, + "learning_rate": 4.2025123715264565e-06, + "loss": 1.474, + "step": 727 + }, + { + "epoch": 0.7478171545968156, + "grad_norm": 2.306068181991577, + "learning_rate": 4.20060905976399e-06, + "loss": 1.42, + "step": 728 + }, + { + "epoch": 0.74884437596302, + "grad_norm": 2.391889810562134, + "learning_rate": 4.198705748001523e-06, + "loss": 1.5178, + "step": 729 + }, + { + "epoch": 0.7498715973292245, + "grad_norm": 2.3151705265045166, + "learning_rate": 4.1968024362390566e-06, + "loss": 1.4808, + "step": 730 + }, + { + "epoch": 0.7498715973292245, + "eval_loss": 1.8157070875167847, + "eval_runtime": 37.8799, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 730 + }, + { + "epoch": 0.7508988186954288, + "grad_norm": 2.336479902267456, + "learning_rate": 4.19489912447659e-06, + "loss": 1.5438, + "step": 731 + }, + { + "epoch": 0.7519260400616333, + "grad_norm": 2.338242292404175, + "learning_rate": 4.192995812714123e-06, + "loss": 1.4869, + "step": 732 + }, + { + "epoch": 0.7529532614278377, + "grad_norm": 2.268650770187378, + "learning_rate": 4.191092500951656e-06, + "loss": 1.4787, + "step": 733 + }, + { + "epoch": 0.7539804827940421, + "grad_norm": 2.3010005950927734, + "learning_rate": 4.189189189189189e-06, + "loss": 1.4711, + "step": 734 + }, + { + "epoch": 0.7550077041602465, + "grad_norm": 2.54770827293396, + "learning_rate": 4.1872858774267225e-06, + "loss": 1.5351, + "step": 735 + }, + { + "epoch": 0.7560349255264509, + "grad_norm": 2.3412787914276123, + "learning_rate": 4.185382565664256e-06, + "loss": 1.4655, + "step": 736 + }, + { + "epoch": 0.7570621468926554, + "grad_norm": 2.266413927078247, + "learning_rate": 4.183479253901789e-06, + "loss": 1.4402, + "step": 737 + }, + { + "epoch": 0.7580893682588598, + "grad_norm": 2.2785308361053467, + "learning_rate": 4.181575942139323e-06, + "loss": 1.4874, + "step": 738 + }, + { + "epoch": 0.7591165896250642, + "grad_norm": 2.2710752487182617, + "learning_rate": 4.179672630376856e-06, + "loss": 1.4644, + "step": 739 + }, + { + "epoch": 0.7601438109912686, + "grad_norm": 2.2460498809814453, + "learning_rate": 4.177769318614389e-06, + "loss": 1.5267, + "step": 740 + }, + { + "epoch": 0.7601438109912686, + "eval_loss": 1.8187947273254395, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 740 + }, + { + "epoch": 0.7611710323574731, + "grad_norm": 2.3589370250701904, + "learning_rate": 4.175866006851923e-06, + "loss": 1.5115, + "step": 741 + }, + { + "epoch": 0.7621982537236774, + "grad_norm": 2.2818758487701416, + "learning_rate": 4.173962695089456e-06, + "loss": 1.4945, + "step": 742 + }, + { + "epoch": 0.7632254750898819, + "grad_norm": 2.338085889816284, + "learning_rate": 4.1720593833269894e-06, + "loss": 1.5076, + "step": 743 + }, + { + "epoch": 0.7642526964560863, + "grad_norm": 2.3874809741973877, + "learning_rate": 4.170156071564523e-06, + "loss": 1.5665, + "step": 744 + }, + { + "epoch": 0.7652799178222907, + "grad_norm": 2.2418739795684814, + "learning_rate": 4.168252759802056e-06, + "loss": 1.4827, + "step": 745 + }, + { + "epoch": 0.7663071391884951, + "grad_norm": 2.441315174102783, + "learning_rate": 4.1663494480395895e-06, + "loss": 1.5171, + "step": 746 + }, + { + "epoch": 0.7673343605546995, + "grad_norm": 2.3201963901519775, + "learning_rate": 4.164446136277123e-06, + "loss": 1.5153, + "step": 747 + }, + { + "epoch": 0.768361581920904, + "grad_norm": 2.3395135402679443, + "learning_rate": 4.162542824514656e-06, + "loss": 1.4911, + "step": 748 + }, + { + "epoch": 0.7693888032871083, + "grad_norm": 2.3455941677093506, + "learning_rate": 4.16063951275219e-06, + "loss": 1.4493, + "step": 749 + }, + { + "epoch": 0.7704160246533128, + "grad_norm": 2.3737382888793945, + "learning_rate": 4.158736200989722e-06, + "loss": 1.4893, + "step": 750 + }, + { + "epoch": 0.7704160246533128, + "eval_loss": 1.8153141736984253, + "eval_runtime": 37.9216, + "eval_samples_per_second": 26.37, + "eval_steps_per_second": 1.108, + "step": 750 + }, + { + "epoch": 0.7714432460195172, + "grad_norm": 2.503803014755249, + "learning_rate": 4.1568328892272555e-06, + "loss": 1.499, + "step": 751 + }, + { + "epoch": 0.7724704673857217, + "grad_norm": 2.469083547592163, + "learning_rate": 4.154929577464789e-06, + "loss": 1.4984, + "step": 752 + }, + { + "epoch": 0.773497688751926, + "grad_norm": 2.387991428375244, + "learning_rate": 4.153026265702322e-06, + "loss": 1.4283, + "step": 753 + }, + { + "epoch": 0.7745249101181304, + "grad_norm": 2.3355393409729004, + "learning_rate": 4.1511229539398556e-06, + "loss": 1.4869, + "step": 754 + }, + { + "epoch": 0.7755521314843349, + "grad_norm": 2.629345417022705, + "learning_rate": 4.149219642177389e-06, + "loss": 1.4885, + "step": 755 + }, + { + "epoch": 0.7765793528505393, + "grad_norm": 2.458944797515869, + "learning_rate": 4.147316330414922e-06, + "loss": 1.415, + "step": 756 + }, + { + "epoch": 0.7776065742167437, + "grad_norm": 2.40051531791687, + "learning_rate": 4.145413018652456e-06, + "loss": 1.5129, + "step": 757 + }, + { + "epoch": 0.7786337955829481, + "grad_norm": 2.5249011516571045, + "learning_rate": 4.143509706889989e-06, + "loss": 1.5002, + "step": 758 + }, + { + "epoch": 0.7796610169491526, + "grad_norm": 2.413422107696533, + "learning_rate": 4.141606395127522e-06, + "loss": 1.4837, + "step": 759 + }, + { + "epoch": 0.7806882383153569, + "grad_norm": 2.3438029289245605, + "learning_rate": 4.139703083365056e-06, + "loss": 1.5213, + "step": 760 + }, + { + "epoch": 0.7806882383153569, + "eval_loss": 1.8037596940994263, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 760 + }, + { + "epoch": 0.7817154596815614, + "grad_norm": 2.326828718185425, + "learning_rate": 4.137799771602588e-06, + "loss": 1.5031, + "step": 761 + }, + { + "epoch": 0.7827426810477658, + "grad_norm": 2.4902493953704834, + "learning_rate": 4.135896459840122e-06, + "loss": 1.4602, + "step": 762 + }, + { + "epoch": 0.7837699024139703, + "grad_norm": 2.488004684448242, + "learning_rate": 4.133993148077656e-06, + "loss": 1.4764, + "step": 763 + }, + { + "epoch": 0.7847971237801746, + "grad_norm": 2.5515761375427246, + "learning_rate": 4.132089836315189e-06, + "loss": 1.4946, + "step": 764 + }, + { + "epoch": 0.785824345146379, + "grad_norm": 2.556278705596924, + "learning_rate": 4.1301865245527226e-06, + "loss": 1.4338, + "step": 765 + }, + { + "epoch": 0.7868515665125835, + "grad_norm": 2.3017969131469727, + "learning_rate": 4.128283212790256e-06, + "loss": 1.4053, + "step": 766 + }, + { + "epoch": 0.7878787878787878, + "grad_norm": 2.4371304512023926, + "learning_rate": 4.1263799010277884e-06, + "loss": 1.5238, + "step": 767 + }, + { + "epoch": 0.7889060092449923, + "grad_norm": 2.6256046295166016, + "learning_rate": 4.124476589265322e-06, + "loss": 1.4672, + "step": 768 + }, + { + "epoch": 0.7899332306111967, + "grad_norm": 2.3725669384002686, + "learning_rate": 4.122573277502855e-06, + "loss": 1.4677, + "step": 769 + }, + { + "epoch": 0.7909604519774012, + "grad_norm": 2.2928454875946045, + "learning_rate": 4.1206699657403885e-06, + "loss": 1.4927, + "step": 770 + }, + { + "epoch": 0.7909604519774012, + "eval_loss": 1.792726755142212, + "eval_runtime": 37.9273, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 770 + }, + { + "epoch": 0.7919876733436055, + "grad_norm": 2.315674066543579, + "learning_rate": 4.118766653977922e-06, + "loss": 1.4632, + "step": 771 + }, + { + "epoch": 0.79301489470981, + "grad_norm": 2.418849468231201, + "learning_rate": 4.116863342215455e-06, + "loss": 1.446, + "step": 772 + }, + { + "epoch": 0.7940421160760144, + "grad_norm": 2.3209879398345947, + "learning_rate": 4.114960030452989e-06, + "loss": 1.4755, + "step": 773 + }, + { + "epoch": 0.7950693374422187, + "grad_norm": 2.2322115898132324, + "learning_rate": 4.113056718690522e-06, + "loss": 1.4149, + "step": 774 + }, + { + "epoch": 0.7960965588084232, + "grad_norm": 2.6517558097839355, + "learning_rate": 4.111153406928055e-06, + "loss": 1.5323, + "step": 775 + }, + { + "epoch": 0.7971237801746276, + "grad_norm": 2.4602737426757812, + "learning_rate": 4.109250095165589e-06, + "loss": 1.5179, + "step": 776 + }, + { + "epoch": 0.7981510015408321, + "grad_norm": 2.4015536308288574, + "learning_rate": 4.107346783403122e-06, + "loss": 1.4711, + "step": 777 + }, + { + "epoch": 0.7991782229070364, + "grad_norm": 2.4367337226867676, + "learning_rate": 4.105443471640655e-06, + "loss": 1.4425, + "step": 778 + }, + { + "epoch": 0.8002054442732409, + "grad_norm": 2.5161020755767822, + "learning_rate": 4.103540159878188e-06, + "loss": 1.4832, + "step": 779 + }, + { + "epoch": 0.8012326656394453, + "grad_norm": 2.483154773712158, + "learning_rate": 4.101636848115721e-06, + "loss": 1.4508, + "step": 780 + }, + { + "epoch": 0.8012326656394453, + "eval_loss": 1.792397379875183, + "eval_runtime": 37.8937, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 780 + }, + { + "epoch": 0.8022598870056498, + "grad_norm": 2.4649806022644043, + "learning_rate": 4.099733536353255e-06, + "loss": 1.4919, + "step": 781 + }, + { + "epoch": 0.8032871083718541, + "grad_norm": 2.578634262084961, + "learning_rate": 4.097830224590788e-06, + "loss": 1.565, + "step": 782 + }, + { + "epoch": 0.8043143297380585, + "grad_norm": 2.444427967071533, + "learning_rate": 4.095926912828322e-06, + "loss": 1.5309, + "step": 783 + }, + { + "epoch": 0.805341551104263, + "grad_norm": 2.32051420211792, + "learning_rate": 4.094023601065855e-06, + "loss": 1.4777, + "step": 784 + }, + { + "epoch": 0.8063687724704673, + "grad_norm": 2.3207180500030518, + "learning_rate": 4.092120289303388e-06, + "loss": 1.4863, + "step": 785 + }, + { + "epoch": 0.8073959938366718, + "grad_norm": 2.1404120922088623, + "learning_rate": 4.0902169775409215e-06, + "loss": 1.437, + "step": 786 + }, + { + "epoch": 0.8084232152028762, + "grad_norm": 2.2980856895446777, + "learning_rate": 4.088313665778455e-06, + "loss": 1.4571, + "step": 787 + }, + { + "epoch": 0.8094504365690807, + "grad_norm": 2.2555243968963623, + "learning_rate": 4.086410354015988e-06, + "loss": 1.4451, + "step": 788 + }, + { + "epoch": 0.810477657935285, + "grad_norm": 2.3837077617645264, + "learning_rate": 4.0845070422535216e-06, + "loss": 1.4641, + "step": 789 + }, + { + "epoch": 0.8115048793014895, + "grad_norm": 2.5199060440063477, + "learning_rate": 4.082603730491055e-06, + "loss": 1.5048, + "step": 790 + }, + { + "epoch": 0.8115048793014895, + "eval_loss": 1.8019492626190186, + "eval_runtime": 37.8805, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 790 + }, + { + "epoch": 0.8125321006676939, + "grad_norm": 2.4010138511657715, + "learning_rate": 4.080700418728588e-06, + "loss": 1.4881, + "step": 791 + }, + { + "epoch": 0.8135593220338984, + "grad_norm": 2.342094898223877, + "learning_rate": 4.078797106966122e-06, + "loss": 1.3792, + "step": 792 + }, + { + "epoch": 0.8145865434001027, + "grad_norm": 2.276772975921631, + "learning_rate": 4.076893795203655e-06, + "loss": 1.4539, + "step": 793 + }, + { + "epoch": 0.8156137647663071, + "grad_norm": 2.569275140762329, + "learning_rate": 4.074990483441188e-06, + "loss": 1.5059, + "step": 794 + }, + { + "epoch": 0.8166409861325116, + "grad_norm": 2.345761299133301, + "learning_rate": 4.073087171678721e-06, + "loss": 1.494, + "step": 795 + }, + { + "epoch": 0.8176682074987159, + "grad_norm": 2.4425504207611084, + "learning_rate": 4.071183859916254e-06, + "loss": 1.471, + "step": 796 + }, + { + "epoch": 0.8186954288649204, + "grad_norm": 2.5758635997772217, + "learning_rate": 4.069280548153788e-06, + "loss": 1.4378, + "step": 797 + }, + { + "epoch": 0.8197226502311248, + "grad_norm": 2.4320778846740723, + "learning_rate": 4.067377236391321e-06, + "loss": 1.474, + "step": 798 + }, + { + "epoch": 0.8207498715973293, + "grad_norm": 2.466978073120117, + "learning_rate": 4.065473924628854e-06, + "loss": 1.4997, + "step": 799 + }, + { + "epoch": 0.8217770929635336, + "grad_norm": 2.462489128112793, + "learning_rate": 4.063570612866388e-06, + "loss": 1.5, + "step": 800 + }, + { + "epoch": 0.8217770929635336, + "eval_loss": 1.783312201499939, + "eval_runtime": 37.8832, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 800 + }, + { + "epoch": 0.8228043143297381, + "grad_norm": 2.367455005645752, + "learning_rate": 4.061667301103921e-06, + "loss": 1.4658, + "step": 801 + }, + { + "epoch": 0.8238315356959425, + "grad_norm": 2.292250394821167, + "learning_rate": 4.0597639893414544e-06, + "loss": 1.476, + "step": 802 + }, + { + "epoch": 0.8248587570621468, + "grad_norm": 2.3460025787353516, + "learning_rate": 4.057860677578988e-06, + "loss": 1.4932, + "step": 803 + }, + { + "epoch": 0.8258859784283513, + "grad_norm": 2.3708906173706055, + "learning_rate": 4.055957365816521e-06, + "loss": 1.4396, + "step": 804 + }, + { + "epoch": 0.8269131997945557, + "grad_norm": 2.353513479232788, + "learning_rate": 4.0540540540540545e-06, + "loss": 1.4285, + "step": 805 + }, + { + "epoch": 0.8279404211607602, + "grad_norm": 2.319577932357788, + "learning_rate": 4.052150742291588e-06, + "loss": 1.4326, + "step": 806 + }, + { + "epoch": 0.8289676425269645, + "grad_norm": 2.450364112854004, + "learning_rate": 4.050247430529121e-06, + "loss": 1.5109, + "step": 807 + }, + { + "epoch": 0.829994863893169, + "grad_norm": 2.393907070159912, + "learning_rate": 4.048344118766655e-06, + "loss": 1.4948, + "step": 808 + }, + { + "epoch": 0.8310220852593734, + "grad_norm": 2.336002826690674, + "learning_rate": 4.046440807004188e-06, + "loss": 1.4851, + "step": 809 + }, + { + "epoch": 0.8320493066255779, + "grad_norm": 2.281139850616455, + "learning_rate": 4.044537495241721e-06, + "loss": 1.4445, + "step": 810 + }, + { + "epoch": 0.8320493066255779, + "eval_loss": 1.784231185913086, + "eval_runtime": 37.6614, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 810 + }, + { + "epoch": 0.8330765279917822, + "grad_norm": 2.4848175048828125, + "learning_rate": 4.042634183479255e-06, + "loss": 1.4982, + "step": 811 + }, + { + "epoch": 0.8341037493579867, + "grad_norm": 2.263347864151001, + "learning_rate": 4.040730871716787e-06, + "loss": 1.4795, + "step": 812 + }, + { + "epoch": 0.8351309707241911, + "grad_norm": 2.394005060195923, + "learning_rate": 4.038827559954321e-06, + "loss": 1.5231, + "step": 813 + }, + { + "epoch": 0.8361581920903954, + "grad_norm": 2.5051357746124268, + "learning_rate": 4.036924248191854e-06, + "loss": 1.4886, + "step": 814 + }, + { + "epoch": 0.8371854134565999, + "grad_norm": 2.3580374717712402, + "learning_rate": 4.035020936429387e-06, + "loss": 1.4713, + "step": 815 + }, + { + "epoch": 0.8382126348228043, + "grad_norm": 2.3179264068603516, + "learning_rate": 4.033117624666921e-06, + "loss": 1.4369, + "step": 816 + }, + { + "epoch": 0.8392398561890088, + "grad_norm": 2.269911289215088, + "learning_rate": 4.031214312904454e-06, + "loss": 1.4326, + "step": 817 + }, + { + "epoch": 0.8402670775552131, + "grad_norm": 2.3177807331085205, + "learning_rate": 4.029311001141987e-06, + "loss": 1.4096, + "step": 818 + }, + { + "epoch": 0.8412942989214176, + "grad_norm": 2.394324541091919, + "learning_rate": 4.027407689379521e-06, + "loss": 1.4647, + "step": 819 + }, + { + "epoch": 0.842321520287622, + "grad_norm": 2.288536548614502, + "learning_rate": 4.025504377617054e-06, + "loss": 1.4472, + "step": 820 + }, + { + "epoch": 0.842321520287622, + "eval_loss": 1.7811495065689087, + "eval_runtime": 37.8132, + "eval_samples_per_second": 26.446, + "eval_steps_per_second": 1.111, + "step": 820 + }, + { + "epoch": 0.8433487416538265, + "grad_norm": 2.376574754714966, + "learning_rate": 4.0236010658545875e-06, + "loss": 1.4918, + "step": 821 + }, + { + "epoch": 0.8443759630200308, + "grad_norm": 2.5850327014923096, + "learning_rate": 4.02169775409212e-06, + "loss": 1.5342, + "step": 822 + }, + { + "epoch": 0.8454031843862352, + "grad_norm": 2.247206211090088, + "learning_rate": 4.019794442329653e-06, + "loss": 1.4177, + "step": 823 + }, + { + "epoch": 0.8464304057524397, + "grad_norm": 2.4411795139312744, + "learning_rate": 4.0178911305671876e-06, + "loss": 1.4992, + "step": 824 + }, + { + "epoch": 0.847457627118644, + "grad_norm": 2.3377397060394287, + "learning_rate": 4.015987818804721e-06, + "loss": 1.4359, + "step": 825 + }, + { + "epoch": 0.8484848484848485, + "grad_norm": 2.329857349395752, + "learning_rate": 4.014084507042254e-06, + "loss": 1.4096, + "step": 826 + }, + { + "epoch": 0.8495120698510529, + "grad_norm": 2.563685178756714, + "learning_rate": 4.012181195279788e-06, + "loss": 1.5174, + "step": 827 + }, + { + "epoch": 0.8505392912172574, + "grad_norm": 2.4521079063415527, + "learning_rate": 4.010277883517321e-06, + "loss": 1.5028, + "step": 828 + }, + { + "epoch": 0.8515665125834617, + "grad_norm": 2.4091477394104004, + "learning_rate": 4.0083745717548535e-06, + "loss": 1.4615, + "step": 829 + }, + { + "epoch": 0.8525937339496662, + "grad_norm": 2.4010677337646484, + "learning_rate": 4.006471259992387e-06, + "loss": 1.4834, + "step": 830 + }, + { + "epoch": 0.8525937339496662, + "eval_loss": 1.7797678709030151, + "eval_runtime": 37.9378, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 830 + }, + { + "epoch": 0.8536209553158706, + "grad_norm": 2.3712635040283203, + "learning_rate": 4.00456794822992e-06, + "loss": 1.5196, + "step": 831 + }, + { + "epoch": 0.8546481766820749, + "grad_norm": 2.4096601009368896, + "learning_rate": 4.002664636467454e-06, + "loss": 1.4769, + "step": 832 + }, + { + "epoch": 0.8556753980482794, + "grad_norm": 2.408466100692749, + "learning_rate": 4.000761324704987e-06, + "loss": 1.4268, + "step": 833 + }, + { + "epoch": 0.8567026194144838, + "grad_norm": 2.6857047080993652, + "learning_rate": 3.99885801294252e-06, + "loss": 1.492, + "step": 834 + }, + { + "epoch": 0.8577298407806883, + "grad_norm": 2.297550916671753, + "learning_rate": 3.996954701180054e-06, + "loss": 1.4184, + "step": 835 + }, + { + "epoch": 0.8587570621468926, + "grad_norm": 2.423224925994873, + "learning_rate": 3.995051389417587e-06, + "loss": 1.4337, + "step": 836 + }, + { + "epoch": 0.8597842835130971, + "grad_norm": 2.260493040084839, + "learning_rate": 3.9931480776551204e-06, + "loss": 1.5267, + "step": 837 + }, + { + "epoch": 0.8608115048793015, + "grad_norm": 2.3803012371063232, + "learning_rate": 3.991244765892654e-06, + "loss": 1.4426, + "step": 838 + }, + { + "epoch": 0.861838726245506, + "grad_norm": 2.1425721645355225, + "learning_rate": 3.989341454130186e-06, + "loss": 1.4148, + "step": 839 + }, + { + "epoch": 0.8628659476117103, + "grad_norm": 2.316624641418457, + "learning_rate": 3.98743814236772e-06, + "loss": 1.4544, + "step": 840 + }, + { + "epoch": 0.8628659476117103, + "eval_loss": 1.7824597358703613, + "eval_runtime": 37.9676, + "eval_samples_per_second": 26.338, + "eval_steps_per_second": 1.106, + "step": 840 + }, + { + "epoch": 0.8638931689779148, + "grad_norm": 2.3723156452178955, + "learning_rate": 3.985534830605253e-06, + "loss": 1.473, + "step": 841 + }, + { + "epoch": 0.8649203903441192, + "grad_norm": 2.5167760848999023, + "learning_rate": 3.983631518842786e-06, + "loss": 1.4919, + "step": 842 + }, + { + "epoch": 0.8659476117103235, + "grad_norm": 2.3018670082092285, + "learning_rate": 3.98172820708032e-06, + "loss": 1.447, + "step": 843 + }, + { + "epoch": 0.866974833076528, + "grad_norm": 2.420483112335205, + "learning_rate": 3.979824895317854e-06, + "loss": 1.4159, + "step": 844 + }, + { + "epoch": 0.8680020544427324, + "grad_norm": 2.39443039894104, + "learning_rate": 3.977921583555387e-06, + "loss": 1.4984, + "step": 845 + }, + { + "epoch": 0.8690292758089369, + "grad_norm": 2.2953920364379883, + "learning_rate": 3.97601827179292e-06, + "loss": 1.4297, + "step": 846 + }, + { + "epoch": 0.8700564971751412, + "grad_norm": 2.2919528484344482, + "learning_rate": 3.974114960030453e-06, + "loss": 1.4728, + "step": 847 + }, + { + "epoch": 0.8710837185413457, + "grad_norm": 2.327914237976074, + "learning_rate": 3.9722116482679866e-06, + "loss": 1.4584, + "step": 848 + }, + { + "epoch": 0.8721109399075501, + "grad_norm": 2.305720567703247, + "learning_rate": 3.97030833650552e-06, + "loss": 1.4688, + "step": 849 + }, + { + "epoch": 0.8731381612737545, + "grad_norm": 2.2969236373901367, + "learning_rate": 3.968405024743053e-06, + "loss": 1.3972, + "step": 850 + }, + { + "epoch": 0.8731381612737545, + "eval_loss": 1.775634765625, + "eval_runtime": 37.9916, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 850 + }, + { + "epoch": 0.8741653826399589, + "grad_norm": 2.314387083053589, + "learning_rate": 3.966501712980587e-06, + "loss": 1.4476, + "step": 851 + }, + { + "epoch": 0.8751926040061633, + "grad_norm": 2.3271584510803223, + "learning_rate": 3.96459840121812e-06, + "loss": 1.4615, + "step": 852 + }, + { + "epoch": 0.8762198253723678, + "grad_norm": 2.494795322418213, + "learning_rate": 3.962695089455653e-06, + "loss": 1.5062, + "step": 853 + }, + { + "epoch": 0.8772470467385721, + "grad_norm": 2.4132444858551025, + "learning_rate": 3.960791777693187e-06, + "loss": 1.4861, + "step": 854 + }, + { + "epoch": 0.8782742681047766, + "grad_norm": 2.3869900703430176, + "learning_rate": 3.95888846593072e-06, + "loss": 1.4923, + "step": 855 + }, + { + "epoch": 0.879301489470981, + "grad_norm": 2.5167367458343506, + "learning_rate": 3.956985154168253e-06, + "loss": 1.5137, + "step": 856 + }, + { + "epoch": 0.8803287108371854, + "grad_norm": 2.510361671447754, + "learning_rate": 3.955081842405786e-06, + "loss": 1.5414, + "step": 857 + }, + { + "epoch": 0.8813559322033898, + "grad_norm": 2.554427146911621, + "learning_rate": 3.953178530643319e-06, + "loss": 1.5569, + "step": 858 + }, + { + "epoch": 0.8823831535695943, + "grad_norm": 2.306509256362915, + "learning_rate": 3.951275218880853e-06, + "loss": 1.3982, + "step": 859 + }, + { + "epoch": 0.8834103749357987, + "grad_norm": 2.3805363178253174, + "learning_rate": 3.949371907118386e-06, + "loss": 1.4977, + "step": 860 + }, + { + "epoch": 0.8834103749357987, + "eval_loss": 1.7882694005966187, + "eval_runtime": 38.0072, + "eval_samples_per_second": 26.311, + "eval_steps_per_second": 1.105, + "step": 860 + }, + { + "epoch": 0.884437596302003, + "grad_norm": 2.5413448810577393, + "learning_rate": 3.9474685953559194e-06, + "loss": 1.472, + "step": 861 + }, + { + "epoch": 0.8854648176682075, + "grad_norm": 2.484834909439087, + "learning_rate": 3.945565283593453e-06, + "loss": 1.4597, + "step": 862 + }, + { + "epoch": 0.8864920390344119, + "grad_norm": 2.6060585975646973, + "learning_rate": 3.943661971830986e-06, + "loss": 1.509, + "step": 863 + }, + { + "epoch": 0.8875192604006163, + "grad_norm": 2.6536927223205566, + "learning_rate": 3.9417586600685195e-06, + "loss": 1.4836, + "step": 864 + }, + { + "epoch": 0.8885464817668207, + "grad_norm": 2.5198960304260254, + "learning_rate": 3.939855348306053e-06, + "loss": 1.486, + "step": 865 + }, + { + "epoch": 0.8895737031330252, + "grad_norm": 2.335204839706421, + "learning_rate": 3.937952036543586e-06, + "loss": 1.4701, + "step": 866 + }, + { + "epoch": 0.8906009244992296, + "grad_norm": 2.7058098316192627, + "learning_rate": 3.93604872478112e-06, + "loss": 1.4798, + "step": 867 + }, + { + "epoch": 0.891628145865434, + "grad_norm": 2.4798922538757324, + "learning_rate": 3.934145413018653e-06, + "loss": 1.4364, + "step": 868 + }, + { + "epoch": 0.8926553672316384, + "grad_norm": 2.3090760707855225, + "learning_rate": 3.932242101256186e-06, + "loss": 1.4405, + "step": 869 + }, + { + "epoch": 0.8936825885978429, + "grad_norm": 2.4251503944396973, + "learning_rate": 3.93033878949372e-06, + "loss": 1.4113, + "step": 870 + }, + { + "epoch": 0.8936825885978429, + "eval_loss": 1.794334888458252, + "eval_runtime": 38.0197, + "eval_samples_per_second": 26.302, + "eval_steps_per_second": 1.105, + "step": 870 + }, + { + "epoch": 0.8947098099640473, + "grad_norm": 2.594945192337036, + "learning_rate": 3.928435477731253e-06, + "loss": 1.502, + "step": 871 + }, + { + "epoch": 0.8957370313302516, + "grad_norm": 2.4224843978881836, + "learning_rate": 3.9265321659687864e-06, + "loss": 1.5005, + "step": 872 + }, + { + "epoch": 0.8967642526964561, + "grad_norm": 2.3968186378479004, + "learning_rate": 3.924628854206319e-06, + "loss": 1.481, + "step": 873 + }, + { + "epoch": 0.8977914740626605, + "grad_norm": 2.4923524856567383, + "learning_rate": 3.922725542443852e-06, + "loss": 1.4259, + "step": 874 + }, + { + "epoch": 0.8988186954288649, + "grad_norm": 2.5386104583740234, + "learning_rate": 3.920822230681386e-06, + "loss": 1.4497, + "step": 875 + }, + { + "epoch": 0.8998459167950693, + "grad_norm": 2.4014909267425537, + "learning_rate": 3.918918918918919e-06, + "loss": 1.4816, + "step": 876 + }, + { + "epoch": 0.9008731381612738, + "grad_norm": 2.4019317626953125, + "learning_rate": 3.917015607156452e-06, + "loss": 1.4669, + "step": 877 + }, + { + "epoch": 0.9019003595274782, + "grad_norm": 2.5408380031585693, + "learning_rate": 3.915112295393986e-06, + "loss": 1.5343, + "step": 878 + }, + { + "epoch": 0.9029275808936826, + "grad_norm": 2.364053964614868, + "learning_rate": 3.913208983631519e-06, + "loss": 1.5229, + "step": 879 + }, + { + "epoch": 0.903954802259887, + "grad_norm": 2.307412624359131, + "learning_rate": 3.9113056718690525e-06, + "loss": 1.4444, + "step": 880 + }, + { + "epoch": 0.903954802259887, + "eval_loss": 1.7910975217819214, + "eval_runtime": 38.0659, + "eval_samples_per_second": 26.27, + "eval_steps_per_second": 1.103, + "step": 880 + }, + { + "epoch": 0.9049820236260914, + "grad_norm": 2.496135950088501, + "learning_rate": 3.909402360106586e-06, + "loss": 1.4853, + "step": 881 + }, + { + "epoch": 0.9060092449922958, + "grad_norm": 2.4825589656829834, + "learning_rate": 3.907499048344119e-06, + "loss": 1.4886, + "step": 882 + }, + { + "epoch": 0.9070364663585002, + "grad_norm": 2.3458824157714844, + "learning_rate": 3.905595736581652e-06, + "loss": 1.4882, + "step": 883 + }, + { + "epoch": 0.9080636877247047, + "grad_norm": 2.2699403762817383, + "learning_rate": 3.903692424819186e-06, + "loss": 1.4417, + "step": 884 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 2.348593235015869, + "learning_rate": 3.901789113056719e-06, + "loss": 1.4991, + "step": 885 + }, + { + "epoch": 0.9101181304571135, + "grad_norm": 2.3292129039764404, + "learning_rate": 3.899885801294253e-06, + "loss": 1.4752, + "step": 886 + }, + { + "epoch": 0.9111453518233179, + "grad_norm": 2.487183094024658, + "learning_rate": 3.897982489531786e-06, + "loss": 1.5317, + "step": 887 + }, + { + "epoch": 0.9121725731895224, + "grad_norm": 2.3300845623016357, + "learning_rate": 3.896079177769319e-06, + "loss": 1.4964, + "step": 888 + }, + { + "epoch": 0.9131997945557268, + "grad_norm": 2.3292312622070312, + "learning_rate": 3.894175866006853e-06, + "loss": 1.5221, + "step": 889 + }, + { + "epoch": 0.9142270159219312, + "grad_norm": 2.3556439876556396, + "learning_rate": 3.892272554244385e-06, + "loss": 1.4471, + "step": 890 + }, + { + "epoch": 0.9142270159219312, + "eval_loss": 1.7807039022445679, + "eval_runtime": 38.0614, + "eval_samples_per_second": 26.273, + "eval_steps_per_second": 1.103, + "step": 890 + }, + { + "epoch": 0.9152542372881356, + "grad_norm": 2.3302760124206543, + "learning_rate": 3.890369242481919e-06, + "loss": 1.5179, + "step": 891 + }, + { + "epoch": 0.91628145865434, + "grad_norm": 2.327072858810425, + "learning_rate": 3.888465930719452e-06, + "loss": 1.4474, + "step": 892 + }, + { + "epoch": 0.9173086800205444, + "grad_norm": 2.2773101329803467, + "learning_rate": 3.886562618956985e-06, + "loss": 1.4783, + "step": 893 + }, + { + "epoch": 0.9183359013867488, + "grad_norm": 2.2717974185943604, + "learning_rate": 3.884659307194519e-06, + "loss": 1.4062, + "step": 894 + }, + { + "epoch": 0.9193631227529533, + "grad_norm": 2.3350367546081543, + "learning_rate": 3.882755995432052e-06, + "loss": 1.483, + "step": 895 + }, + { + "epoch": 0.9203903441191577, + "grad_norm": 2.302811622619629, + "learning_rate": 3.8808526836695854e-06, + "loss": 1.4407, + "step": 896 + }, + { + "epoch": 0.9214175654853621, + "grad_norm": 2.4031271934509277, + "learning_rate": 3.878949371907119e-06, + "loss": 1.4331, + "step": 897 + }, + { + "epoch": 0.9224447868515665, + "grad_norm": 2.4806716442108154, + "learning_rate": 3.877046060144652e-06, + "loss": 1.4341, + "step": 898 + }, + { + "epoch": 0.923472008217771, + "grad_norm": 2.4243357181549072, + "learning_rate": 3.8751427483821855e-06, + "loss": 1.4701, + "step": 899 + }, + { + "epoch": 0.9244992295839753, + "grad_norm": 2.4953269958496094, + "learning_rate": 3.873239436619718e-06, + "loss": 1.4287, + "step": 900 + }, + { + "epoch": 0.9244992295839753, + "eval_loss": 1.7872728109359741, + "eval_runtime": 38.0906, + "eval_samples_per_second": 26.253, + "eval_steps_per_second": 1.103, + "step": 900 + }, + { + "epoch": 0.9255264509501797, + "grad_norm": 2.307490587234497, + "learning_rate": 3.871336124857251e-06, + "loss": 1.4693, + "step": 901 + }, + { + "epoch": 0.9265536723163842, + "grad_norm": 2.4745755195617676, + "learning_rate": 3.869432813094785e-06, + "loss": 1.4779, + "step": 902 + }, + { + "epoch": 0.9275808936825886, + "grad_norm": 2.367715358734131, + "learning_rate": 3.867529501332318e-06, + "loss": 1.4771, + "step": 903 + }, + { + "epoch": 0.928608115048793, + "grad_norm": 2.388549566268921, + "learning_rate": 3.8656261895698515e-06, + "loss": 1.4648, + "step": 904 + }, + { + "epoch": 0.9296353364149974, + "grad_norm": 2.39432954788208, + "learning_rate": 3.863722877807386e-06, + "loss": 1.4932, + "step": 905 + }, + { + "epoch": 0.9306625577812019, + "grad_norm": 2.3921315670013428, + "learning_rate": 3.861819566044919e-06, + "loss": 1.5312, + "step": 906 + }, + { + "epoch": 0.9316897791474062, + "grad_norm": 2.400240659713745, + "learning_rate": 3.859916254282452e-06, + "loss": 1.4651, + "step": 907 + }, + { + "epoch": 0.9327170005136107, + "grad_norm": 2.394559860229492, + "learning_rate": 3.858012942519985e-06, + "loss": 1.4759, + "step": 908 + }, + { + "epoch": 0.9337442218798151, + "grad_norm": 2.3365273475646973, + "learning_rate": 3.856109630757518e-06, + "loss": 1.4502, + "step": 909 + }, + { + "epoch": 0.9347714432460195, + "grad_norm": 2.465691328048706, + "learning_rate": 3.854206318995052e-06, + "loss": 1.5036, + "step": 910 + }, + { + "epoch": 0.9347714432460195, + "eval_loss": 1.7682362794876099, + "eval_runtime": 37.6624, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 910 + }, + { + "epoch": 0.9357986646122239, + "grad_norm": 2.3376758098602295, + "learning_rate": 3.852303007232585e-06, + "loss": 1.4342, + "step": 911 + }, + { + "epoch": 0.9368258859784283, + "grad_norm": 2.345855236053467, + "learning_rate": 3.850399695470118e-06, + "loss": 1.4444, + "step": 912 + }, + { + "epoch": 0.9378531073446328, + "grad_norm": 2.4444501399993896, + "learning_rate": 3.848496383707652e-06, + "loss": 1.4315, + "step": 913 + }, + { + "epoch": 0.9388803287108372, + "grad_norm": 2.4896373748779297, + "learning_rate": 3.846593071945185e-06, + "loss": 1.4228, + "step": 914 + }, + { + "epoch": 0.9399075500770416, + "grad_norm": 2.3970463275909424, + "learning_rate": 3.8446897601827185e-06, + "loss": 1.5283, + "step": 915 + }, + { + "epoch": 0.940934771443246, + "grad_norm": 2.5194883346557617, + "learning_rate": 3.842786448420252e-06, + "loss": 1.5445, + "step": 916 + }, + { + "epoch": 0.9419619928094505, + "grad_norm": 2.446465253829956, + "learning_rate": 3.840883136657785e-06, + "loss": 1.5248, + "step": 917 + }, + { + "epoch": 0.9429892141756548, + "grad_norm": 2.348477840423584, + "learning_rate": 3.838979824895318e-06, + "loss": 1.402, + "step": 918 + }, + { + "epoch": 0.9440164355418593, + "grad_norm": 2.4511539936065674, + "learning_rate": 3.837076513132851e-06, + "loss": 1.4893, + "step": 919 + }, + { + "epoch": 0.9450436569080637, + "grad_norm": 2.3801777362823486, + "learning_rate": 3.8351732013703844e-06, + "loss": 1.4457, + "step": 920 + }, + { + "epoch": 0.9450436569080637, + "eval_loss": 1.7810813188552856, + "eval_runtime": 37.7647, + "eval_samples_per_second": 26.48, + "eval_steps_per_second": 1.112, + "step": 920 + }, + { + "epoch": 0.9460708782742681, + "grad_norm": 2.5260043144226074, + "learning_rate": 3.833269889607918e-06, + "loss": 1.4267, + "step": 921 + }, + { + "epoch": 0.9470980996404725, + "grad_norm": 2.270653009414673, + "learning_rate": 3.831366577845451e-06, + "loss": 1.4313, + "step": 922 + }, + { + "epoch": 0.9481253210066769, + "grad_norm": 2.5123250484466553, + "learning_rate": 3.8294632660829845e-06, + "loss": 1.4588, + "step": 923 + }, + { + "epoch": 0.9491525423728814, + "grad_norm": 2.4242913722991943, + "learning_rate": 3.827559954320518e-06, + "loss": 1.4659, + "step": 924 + }, + { + "epoch": 0.9501797637390857, + "grad_norm": 2.4337117671966553, + "learning_rate": 3.825656642558051e-06, + "loss": 1.4838, + "step": 925 + }, + { + "epoch": 0.9512069851052902, + "grad_norm": 2.3577871322631836, + "learning_rate": 3.823753330795585e-06, + "loss": 1.4674, + "step": 926 + }, + { + "epoch": 0.9522342064714946, + "grad_norm": 2.2574679851531982, + "learning_rate": 3.821850019033118e-06, + "loss": 1.4011, + "step": 927 + }, + { + "epoch": 0.9532614278376991, + "grad_norm": 2.38043475151062, + "learning_rate": 3.819946707270651e-06, + "loss": 1.459, + "step": 928 + }, + { + "epoch": 0.9542886492039034, + "grad_norm": 2.4249954223632812, + "learning_rate": 3.818043395508185e-06, + "loss": 1.4649, + "step": 929 + }, + { + "epoch": 0.9553158705701078, + "grad_norm": 2.4420788288116455, + "learning_rate": 3.816140083745718e-06, + "loss": 1.532, + "step": 930 + }, + { + "epoch": 0.9553158705701078, + "eval_loss": 1.7805145978927612, + "eval_runtime": 37.8606, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 930 + }, + { + "epoch": 0.9563430919363123, + "grad_norm": 2.391042470932007, + "learning_rate": 3.8142367719832514e-06, + "loss": 1.4365, + "step": 931 + }, + { + "epoch": 0.9573703133025167, + "grad_norm": 2.462303876876831, + "learning_rate": 3.8123334602207844e-06, + "loss": 1.4367, + "step": 932 + }, + { + "epoch": 0.9583975346687211, + "grad_norm": 2.330714702606201, + "learning_rate": 3.8104301484583177e-06, + "loss": 1.427, + "step": 933 + }, + { + "epoch": 0.9594247560349255, + "grad_norm": 2.3475441932678223, + "learning_rate": 3.808526836695851e-06, + "loss": 1.4836, + "step": 934 + }, + { + "epoch": 0.96045197740113, + "grad_norm": 2.4546470642089844, + "learning_rate": 3.8066235249333845e-06, + "loss": 1.4601, + "step": 935 + }, + { + "epoch": 0.9614791987673343, + "grad_norm": 2.4604640007019043, + "learning_rate": 3.804720213170918e-06, + "loss": 1.4616, + "step": 936 + }, + { + "epoch": 0.9625064201335388, + "grad_norm": 2.3087422847747803, + "learning_rate": 3.8028169014084508e-06, + "loss": 1.4816, + "step": 937 + }, + { + "epoch": 0.9635336414997432, + "grad_norm": 2.375377893447876, + "learning_rate": 3.800913589645984e-06, + "loss": 1.471, + "step": 938 + }, + { + "epoch": 0.9645608628659477, + "grad_norm": 2.358741283416748, + "learning_rate": 3.7990102778835175e-06, + "loss": 1.4754, + "step": 939 + }, + { + "epoch": 0.965588084232152, + "grad_norm": 2.3990139961242676, + "learning_rate": 3.797106966121051e-06, + "loss": 1.534, + "step": 940 + }, + { + "epoch": 0.965588084232152, + "eval_loss": 1.7745739221572876, + "eval_runtime": 37.8883, + "eval_samples_per_second": 26.393, + "eval_steps_per_second": 1.109, + "step": 940 + }, + { + "epoch": 0.9666153055983564, + "grad_norm": 2.417396068572998, + "learning_rate": 3.7952036543585842e-06, + "loss": 1.4994, + "step": 941 + }, + { + "epoch": 0.9676425269645609, + "grad_norm": 2.323125123977661, + "learning_rate": 3.793300342596117e-06, + "loss": 1.5045, + "step": 942 + }, + { + "epoch": 0.9686697483307652, + "grad_norm": 2.4424829483032227, + "learning_rate": 3.7913970308336505e-06, + "loss": 1.4936, + "step": 943 + }, + { + "epoch": 0.9696969696969697, + "grad_norm": 2.2769429683685303, + "learning_rate": 3.789493719071184e-06, + "loss": 1.4453, + "step": 944 + }, + { + "epoch": 0.9707241910631741, + "grad_norm": 2.183701276779175, + "learning_rate": 3.7875904073087177e-06, + "loss": 1.4005, + "step": 945 + }, + { + "epoch": 0.9717514124293786, + "grad_norm": 3.4538655281066895, + "learning_rate": 3.785687095546251e-06, + "loss": 1.4521, + "step": 946 + }, + { + "epoch": 0.9727786337955829, + "grad_norm": 2.3735523223876953, + "learning_rate": 3.7837837837837844e-06, + "loss": 1.4328, + "step": 947 + }, + { + "epoch": 0.9738058551617874, + "grad_norm": 2.6214165687561035, + "learning_rate": 3.7818804720213178e-06, + "loss": 1.4004, + "step": 948 + }, + { + "epoch": 0.9748330765279918, + "grad_norm": 2.365091562271118, + "learning_rate": 3.7799771602588507e-06, + "loss": 1.4006, + "step": 949 + }, + { + "epoch": 0.9758602978941961, + "grad_norm": 2.1926634311676025, + "learning_rate": 3.778073848496384e-06, + "loss": 1.4587, + "step": 950 + }, + { + "epoch": 0.9758602978941961, + "eval_loss": 1.767502784729004, + "eval_runtime": 37.9287, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 950 + }, + { + "epoch": 0.9768875192604006, + "grad_norm": 2.3047192096710205, + "learning_rate": 3.7761705367339174e-06, + "loss": 1.4826, + "step": 951 + }, + { + "epoch": 0.977914740626605, + "grad_norm": 2.3306028842926025, + "learning_rate": 3.7742672249714508e-06, + "loss": 1.4914, + "step": 952 + }, + { + "epoch": 0.9789419619928095, + "grad_norm": 2.3409693241119385, + "learning_rate": 3.772363913208984e-06, + "loss": 1.4556, + "step": 953 + }, + { + "epoch": 0.9799691833590138, + "grad_norm": 2.3337934017181396, + "learning_rate": 3.770460601446517e-06, + "loss": 1.4751, + "step": 954 + }, + { + "epoch": 0.9809964047252183, + "grad_norm": 2.3354830741882324, + "learning_rate": 3.7685572896840504e-06, + "loss": 1.4176, + "step": 955 + }, + { + "epoch": 0.9820236260914227, + "grad_norm": 2.4049782752990723, + "learning_rate": 3.766653977921584e-06, + "loss": 1.3715, + "step": 956 + }, + { + "epoch": 0.9830508474576272, + "grad_norm": 2.5127065181732178, + "learning_rate": 3.764750666159117e-06, + "loss": 1.468, + "step": 957 + }, + { + "epoch": 0.9840780688238315, + "grad_norm": 2.5576820373535156, + "learning_rate": 3.7628473543966505e-06, + "loss": 1.4237, + "step": 958 + }, + { + "epoch": 0.9851052901900359, + "grad_norm": 2.394098997116089, + "learning_rate": 3.7609440426341835e-06, + "loss": 1.5126, + "step": 959 + }, + { + "epoch": 0.9861325115562404, + "grad_norm": 2.3358702659606934, + "learning_rate": 3.759040730871717e-06, + "loss": 1.4109, + "step": 960 + }, + { + "epoch": 0.9861325115562404, + "eval_loss": 1.7665778398513794, + "eval_runtime": 37.9028, + "eval_samples_per_second": 26.383, + "eval_steps_per_second": 1.108, + "step": 960 + }, + { + "epoch": 0.9871597329224447, + "grad_norm": 2.4213271141052246, + "learning_rate": 3.75713741910925e-06, + "loss": 1.3995, + "step": 961 + }, + { + "epoch": 0.9881869542886492, + "grad_norm": 2.367835760116577, + "learning_rate": 3.7552341073467836e-06, + "loss": 1.4491, + "step": 962 + }, + { + "epoch": 0.9892141756548536, + "grad_norm": 2.433840751647949, + "learning_rate": 3.753330795584317e-06, + "loss": 1.5078, + "step": 963 + }, + { + "epoch": 0.9902413970210581, + "grad_norm": 2.3976447582244873, + "learning_rate": 3.75142748382185e-06, + "loss": 1.4107, + "step": 964 + }, + { + "epoch": 0.9912686183872624, + "grad_norm": 2.4369053840637207, + "learning_rate": 3.749524172059384e-06, + "loss": 1.4569, + "step": 965 + }, + { + "epoch": 0.9922958397534669, + "grad_norm": 2.3203182220458984, + "learning_rate": 3.747620860296917e-06, + "loss": 1.4976, + "step": 966 + }, + { + "epoch": 0.9933230611196713, + "grad_norm": 2.431885242462158, + "learning_rate": 3.7457175485344504e-06, + "loss": 1.4945, + "step": 967 + }, + { + "epoch": 0.9943502824858758, + "grad_norm": 2.2997021675109863, + "learning_rate": 3.7438142367719837e-06, + "loss": 1.48, + "step": 968 + }, + { + "epoch": 0.9953775038520801, + "grad_norm": 2.2502849102020264, + "learning_rate": 3.741910925009517e-06, + "loss": 1.4322, + "step": 969 + }, + { + "epoch": 0.9964047252182845, + "grad_norm": 2.311049222946167, + "learning_rate": 3.7400076132470505e-06, + "loss": 1.4498, + "step": 970 + }, + { + "epoch": 0.9964047252182845, + "eval_loss": 1.771975040435791, + "eval_runtime": 37.852, + "eval_samples_per_second": 26.419, + "eval_steps_per_second": 1.11, + "step": 970 + }, + { + "epoch": 0.997431946584489, + "grad_norm": 2.431821584701538, + "learning_rate": 3.7381043014845834e-06, + "loss": 1.4796, + "step": 971 + }, + { + "epoch": 0.9984591679506933, + "grad_norm": 2.319145441055298, + "learning_rate": 3.7362009897221168e-06, + "loss": 1.459, + "step": 972 + }, + { + "epoch": 0.9994863893168978, + "grad_norm": 2.4445269107818604, + "learning_rate": 3.73429767795965e-06, + "loss": 1.4764, + "step": 973 + }, + { + "epoch": 1.0005136106831023, + "grad_norm": 2.272233009338379, + "learning_rate": 3.7323943661971835e-06, + "loss": 1.3799, + "step": 974 + }, + { + "epoch": 1.0015408320493067, + "grad_norm": 2.5421202182769775, + "learning_rate": 3.730491054434717e-06, + "loss": 1.3257, + "step": 975 + }, + { + "epoch": 1.002568053415511, + "grad_norm": 2.3988611698150635, + "learning_rate": 3.72858774267225e-06, + "loss": 1.4244, + "step": 976 + }, + { + "epoch": 1.0035952747817154, + "grad_norm": 2.5080084800720215, + "learning_rate": 3.726684430909783e-06, + "loss": 1.4461, + "step": 977 + }, + { + "epoch": 1.00462249614792, + "grad_norm": 2.411081314086914, + "learning_rate": 3.7247811191473165e-06, + "loss": 1.45, + "step": 978 + }, + { + "epoch": 1.0056497175141244, + "grad_norm": 2.729546070098877, + "learning_rate": 3.72287780738485e-06, + "loss": 1.4296, + "step": 979 + }, + { + "epoch": 1.0066769388803287, + "grad_norm": 2.5932579040527344, + "learning_rate": 3.7209744956223832e-06, + "loss": 1.4312, + "step": 980 + }, + { + "epoch": 1.0066769388803287, + "eval_loss": 1.770633339881897, + "eval_runtime": 37.8977, + "eval_samples_per_second": 26.387, + "eval_steps_per_second": 1.108, + "step": 980 + }, + { + "epoch": 1.007704160246533, + "grad_norm": 2.4512336254119873, + "learning_rate": 3.719071183859916e-06, + "loss": 1.3637, + "step": 981 + }, + { + "epoch": 1.0087313816127375, + "grad_norm": 2.5416886806488037, + "learning_rate": 3.7171678720974495e-06, + "loss": 1.4591, + "step": 982 + }, + { + "epoch": 1.009758602978942, + "grad_norm": 2.464679002761841, + "learning_rate": 3.715264560334983e-06, + "loss": 1.4306, + "step": 983 + }, + { + "epoch": 1.0107858243451464, + "grad_norm": 2.3745839595794678, + "learning_rate": 3.7133612485725163e-06, + "loss": 1.4426, + "step": 984 + }, + { + "epoch": 1.0118130457113508, + "grad_norm": 2.3606340885162354, + "learning_rate": 3.71145793681005e-06, + "loss": 1.3618, + "step": 985 + }, + { + "epoch": 1.0128402670775551, + "grad_norm": 2.5836591720581055, + "learning_rate": 3.7095546250475834e-06, + "loss": 1.4776, + "step": 986 + }, + { + "epoch": 1.0138674884437597, + "grad_norm": 2.4108684062957764, + "learning_rate": 3.7076513132851168e-06, + "loss": 1.4047, + "step": 987 + }, + { + "epoch": 1.014894709809964, + "grad_norm": 2.4316983222961426, + "learning_rate": 3.7057480015226497e-06, + "loss": 1.3772, + "step": 988 + }, + { + "epoch": 1.0159219311761685, + "grad_norm": 2.5510315895080566, + "learning_rate": 3.703844689760183e-06, + "loss": 1.4536, + "step": 989 + }, + { + "epoch": 1.0169491525423728, + "grad_norm": 2.4947116374969482, + "learning_rate": 3.7019413779977164e-06, + "loss": 1.434, + "step": 990 + }, + { + "epoch": 1.0169491525423728, + "eval_loss": 1.7928223609924316, + "eval_runtime": 37.8838, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 990 + }, + { + "epoch": 1.0179763739085772, + "grad_norm": 2.5938122272491455, + "learning_rate": 3.70003806623525e-06, + "loss": 1.4083, + "step": 991 + }, + { + "epoch": 1.0190035952747818, + "grad_norm": 2.472865581512451, + "learning_rate": 3.698134754472783e-06, + "loss": 1.4696, + "step": 992 + }, + { + "epoch": 1.0200308166409862, + "grad_norm": 2.374344825744629, + "learning_rate": 3.696231442710316e-06, + "loss": 1.403, + "step": 993 + }, + { + "epoch": 1.0210580380071905, + "grad_norm": 2.6928529739379883, + "learning_rate": 3.6943281309478495e-06, + "loss": 1.4216, + "step": 994 + }, + { + "epoch": 1.022085259373395, + "grad_norm": 2.640890598297119, + "learning_rate": 3.692424819185383e-06, + "loss": 1.4362, + "step": 995 + }, + { + "epoch": 1.0231124807395995, + "grad_norm": 2.45357084274292, + "learning_rate": 3.690521507422916e-06, + "loss": 1.382, + "step": 996 + }, + { + "epoch": 1.0241397021058039, + "grad_norm": 2.712517023086548, + "learning_rate": 3.6886181956604496e-06, + "loss": 1.395, + "step": 997 + }, + { + "epoch": 1.0251669234720082, + "grad_norm": 2.7127771377563477, + "learning_rate": 3.686714883897983e-06, + "loss": 1.4589, + "step": 998 + }, + { + "epoch": 1.0261941448382126, + "grad_norm": 2.5379409790039062, + "learning_rate": 3.684811572135516e-06, + "loss": 1.4551, + "step": 999 + }, + { + "epoch": 1.027221366204417, + "grad_norm": 2.5015134811401367, + "learning_rate": 3.6829082603730492e-06, + "loss": 1.4112, + "step": 1000 + }, + { + "epoch": 1.027221366204417, + "eval_loss": 1.7806837558746338, + "eval_runtime": 37.9091, + "eval_samples_per_second": 26.379, + "eval_steps_per_second": 1.108, + "step": 1000 + }, + { + "epoch": 1.0282485875706215, + "grad_norm": 2.481306314468384, + "learning_rate": 3.6810049486105826e-06, + "loss": 1.4469, + "step": 1001 + }, + { + "epoch": 1.029275808936826, + "grad_norm": 2.481520891189575, + "learning_rate": 3.679101636848116e-06, + "loss": 1.4284, + "step": 1002 + }, + { + "epoch": 1.0303030303030303, + "grad_norm": 2.4291019439697266, + "learning_rate": 3.6771983250856493e-06, + "loss": 1.4067, + "step": 1003 + }, + { + "epoch": 1.0313302516692346, + "grad_norm": 2.5361173152923584, + "learning_rate": 3.6752950133231822e-06, + "loss": 1.4347, + "step": 1004 + }, + { + "epoch": 1.0323574730354392, + "grad_norm": 2.5588085651397705, + "learning_rate": 3.6733917015607156e-06, + "loss": 1.4455, + "step": 1005 + }, + { + "epoch": 1.0333846944016436, + "grad_norm": 2.555091619491577, + "learning_rate": 3.6714883897982494e-06, + "loss": 1.399, + "step": 1006 + }, + { + "epoch": 1.034411915767848, + "grad_norm": 2.5489652156829834, + "learning_rate": 3.6695850780357828e-06, + "loss": 1.3894, + "step": 1007 + }, + { + "epoch": 1.0354391371340523, + "grad_norm": 2.5258426666259766, + "learning_rate": 3.667681766273316e-06, + "loss": 1.3962, + "step": 1008 + }, + { + "epoch": 1.0364663585002567, + "grad_norm": 2.455916166305542, + "learning_rate": 3.6657784545108495e-06, + "loss": 1.4773, + "step": 1009 + }, + { + "epoch": 1.0374935798664613, + "grad_norm": 2.5466208457946777, + "learning_rate": 3.6638751427483824e-06, + "loss": 1.4337, + "step": 1010 + }, + { + "epoch": 1.0374935798664613, + "eval_loss": 1.7840261459350586, + "eval_runtime": 37.6685, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 1010 + }, + { + "epoch": 1.0385208012326657, + "grad_norm": 2.5446081161499023, + "learning_rate": 3.6619718309859158e-06, + "loss": 1.4777, + "step": 1011 + }, + { + "epoch": 1.03954802259887, + "grad_norm": 2.509493350982666, + "learning_rate": 3.660068519223449e-06, + "loss": 1.4295, + "step": 1012 + }, + { + "epoch": 1.0405752439650744, + "grad_norm": 2.417632818222046, + "learning_rate": 3.6581652074609825e-06, + "loss": 1.4044, + "step": 1013 + }, + { + "epoch": 1.041602465331279, + "grad_norm": 2.457902193069458, + "learning_rate": 3.656261895698516e-06, + "loss": 1.3948, + "step": 1014 + }, + { + "epoch": 1.0426296866974833, + "grad_norm": 2.4869253635406494, + "learning_rate": 3.6543585839360492e-06, + "loss": 1.397, + "step": 1015 + }, + { + "epoch": 1.0436569080636877, + "grad_norm": 2.495419979095459, + "learning_rate": 3.652455272173582e-06, + "loss": 1.4037, + "step": 1016 + }, + { + "epoch": 1.044684129429892, + "grad_norm": 2.5231642723083496, + "learning_rate": 3.6505519604111155e-06, + "loss": 1.4156, + "step": 1017 + }, + { + "epoch": 1.0457113507960965, + "grad_norm": 2.8302087783813477, + "learning_rate": 3.648648648648649e-06, + "loss": 1.4529, + "step": 1018 + }, + { + "epoch": 1.046738572162301, + "grad_norm": 2.467430353164673, + "learning_rate": 3.6467453368861823e-06, + "loss": 1.4078, + "step": 1019 + }, + { + "epoch": 1.0477657935285054, + "grad_norm": 2.469128131866455, + "learning_rate": 3.6448420251237156e-06, + "loss": 1.4344, + "step": 1020 + }, + { + "epoch": 1.0477657935285054, + "eval_loss": 1.7800015211105347, + "eval_runtime": 37.854, + "eval_samples_per_second": 26.417, + "eval_steps_per_second": 1.11, + "step": 1020 + }, + { + "epoch": 1.0487930148947098, + "grad_norm": 2.59389328956604, + "learning_rate": 3.6429387133612486e-06, + "loss": 1.451, + "step": 1021 + }, + { + "epoch": 1.0498202362609141, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.641035401598782e-06, + "loss": 1.4026, + "step": 1022 + }, + { + "epoch": 1.0508474576271187, + "grad_norm": 2.601926565170288, + "learning_rate": 3.6391320898363153e-06, + "loss": 1.4382, + "step": 1023 + }, + { + "epoch": 1.051874678993323, + "grad_norm": 2.5336716175079346, + "learning_rate": 3.6372287780738487e-06, + "loss": 1.4184, + "step": 1024 + }, + { + "epoch": 1.0529019003595275, + "grad_norm": 2.503507614135742, + "learning_rate": 3.635325466311382e-06, + "loss": 1.4426, + "step": 1025 + }, + { + "epoch": 1.0539291217257318, + "grad_norm": 2.483013153076172, + "learning_rate": 3.633422154548916e-06, + "loss": 1.4563, + "step": 1026 + }, + { + "epoch": 1.0549563430919364, + "grad_norm": 2.4803354740142822, + "learning_rate": 3.6315188427864487e-06, + "loss": 1.4284, + "step": 1027 + }, + { + "epoch": 1.0559835644581408, + "grad_norm": 2.3735713958740234, + "learning_rate": 3.629615531023982e-06, + "loss": 1.3772, + "step": 1028 + }, + { + "epoch": 1.0570107858243452, + "grad_norm": 2.471226215362549, + "learning_rate": 3.6277122192615155e-06, + "loss": 1.4399, + "step": 1029 + }, + { + "epoch": 1.0580380071905495, + "grad_norm": 2.3827438354492188, + "learning_rate": 3.625808907499049e-06, + "loss": 1.3951, + "step": 1030 + }, + { + "epoch": 1.0580380071905495, + "eval_loss": 1.7787058353424072, + "eval_runtime": 37.9484, + "eval_samples_per_second": 26.352, + "eval_steps_per_second": 1.107, + "step": 1030 + }, + { + "epoch": 1.059065228556754, + "grad_norm": 2.403416633605957, + "learning_rate": 3.623905595736582e-06, + "loss": 1.3946, + "step": 1031 + }, + { + "epoch": 1.0600924499229585, + "grad_norm": 2.455740213394165, + "learning_rate": 3.6220022839741156e-06, + "loss": 1.3859, + "step": 1032 + }, + { + "epoch": 1.0611196712891628, + "grad_norm": 2.6349902153015137, + "learning_rate": 3.6200989722116485e-06, + "loss": 1.4092, + "step": 1033 + }, + { + "epoch": 1.0621468926553672, + "grad_norm": 2.6292974948883057, + "learning_rate": 3.618195660449182e-06, + "loss": 1.419, + "step": 1034 + }, + { + "epoch": 1.0631741140215716, + "grad_norm": 2.4338958263397217, + "learning_rate": 3.6162923486867152e-06, + "loss": 1.3546, + "step": 1035 + }, + { + "epoch": 1.064201335387776, + "grad_norm": 2.376420259475708, + "learning_rate": 3.6143890369242486e-06, + "loss": 1.3899, + "step": 1036 + }, + { + "epoch": 1.0652285567539805, + "grad_norm": 2.428738832473755, + "learning_rate": 3.612485725161782e-06, + "loss": 1.3955, + "step": 1037 + }, + { + "epoch": 1.066255778120185, + "grad_norm": 2.4316651821136475, + "learning_rate": 3.610582413399315e-06, + "loss": 1.3875, + "step": 1038 + }, + { + "epoch": 1.0672829994863893, + "grad_norm": 2.3969266414642334, + "learning_rate": 3.6086791016368482e-06, + "loss": 1.404, + "step": 1039 + }, + { + "epoch": 1.0683102208525936, + "grad_norm": 2.662283420562744, + "learning_rate": 3.6067757898743816e-06, + "loss": 1.4639, + "step": 1040 + }, + { + "epoch": 1.0683102208525936, + "eval_loss": 1.7603929042816162, + "eval_runtime": 38.0098, + "eval_samples_per_second": 26.309, + "eval_steps_per_second": 1.105, + "step": 1040 + }, + { + "epoch": 1.0693374422187982, + "grad_norm": 2.601452112197876, + "learning_rate": 3.604872478111915e-06, + "loss": 1.47, + "step": 1041 + }, + { + "epoch": 1.0703646635850026, + "grad_norm": 2.4221863746643066, + "learning_rate": 3.6029691663494483e-06, + "loss": 1.4162, + "step": 1042 + }, + { + "epoch": 1.071391884951207, + "grad_norm": 2.4996650218963623, + "learning_rate": 3.6010658545869813e-06, + "loss": 1.443, + "step": 1043 + }, + { + "epoch": 1.0724191063174113, + "grad_norm": 2.547239303588867, + "learning_rate": 3.5991625428245146e-06, + "loss": 1.4464, + "step": 1044 + }, + { + "epoch": 1.073446327683616, + "grad_norm": 2.481187343597412, + "learning_rate": 3.597259231062048e-06, + "loss": 1.3749, + "step": 1045 + }, + { + "epoch": 1.0744735490498203, + "grad_norm": 2.5504934787750244, + "learning_rate": 3.5953559192995818e-06, + "loss": 1.4231, + "step": 1046 + }, + { + "epoch": 1.0755007704160247, + "grad_norm": 2.465390205383301, + "learning_rate": 3.593452607537115e-06, + "loss": 1.4805, + "step": 1047 + }, + { + "epoch": 1.076527991782229, + "grad_norm": 2.4399969577789307, + "learning_rate": 3.5915492957746485e-06, + "loss": 1.4091, + "step": 1048 + }, + { + "epoch": 1.0775552131484334, + "grad_norm": 2.4819061756134033, + "learning_rate": 3.589645984012182e-06, + "loss": 1.4367, + "step": 1049 + }, + { + "epoch": 1.078582434514638, + "grad_norm": 2.459517002105713, + "learning_rate": 3.587742672249715e-06, + "loss": 1.3882, + "step": 1050 + }, + { + "epoch": 1.078582434514638, + "eval_loss": 1.7557458877563477, + "eval_runtime": 38.0271, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1050 + }, + { + "epoch": 1.0796096558808423, + "grad_norm": 2.5434813499450684, + "learning_rate": 3.585839360487248e-06, + "loss": 1.3926, + "step": 1051 + }, + { + "epoch": 1.0806368772470467, + "grad_norm": 2.6173205375671387, + "learning_rate": 3.5839360487247815e-06, + "loss": 1.4044, + "step": 1052 + }, + { + "epoch": 1.081664098613251, + "grad_norm": 2.677584171295166, + "learning_rate": 3.582032736962315e-06, + "loss": 1.4695, + "step": 1053 + }, + { + "epoch": 1.0826913199794557, + "grad_norm": 2.4151623249053955, + "learning_rate": 3.5801294251998483e-06, + "loss": 1.3699, + "step": 1054 + }, + { + "epoch": 1.08371854134566, + "grad_norm": 2.4683942794799805, + "learning_rate": 3.578226113437381e-06, + "loss": 1.4132, + "step": 1055 + }, + { + "epoch": 1.0847457627118644, + "grad_norm": 2.6076581478118896, + "learning_rate": 3.5763228016749146e-06, + "loss": 1.4729, + "step": 1056 + }, + { + "epoch": 1.0857729840780688, + "grad_norm": 2.4766876697540283, + "learning_rate": 3.574419489912448e-06, + "loss": 1.4196, + "step": 1057 + }, + { + "epoch": 1.0868002054442731, + "grad_norm": 2.466458559036255, + "learning_rate": 3.5725161781499813e-06, + "loss": 1.4049, + "step": 1058 + }, + { + "epoch": 1.0878274268104777, + "grad_norm": 2.4220409393310547, + "learning_rate": 3.5706128663875147e-06, + "loss": 1.4246, + "step": 1059 + }, + { + "epoch": 1.088854648176682, + "grad_norm": 2.5571770668029785, + "learning_rate": 3.5687095546250476e-06, + "loss": 1.4397, + "step": 1060 + }, + { + "epoch": 1.088854648176682, + "eval_loss": 1.7701127529144287, + "eval_runtime": 38.0754, + "eval_samples_per_second": 26.264, + "eval_steps_per_second": 1.103, + "step": 1060 + }, + { + "epoch": 1.0898818695428865, + "grad_norm": 2.648768186569214, + "learning_rate": 3.566806242862581e-06, + "loss": 1.423, + "step": 1061 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 2.403651714324951, + "learning_rate": 3.5649029311001143e-06, + "loss": 1.3995, + "step": 1062 + }, + { + "epoch": 1.0919363122752954, + "grad_norm": 2.465538263320923, + "learning_rate": 3.5629996193376477e-06, + "loss": 1.4139, + "step": 1063 + }, + { + "epoch": 1.0929635336414998, + "grad_norm": 2.464662551879883, + "learning_rate": 3.561096307575181e-06, + "loss": 1.4499, + "step": 1064 + }, + { + "epoch": 1.0939907550077042, + "grad_norm": 2.4943597316741943, + "learning_rate": 3.559192995812714e-06, + "loss": 1.4555, + "step": 1065 + }, + { + "epoch": 1.0950179763739085, + "grad_norm": 2.6332199573516846, + "learning_rate": 3.557289684050248e-06, + "loss": 1.4484, + "step": 1066 + }, + { + "epoch": 1.0960451977401129, + "grad_norm": 2.4370667934417725, + "learning_rate": 3.555386372287781e-06, + "loss": 1.4588, + "step": 1067 + }, + { + "epoch": 1.0970724191063175, + "grad_norm": 2.568368673324585, + "learning_rate": 3.5534830605253145e-06, + "loss": 1.4269, + "step": 1068 + }, + { + "epoch": 1.0980996404725218, + "grad_norm": 2.4618077278137207, + "learning_rate": 3.551579748762848e-06, + "loss": 1.4165, + "step": 1069 + }, + { + "epoch": 1.0991268618387262, + "grad_norm": 2.551320791244507, + "learning_rate": 3.5496764370003812e-06, + "loss": 1.3991, + "step": 1070 + }, + { + "epoch": 1.0991268618387262, + "eval_loss": 1.7648736238479614, + "eval_runtime": 38.0273, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1070 + }, + { + "epoch": 1.1001540832049306, + "grad_norm": 2.350905418395996, + "learning_rate": 3.5477731252379146e-06, + "loss": 1.3556, + "step": 1071 + }, + { + "epoch": 1.1011813045711352, + "grad_norm": 2.4294588565826416, + "learning_rate": 3.5458698134754475e-06, + "loss": 1.3801, + "step": 1072 + }, + { + "epoch": 1.1022085259373395, + "grad_norm": 2.576774835586548, + "learning_rate": 3.543966501712981e-06, + "loss": 1.3876, + "step": 1073 + }, + { + "epoch": 1.103235747303544, + "grad_norm": 2.716935873031616, + "learning_rate": 3.5420631899505142e-06, + "loss": 1.4907, + "step": 1074 + }, + { + "epoch": 1.1042629686697483, + "grad_norm": 2.425305128097534, + "learning_rate": 3.5401598781880476e-06, + "loss": 1.3161, + "step": 1075 + }, + { + "epoch": 1.1052901900359529, + "grad_norm": 2.54738712310791, + "learning_rate": 3.538256566425581e-06, + "loss": 1.4462, + "step": 1076 + }, + { + "epoch": 1.1063174114021572, + "grad_norm": 2.5109341144561768, + "learning_rate": 3.536353254663114e-06, + "loss": 1.4022, + "step": 1077 + }, + { + "epoch": 1.1073446327683616, + "grad_norm": 2.468749523162842, + "learning_rate": 3.5344499429006473e-06, + "loss": 1.3857, + "step": 1078 + }, + { + "epoch": 1.108371854134566, + "grad_norm": 2.5285134315490723, + "learning_rate": 3.5325466311381806e-06, + "loss": 1.4309, + "step": 1079 + }, + { + "epoch": 1.1093990755007703, + "grad_norm": 2.6391055583953857, + "learning_rate": 3.530643319375714e-06, + "loss": 1.3644, + "step": 1080 + }, + { + "epoch": 1.1093990755007703, + "eval_loss": 1.7846598625183105, + "eval_runtime": 38.0604, + "eval_samples_per_second": 26.274, + "eval_steps_per_second": 1.104, + "step": 1080 + }, + { + "epoch": 1.110426296866975, + "grad_norm": 2.660356283187866, + "learning_rate": 3.5287400076132474e-06, + "loss": 1.3922, + "step": 1081 + }, + { + "epoch": 1.1114535182331793, + "grad_norm": 2.5648741722106934, + "learning_rate": 3.5268366958507803e-06, + "loss": 1.4282, + "step": 1082 + }, + { + "epoch": 1.1124807395993837, + "grad_norm": 2.6862876415252686, + "learning_rate": 3.5249333840883137e-06, + "loss": 1.4208, + "step": 1083 + }, + { + "epoch": 1.113507960965588, + "grad_norm": 2.6200191974639893, + "learning_rate": 3.523030072325847e-06, + "loss": 1.3883, + "step": 1084 + }, + { + "epoch": 1.1145351823317924, + "grad_norm": 2.5802111625671387, + "learning_rate": 3.5211267605633804e-06, + "loss": 1.4033, + "step": 1085 + }, + { + "epoch": 1.115562403697997, + "grad_norm": 2.4148080348968506, + "learning_rate": 3.519223448800914e-06, + "loss": 1.4085, + "step": 1086 + }, + { + "epoch": 1.1165896250642013, + "grad_norm": 2.7452127933502197, + "learning_rate": 3.5173201370384475e-06, + "loss": 1.4021, + "step": 1087 + }, + { + "epoch": 1.1176168464304057, + "grad_norm": 2.6437106132507324, + "learning_rate": 3.515416825275981e-06, + "loss": 1.4161, + "step": 1088 + }, + { + "epoch": 1.11864406779661, + "grad_norm": 2.4709842205047607, + "learning_rate": 3.513513513513514e-06, + "loss": 1.4322, + "step": 1089 + }, + { + "epoch": 1.1196712891628147, + "grad_norm": 2.5138099193573, + "learning_rate": 3.511610201751047e-06, + "loss": 1.3435, + "step": 1090 + }, + { + "epoch": 1.1196712891628147, + "eval_loss": 1.7687023878097534, + "eval_runtime": 38.0137, + "eval_samples_per_second": 26.306, + "eval_steps_per_second": 1.105, + "step": 1090 + }, + { + "epoch": 1.120698510529019, + "grad_norm": 2.4194881916046143, + "learning_rate": 3.5097068899885806e-06, + "loss": 1.4133, + "step": 1091 + }, + { + "epoch": 1.1217257318952234, + "grad_norm": 2.6564371585845947, + "learning_rate": 3.507803578226114e-06, + "loss": 1.4759, + "step": 1092 + }, + { + "epoch": 1.1227529532614278, + "grad_norm": 2.4480180740356445, + "learning_rate": 3.5059002664636473e-06, + "loss": 1.3577, + "step": 1093 + }, + { + "epoch": 1.1237801746276324, + "grad_norm": 2.5538978576660156, + "learning_rate": 3.5039969547011802e-06, + "loss": 1.3984, + "step": 1094 + }, + { + "epoch": 1.1248073959938367, + "grad_norm": 2.6467089653015137, + "learning_rate": 3.5020936429387136e-06, + "loss": 1.4049, + "step": 1095 + }, + { + "epoch": 1.125834617360041, + "grad_norm": 2.6502673625946045, + "learning_rate": 3.500190331176247e-06, + "loss": 1.462, + "step": 1096 + }, + { + "epoch": 1.1268618387262455, + "grad_norm": 2.5696403980255127, + "learning_rate": 3.4982870194137803e-06, + "loss": 1.3926, + "step": 1097 + }, + { + "epoch": 1.1278890600924498, + "grad_norm": 2.5951733589172363, + "learning_rate": 3.4963837076513137e-06, + "loss": 1.4031, + "step": 1098 + }, + { + "epoch": 1.1289162814586544, + "grad_norm": 2.4910218715667725, + "learning_rate": 3.4944803958888466e-06, + "loss": 1.3348, + "step": 1099 + }, + { + "epoch": 1.1299435028248588, + "grad_norm": 2.6325888633728027, + "learning_rate": 3.49257708412638e-06, + "loss": 1.4641, + "step": 1100 + }, + { + "epoch": 1.1299435028248588, + "eval_loss": 1.7720683813095093, + "eval_runtime": 37.9585, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 1100 + }, + { + "epoch": 1.1309707241910631, + "grad_norm": 2.6712775230407715, + "learning_rate": 3.4906737723639133e-06, + "loss": 1.4726, + "step": 1101 + }, + { + "epoch": 1.1319979455572675, + "grad_norm": 2.580404043197632, + "learning_rate": 3.4887704606014467e-06, + "loss": 1.4158, + "step": 1102 + }, + { + "epoch": 1.133025166923472, + "grad_norm": 2.467453718185425, + "learning_rate": 3.48686714883898e-06, + "loss": 1.4636, + "step": 1103 + }, + { + "epoch": 1.1340523882896765, + "grad_norm": 2.627136468887329, + "learning_rate": 3.484963837076513e-06, + "loss": 1.4082, + "step": 1104 + }, + { + "epoch": 1.1350796096558808, + "grad_norm": 2.6198604106903076, + "learning_rate": 3.4830605253140464e-06, + "loss": 1.3628, + "step": 1105 + }, + { + "epoch": 1.1361068310220852, + "grad_norm": 2.439512014389038, + "learning_rate": 3.4811572135515797e-06, + "loss": 1.3885, + "step": 1106 + }, + { + "epoch": 1.1371340523882898, + "grad_norm": 2.5543010234832764, + "learning_rate": 3.4792539017891135e-06, + "loss": 1.3674, + "step": 1107 + }, + { + "epoch": 1.1381612737544942, + "grad_norm": 2.7263364791870117, + "learning_rate": 3.477350590026647e-06, + "loss": 1.4387, + "step": 1108 + }, + { + "epoch": 1.1391884951206985, + "grad_norm": 2.5030503273010254, + "learning_rate": 3.4754472782641802e-06, + "loss": 1.4222, + "step": 1109 + }, + { + "epoch": 1.140215716486903, + "grad_norm": 2.5090320110321045, + "learning_rate": 3.4735439665017136e-06, + "loss": 1.3981, + "step": 1110 + }, + { + "epoch": 1.140215716486903, + "eval_loss": 1.769408941268921, + "eval_runtime": 37.608, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 1110 + }, + { + "epoch": 1.1412429378531073, + "grad_norm": 2.575934648513794, + "learning_rate": 3.4716406547392465e-06, + "loss": 1.4074, + "step": 1111 + }, + { + "epoch": 1.1422701592193119, + "grad_norm": 2.6381359100341797, + "learning_rate": 3.46973734297678e-06, + "loss": 1.4162, + "step": 1112 + }, + { + "epoch": 1.1432973805855162, + "grad_norm": 2.7605674266815186, + "learning_rate": 3.4678340312143133e-06, + "loss": 1.4616, + "step": 1113 + }, + { + "epoch": 1.1443246019517206, + "grad_norm": 2.5548479557037354, + "learning_rate": 3.4659307194518466e-06, + "loss": 1.4511, + "step": 1114 + }, + { + "epoch": 1.145351823317925, + "grad_norm": 2.471168279647827, + "learning_rate": 3.46402740768938e-06, + "loss": 1.3777, + "step": 1115 + }, + { + "epoch": 1.1463790446841293, + "grad_norm": 2.4416346549987793, + "learning_rate": 3.462124095926913e-06, + "loss": 1.3805, + "step": 1116 + }, + { + "epoch": 1.147406266050334, + "grad_norm": 2.4972171783447266, + "learning_rate": 3.4602207841644463e-06, + "loss": 1.4282, + "step": 1117 + }, + { + "epoch": 1.1484334874165383, + "grad_norm": 2.4379162788391113, + "learning_rate": 3.4583174724019797e-06, + "loss": 1.4744, + "step": 1118 + }, + { + "epoch": 1.1494607087827426, + "grad_norm": 2.508047580718994, + "learning_rate": 3.456414160639513e-06, + "loss": 1.4303, + "step": 1119 + }, + { + "epoch": 1.150487930148947, + "grad_norm": 2.4299631118774414, + "learning_rate": 3.4545108488770464e-06, + "loss": 1.4179, + "step": 1120 + }, + { + "epoch": 1.150487930148947, + "eval_loss": 1.7768158912658691, + "eval_runtime": 37.736, + "eval_samples_per_second": 26.5, + "eval_steps_per_second": 1.113, + "step": 1120 + }, + { + "epoch": 1.1515151515151516, + "grad_norm": 2.41538143157959, + "learning_rate": 3.4526075371145793e-06, + "loss": 1.4235, + "step": 1121 + }, + { + "epoch": 1.152542372881356, + "grad_norm": 2.5614655017852783, + "learning_rate": 3.4507042253521127e-06, + "loss": 1.4879, + "step": 1122 + }, + { + "epoch": 1.1535695942475603, + "grad_norm": 2.4826605319976807, + "learning_rate": 3.448800913589646e-06, + "loss": 1.3968, + "step": 1123 + }, + { + "epoch": 1.1545968156137647, + "grad_norm": 2.5341391563415527, + "learning_rate": 3.4468976018271794e-06, + "loss": 1.4903, + "step": 1124 + }, + { + "epoch": 1.1556240369799693, + "grad_norm": 2.493818998336792, + "learning_rate": 3.4449942900647128e-06, + "loss": 1.4466, + "step": 1125 + }, + { + "epoch": 1.1566512583461737, + "grad_norm": 2.6057965755462646, + "learning_rate": 3.443090978302246e-06, + "loss": 1.4425, + "step": 1126 + }, + { + "epoch": 1.157678479712378, + "grad_norm": 2.5481390953063965, + "learning_rate": 3.44118766653978e-06, + "loss": 1.4296, + "step": 1127 + }, + { + "epoch": 1.1587057010785824, + "grad_norm": 2.5753650665283203, + "learning_rate": 3.439284354777313e-06, + "loss": 1.4408, + "step": 1128 + }, + { + "epoch": 1.1597329224447868, + "grad_norm": 2.5423803329467773, + "learning_rate": 3.4373810430148462e-06, + "loss": 1.373, + "step": 1129 + }, + { + "epoch": 1.1607601438109914, + "grad_norm": 2.593487024307251, + "learning_rate": 3.4354777312523796e-06, + "loss": 1.3992, + "step": 1130 + }, + { + "epoch": 1.1607601438109914, + "eval_loss": 1.7784411907196045, + "eval_runtime": 37.8687, + "eval_samples_per_second": 26.407, + "eval_steps_per_second": 1.109, + "step": 1130 + }, + { + "epoch": 1.1617873651771957, + "grad_norm": 2.4693503379821777, + "learning_rate": 3.433574419489913e-06, + "loss": 1.4339, + "step": 1131 + }, + { + "epoch": 1.1628145865434, + "grad_norm": 2.5477423667907715, + "learning_rate": 3.4316711077274463e-06, + "loss": 1.4035, + "step": 1132 + }, + { + "epoch": 1.1638418079096045, + "grad_norm": 2.5650672912597656, + "learning_rate": 3.4297677959649792e-06, + "loss": 1.4517, + "step": 1133 + }, + { + "epoch": 1.1648690292758088, + "grad_norm": 2.3217973709106445, + "learning_rate": 3.4278644842025126e-06, + "loss": 1.3983, + "step": 1134 + }, + { + "epoch": 1.1658962506420134, + "grad_norm": 2.674342155456543, + "learning_rate": 3.425961172440046e-06, + "loss": 1.4799, + "step": 1135 + }, + { + "epoch": 1.1669234720082178, + "grad_norm": 2.4861397743225098, + "learning_rate": 3.4240578606775793e-06, + "loss": 1.414, + "step": 1136 + }, + { + "epoch": 1.1679506933744221, + "grad_norm": 2.4306106567382812, + "learning_rate": 3.4221545489151127e-06, + "loss": 1.3918, + "step": 1137 + }, + { + "epoch": 1.1689779147406265, + "grad_norm": 2.6873955726623535, + "learning_rate": 3.4202512371526456e-06, + "loss": 1.4979, + "step": 1138 + }, + { + "epoch": 1.170005136106831, + "grad_norm": 2.559042453765869, + "learning_rate": 3.418347925390179e-06, + "loss": 1.4424, + "step": 1139 + }, + { + "epoch": 1.1710323574730355, + "grad_norm": 2.397489547729492, + "learning_rate": 3.4164446136277124e-06, + "loss": 1.418, + "step": 1140 + }, + { + "epoch": 1.1710323574730355, + "eval_loss": 1.7702736854553223, + "eval_runtime": 37.8963, + "eval_samples_per_second": 26.388, + "eval_steps_per_second": 1.108, + "step": 1140 + }, + { + "epoch": 1.1720595788392398, + "grad_norm": 2.4295787811279297, + "learning_rate": 3.4145413018652457e-06, + "loss": 1.4436, + "step": 1141 + }, + { + "epoch": 1.1730868002054442, + "grad_norm": 2.503202199935913, + "learning_rate": 3.412637990102779e-06, + "loss": 1.4598, + "step": 1142 + }, + { + "epoch": 1.1741140215716488, + "grad_norm": 2.540195941925049, + "learning_rate": 3.4107346783403125e-06, + "loss": 1.3704, + "step": 1143 + }, + { + "epoch": 1.1751412429378532, + "grad_norm": 2.4832522869110107, + "learning_rate": 3.4088313665778454e-06, + "loss": 1.4081, + "step": 1144 + }, + { + "epoch": 1.1761684643040575, + "grad_norm": 2.5639121532440186, + "learning_rate": 3.4069280548153788e-06, + "loss": 1.3866, + "step": 1145 + }, + { + "epoch": 1.177195685670262, + "grad_norm": 2.526610851287842, + "learning_rate": 3.405024743052912e-06, + "loss": 1.4181, + "step": 1146 + }, + { + "epoch": 1.1782229070364663, + "grad_norm": 2.6945254802703857, + "learning_rate": 3.403121431290446e-06, + "loss": 1.3871, + "step": 1147 + }, + { + "epoch": 1.1792501284026708, + "grad_norm": 2.6561484336853027, + "learning_rate": 3.4012181195279793e-06, + "loss": 1.3987, + "step": 1148 + }, + { + "epoch": 1.1802773497688752, + "grad_norm": 2.6104650497436523, + "learning_rate": 3.3993148077655126e-06, + "loss": 1.4057, + "step": 1149 + }, + { + "epoch": 1.1813045711350796, + "grad_norm": 2.571707248687744, + "learning_rate": 3.3974114960030456e-06, + "loss": 1.3606, + "step": 1150 + }, + { + "epoch": 1.1813045711350796, + "eval_loss": 1.7780840396881104, + "eval_runtime": 37.8616, + "eval_samples_per_second": 26.412, + "eval_steps_per_second": 1.109, + "step": 1150 + }, + { + "epoch": 1.182331792501284, + "grad_norm": 2.5804402828216553, + "learning_rate": 3.395508184240579e-06, + "loss": 1.3876, + "step": 1151 + }, + { + "epoch": 1.1833590138674885, + "grad_norm": 2.832446336746216, + "learning_rate": 3.3936048724781123e-06, + "loss": 1.4674, + "step": 1152 + }, + { + "epoch": 1.184386235233693, + "grad_norm": 2.6478688716888428, + "learning_rate": 3.3917015607156457e-06, + "loss": 1.3643, + "step": 1153 + }, + { + "epoch": 1.1854134565998973, + "grad_norm": 2.579416513442993, + "learning_rate": 3.389798248953179e-06, + "loss": 1.3874, + "step": 1154 + }, + { + "epoch": 1.1864406779661016, + "grad_norm": 2.487861156463623, + "learning_rate": 3.387894937190712e-06, + "loss": 1.4323, + "step": 1155 + }, + { + "epoch": 1.1874678993323062, + "grad_norm": 2.5020716190338135, + "learning_rate": 3.3859916254282453e-06, + "loss": 1.423, + "step": 1156 + }, + { + "epoch": 1.1884951206985106, + "grad_norm": 2.6588852405548096, + "learning_rate": 3.3840883136657787e-06, + "loss": 1.4808, + "step": 1157 + }, + { + "epoch": 1.189522342064715, + "grad_norm": 2.5702178478240967, + "learning_rate": 3.382185001903312e-06, + "loss": 1.4007, + "step": 1158 + }, + { + "epoch": 1.1905495634309193, + "grad_norm": 2.5082061290740967, + "learning_rate": 3.3802816901408454e-06, + "loss": 1.4278, + "step": 1159 + }, + { + "epoch": 1.1915767847971237, + "grad_norm": 2.4165053367614746, + "learning_rate": 3.3783783783783788e-06, + "loss": 1.4088, + "step": 1160 + }, + { + "epoch": 1.1915767847971237, + "eval_loss": 1.7722209692001343, + "eval_runtime": 37.8843, + "eval_samples_per_second": 26.396, + "eval_steps_per_second": 1.109, + "step": 1160 + }, + { + "epoch": 1.1926040061633283, + "grad_norm": 2.4510512351989746, + "learning_rate": 3.3764750666159117e-06, + "loss": 1.4089, + "step": 1161 + }, + { + "epoch": 1.1936312275295327, + "grad_norm": 2.5006556510925293, + "learning_rate": 3.374571754853445e-06, + "loss": 1.3988, + "step": 1162 + }, + { + "epoch": 1.194658448895737, + "grad_norm": 2.416264057159424, + "learning_rate": 3.3726684430909784e-06, + "loss": 1.3576, + "step": 1163 + }, + { + "epoch": 1.1956856702619414, + "grad_norm": 2.465613842010498, + "learning_rate": 3.370765131328512e-06, + "loss": 1.3796, + "step": 1164 + }, + { + "epoch": 1.1967128916281458, + "grad_norm": 2.510542154312134, + "learning_rate": 3.368861819566045e-06, + "loss": 1.391, + "step": 1165 + }, + { + "epoch": 1.1977401129943503, + "grad_norm": 2.729495048522949, + "learning_rate": 3.366958507803578e-06, + "loss": 1.431, + "step": 1166 + }, + { + "epoch": 1.1987673343605547, + "grad_norm": 2.4473495483398438, + "learning_rate": 3.365055196041112e-06, + "loss": 1.376, + "step": 1167 + }, + { + "epoch": 1.199794555726759, + "grad_norm": 2.3869211673736572, + "learning_rate": 3.3631518842786452e-06, + "loss": 1.4498, + "step": 1168 + }, + { + "epoch": 1.2008217770929634, + "grad_norm": 2.4202256202697754, + "learning_rate": 3.3612485725161786e-06, + "loss": 1.3754, + "step": 1169 + }, + { + "epoch": 1.201848998459168, + "grad_norm": 2.452707529067993, + "learning_rate": 3.359345260753712e-06, + "loss": 1.4647, + "step": 1170 + }, + { + "epoch": 1.201848998459168, + "eval_loss": 1.769252896308899, + "eval_runtime": 37.9385, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1170 + }, + { + "epoch": 1.2028762198253724, + "grad_norm": 2.56616473197937, + "learning_rate": 3.3574419489912453e-06, + "loss": 1.443, + "step": 1171 + }, + { + "epoch": 1.2039034411915768, + "grad_norm": 2.6483540534973145, + "learning_rate": 3.3555386372287783e-06, + "loss": 1.4755, + "step": 1172 + }, + { + "epoch": 1.2049306625577811, + "grad_norm": 2.5579090118408203, + "learning_rate": 3.3536353254663116e-06, + "loss": 1.3732, + "step": 1173 + }, + { + "epoch": 1.2059578839239857, + "grad_norm": 2.5310683250427246, + "learning_rate": 3.351732013703845e-06, + "loss": 1.4077, + "step": 1174 + }, + { + "epoch": 1.20698510529019, + "grad_norm": 2.616957187652588, + "learning_rate": 3.3498287019413784e-06, + "loss": 1.4237, + "step": 1175 + }, + { + "epoch": 1.2080123266563945, + "grad_norm": 2.6008999347686768, + "learning_rate": 3.3479253901789117e-06, + "loss": 1.4381, + "step": 1176 + }, + { + "epoch": 1.2090395480225988, + "grad_norm": 2.5235824584960938, + "learning_rate": 3.346022078416445e-06, + "loss": 1.3471, + "step": 1177 + }, + { + "epoch": 1.2100667693888032, + "grad_norm": 2.3749704360961914, + "learning_rate": 3.344118766653978e-06, + "loss": 1.417, + "step": 1178 + }, + { + "epoch": 1.2110939907550078, + "grad_norm": 2.493380069732666, + "learning_rate": 3.3422154548915114e-06, + "loss": 1.3924, + "step": 1179 + }, + { + "epoch": 1.2121212121212122, + "grad_norm": 2.4233086109161377, + "learning_rate": 3.3403121431290448e-06, + "loss": 1.4257, + "step": 1180 + }, + { + "epoch": 1.2121212121212122, + "eval_loss": 1.7660146951675415, + "eval_runtime": 37.9268, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 1180 + }, + { + "epoch": 1.2131484334874165, + "grad_norm": 2.535982608795166, + "learning_rate": 3.338408831366578e-06, + "loss": 1.3587, + "step": 1181 + }, + { + "epoch": 1.214175654853621, + "grad_norm": 2.5744845867156982, + "learning_rate": 3.3365055196041115e-06, + "loss": 1.436, + "step": 1182 + }, + { + "epoch": 1.2152028762198253, + "grad_norm": 2.5544042587280273, + "learning_rate": 3.3346022078416444e-06, + "loss": 1.464, + "step": 1183 + }, + { + "epoch": 1.2162300975860298, + "grad_norm": 2.4980483055114746, + "learning_rate": 3.3326988960791778e-06, + "loss": 1.3915, + "step": 1184 + }, + { + "epoch": 1.2172573189522342, + "grad_norm": 2.403468608856201, + "learning_rate": 3.330795584316711e-06, + "loss": 1.3239, + "step": 1185 + }, + { + "epoch": 1.2182845403184386, + "grad_norm": 2.6861846446990967, + "learning_rate": 3.3288922725542445e-06, + "loss": 1.4013, + "step": 1186 + }, + { + "epoch": 1.219311761684643, + "grad_norm": 2.525033712387085, + "learning_rate": 3.326988960791778e-06, + "loss": 1.4135, + "step": 1187 + }, + { + "epoch": 1.2203389830508475, + "grad_norm": 2.6705894470214844, + "learning_rate": 3.3250856490293117e-06, + "loss": 1.4587, + "step": 1188 + }, + { + "epoch": 1.221366204417052, + "grad_norm": 2.6108217239379883, + "learning_rate": 3.323182337266845e-06, + "loss": 1.4482, + "step": 1189 + }, + { + "epoch": 1.2223934257832563, + "grad_norm": 2.5227413177490234, + "learning_rate": 3.321279025504378e-06, + "loss": 1.432, + "step": 1190 + }, + { + "epoch": 1.2223934257832563, + "eval_loss": 1.756821870803833, + "eval_runtime": 37.9929, + "eval_samples_per_second": 26.321, + "eval_steps_per_second": 1.105, + "step": 1190 + }, + { + "epoch": 1.2234206471494606, + "grad_norm": 2.469766139984131, + "learning_rate": 3.3193757137419113e-06, + "loss": 1.4215, + "step": 1191 + }, + { + "epoch": 1.2244478685156652, + "grad_norm": 2.5257861614227295, + "learning_rate": 3.3174724019794447e-06, + "loss": 1.4624, + "step": 1192 + }, + { + "epoch": 1.2254750898818696, + "grad_norm": 2.5281760692596436, + "learning_rate": 3.315569090216978e-06, + "loss": 1.3799, + "step": 1193 + }, + { + "epoch": 1.226502311248074, + "grad_norm": 2.701521396636963, + "learning_rate": 3.3136657784545114e-06, + "loss": 1.4631, + "step": 1194 + }, + { + "epoch": 1.2275295326142783, + "grad_norm": 2.4239444732666016, + "learning_rate": 3.3117624666920443e-06, + "loss": 1.3746, + "step": 1195 + }, + { + "epoch": 1.2285567539804827, + "grad_norm": 2.430377721786499, + "learning_rate": 3.3098591549295777e-06, + "loss": 1.3584, + "step": 1196 + }, + { + "epoch": 1.2295839753466873, + "grad_norm": 2.7177658081054688, + "learning_rate": 3.307955843167111e-06, + "loss": 1.4301, + "step": 1197 + }, + { + "epoch": 1.2306111967128917, + "grad_norm": 2.512943983078003, + "learning_rate": 3.3060525314046444e-06, + "loss": 1.4111, + "step": 1198 + }, + { + "epoch": 1.231638418079096, + "grad_norm": 2.5269784927368164, + "learning_rate": 3.304149219642178e-06, + "loss": 1.4096, + "step": 1199 + }, + { + "epoch": 1.2326656394453004, + "grad_norm": 2.476409673690796, + "learning_rate": 3.3022459078797107e-06, + "loss": 1.4322, + "step": 1200 + }, + { + "epoch": 1.2326656394453004, + "eval_loss": 1.7645230293273926, + "eval_runtime": 37.9898, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1200 + }, + { + "epoch": 1.2336928608115048, + "grad_norm": 2.5339248180389404, + "learning_rate": 3.300342596117244e-06, + "loss": 1.4093, + "step": 1201 + }, + { + "epoch": 1.2347200821777093, + "grad_norm": 2.3505656719207764, + "learning_rate": 3.2984392843547775e-06, + "loss": 1.3623, + "step": 1202 + }, + { + "epoch": 1.2357473035439137, + "grad_norm": 2.400259256362915, + "learning_rate": 3.296535972592311e-06, + "loss": 1.3512, + "step": 1203 + }, + { + "epoch": 1.236774524910118, + "grad_norm": 2.561558723449707, + "learning_rate": 3.294632660829844e-06, + "loss": 1.3946, + "step": 1204 + }, + { + "epoch": 1.2378017462763227, + "grad_norm": 2.623100519180298, + "learning_rate": 3.292729349067377e-06, + "loss": 1.3972, + "step": 1205 + }, + { + "epoch": 1.238828967642527, + "grad_norm": 2.4889469146728516, + "learning_rate": 3.2908260373049105e-06, + "loss": 1.381, + "step": 1206 + }, + { + "epoch": 1.2398561890087314, + "grad_norm": 2.457334518432617, + "learning_rate": 3.288922725542444e-06, + "loss": 1.4345, + "step": 1207 + }, + { + "epoch": 1.2408834103749358, + "grad_norm": 2.440988540649414, + "learning_rate": 3.2870194137799776e-06, + "loss": 1.3901, + "step": 1208 + }, + { + "epoch": 1.2419106317411401, + "grad_norm": 2.4992308616638184, + "learning_rate": 3.285116102017511e-06, + "loss": 1.3694, + "step": 1209 + }, + { + "epoch": 1.2429378531073447, + "grad_norm": 2.611128091812134, + "learning_rate": 3.2832127902550444e-06, + "loss": 1.4344, + "step": 1210 + }, + { + "epoch": 1.2429378531073447, + "eval_loss": 1.763407826423645, + "eval_runtime": 37.7142, + "eval_samples_per_second": 26.515, + "eval_steps_per_second": 1.114, + "step": 1210 + }, + { + "epoch": 1.243965074473549, + "grad_norm": 2.6573522090911865, + "learning_rate": 3.2813094784925777e-06, + "loss": 1.4204, + "step": 1211 + }, + { + "epoch": 1.2449922958397535, + "grad_norm": 2.4785778522491455, + "learning_rate": 3.2794061667301107e-06, + "loss": 1.3617, + "step": 1212 + }, + { + "epoch": 1.2460195172059578, + "grad_norm": 2.4831974506378174, + "learning_rate": 3.277502854967644e-06, + "loss": 1.4004, + "step": 1213 + }, + { + "epoch": 1.2470467385721622, + "grad_norm": 2.6142327785491943, + "learning_rate": 3.2755995432051774e-06, + "loss": 1.3881, + "step": 1214 + }, + { + "epoch": 1.2480739599383668, + "grad_norm": 2.447962999343872, + "learning_rate": 3.2736962314427107e-06, + "loss": 1.3984, + "step": 1215 + }, + { + "epoch": 1.2491011813045712, + "grad_norm": 2.3895885944366455, + "learning_rate": 3.271792919680244e-06, + "loss": 1.4099, + "step": 1216 + }, + { + "epoch": 1.2501284026707755, + "grad_norm": 2.599364995956421, + "learning_rate": 3.269889607917777e-06, + "loss": 1.4575, + "step": 1217 + }, + { + "epoch": 1.25115562403698, + "grad_norm": 2.6294758319854736, + "learning_rate": 3.2679862961553104e-06, + "loss": 1.3752, + "step": 1218 + }, + { + "epoch": 1.2521828454031843, + "grad_norm": 2.50923228263855, + "learning_rate": 3.2660829843928438e-06, + "loss": 1.4017, + "step": 1219 + }, + { + "epoch": 1.2532100667693888, + "grad_norm": 2.626434564590454, + "learning_rate": 3.264179672630377e-06, + "loss": 1.4505, + "step": 1220 + }, + { + "epoch": 1.2532100667693888, + "eval_loss": 1.7672324180603027, + "eval_runtime": 37.8605, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 1220 + }, + { + "epoch": 1.2542372881355932, + "grad_norm": 2.541663885116577, + "learning_rate": 3.2622763608679105e-06, + "loss": 1.4129, + "step": 1221 + }, + { + "epoch": 1.2552645095017976, + "grad_norm": 2.559877395629883, + "learning_rate": 3.2603730491054434e-06, + "loss": 1.3793, + "step": 1222 + }, + { + "epoch": 1.2562917308680022, + "grad_norm": 2.4085795879364014, + "learning_rate": 3.258469737342977e-06, + "loss": 1.3895, + "step": 1223 + }, + { + "epoch": 1.2573189522342065, + "grad_norm": 2.2942678928375244, + "learning_rate": 3.25656642558051e-06, + "loss": 1.3713, + "step": 1224 + }, + { + "epoch": 1.258346173600411, + "grad_norm": 2.561086893081665, + "learning_rate": 3.2546631138180435e-06, + "loss": 1.3835, + "step": 1225 + }, + { + "epoch": 1.2593733949666153, + "grad_norm": 2.4947667121887207, + "learning_rate": 3.252759802055577e-06, + "loss": 1.3793, + "step": 1226 + }, + { + "epoch": 1.2604006163328196, + "grad_norm": 2.6293253898620605, + "learning_rate": 3.25085649029311e-06, + "loss": 1.3935, + "step": 1227 + }, + { + "epoch": 1.2614278376990242, + "grad_norm": 2.5461983680725098, + "learning_rate": 3.248953178530644e-06, + "loss": 1.4201, + "step": 1228 + }, + { + "epoch": 1.2624550590652286, + "grad_norm": 2.381537675857544, + "learning_rate": 3.247049866768177e-06, + "loss": 1.3932, + "step": 1229 + }, + { + "epoch": 1.263482280431433, + "grad_norm": 2.381558418273926, + "learning_rate": 3.2451465550057103e-06, + "loss": 1.3425, + "step": 1230 + }, + { + "epoch": 1.263482280431433, + "eval_loss": 1.7713288068771362, + "eval_runtime": 37.9629, + "eval_samples_per_second": 26.342, + "eval_steps_per_second": 1.106, + "step": 1230 + }, + { + "epoch": 1.2645095017976373, + "grad_norm": 2.5877468585968018, + "learning_rate": 3.2432432432432437e-06, + "loss": 1.3917, + "step": 1231 + }, + { + "epoch": 1.2655367231638417, + "grad_norm": 2.6057848930358887, + "learning_rate": 3.241339931480777e-06, + "loss": 1.4217, + "step": 1232 + }, + { + "epoch": 1.2665639445300463, + "grad_norm": 2.5157413482666016, + "learning_rate": 3.2394366197183104e-06, + "loss": 1.427, + "step": 1233 + }, + { + "epoch": 1.2675911658962506, + "grad_norm": 2.4153692722320557, + "learning_rate": 3.2375333079558434e-06, + "loss": 1.3822, + "step": 1234 + }, + { + "epoch": 1.268618387262455, + "grad_norm": 2.5459582805633545, + "learning_rate": 3.2356299961933767e-06, + "loss": 1.4152, + "step": 1235 + }, + { + "epoch": 1.2696456086286596, + "grad_norm": 2.606468915939331, + "learning_rate": 3.23372668443091e-06, + "loss": 1.4278, + "step": 1236 + }, + { + "epoch": 1.270672829994864, + "grad_norm": 2.5061633586883545, + "learning_rate": 3.2318233726684435e-06, + "loss": 1.4512, + "step": 1237 + }, + { + "epoch": 1.2717000513610683, + "grad_norm": 2.3693087100982666, + "learning_rate": 3.229920060905977e-06, + "loss": 1.3791, + "step": 1238 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 2.479583978652954, + "learning_rate": 3.2280167491435098e-06, + "loss": 1.3814, + "step": 1239 + }, + { + "epoch": 1.273754494093477, + "grad_norm": 2.3258883953094482, + "learning_rate": 3.226113437381043e-06, + "loss": 1.3904, + "step": 1240 + }, + { + "epoch": 1.273754494093477, + "eval_loss": 1.7602728605270386, + "eval_runtime": 37.9604, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 1240 + }, + { + "epoch": 1.2747817154596817, + "grad_norm": 2.4170308113098145, + "learning_rate": 3.2242101256185765e-06, + "loss": 1.4045, + "step": 1241 + }, + { + "epoch": 1.275808936825886, + "grad_norm": 2.578066825866699, + "learning_rate": 3.22230681385611e-06, + "loss": 1.4556, + "step": 1242 + }, + { + "epoch": 1.2768361581920904, + "grad_norm": 2.558187246322632, + "learning_rate": 3.220403502093643e-06, + "loss": 1.4036, + "step": 1243 + }, + { + "epoch": 1.2778633795582948, + "grad_norm": 2.598916530609131, + "learning_rate": 3.218500190331176e-06, + "loss": 1.4462, + "step": 1244 + }, + { + "epoch": 1.2788906009244991, + "grad_norm": 2.4528257846832275, + "learning_rate": 3.2165968785687095e-06, + "loss": 1.3855, + "step": 1245 + }, + { + "epoch": 1.2799178222907037, + "grad_norm": 2.5937459468841553, + "learning_rate": 3.214693566806243e-06, + "loss": 1.4191, + "step": 1246 + }, + { + "epoch": 1.280945043656908, + "grad_norm": 2.5324676036834717, + "learning_rate": 3.2127902550437762e-06, + "loss": 1.4162, + "step": 1247 + }, + { + "epoch": 1.2819722650231125, + "grad_norm": 2.4849791526794434, + "learning_rate": 3.21088694328131e-06, + "loss": 1.4037, + "step": 1248 + }, + { + "epoch": 1.2829994863893168, + "grad_norm": 2.389582633972168, + "learning_rate": 3.2089836315188434e-06, + "loss": 1.3966, + "step": 1249 + }, + { + "epoch": 1.2840267077555212, + "grad_norm": 2.387909173965454, + "learning_rate": 3.2070803197563767e-06, + "loss": 1.396, + "step": 1250 + }, + { + "epoch": 1.2840267077555212, + "eval_loss": 1.7642556428909302, + "eval_runtime": 37.9446, + "eval_samples_per_second": 26.354, + "eval_steps_per_second": 1.107, + "step": 1250 + }, + { + "epoch": 1.2850539291217258, + "grad_norm": 2.5014612674713135, + "learning_rate": 3.2051770079939097e-06, + "loss": 1.4128, + "step": 1251 + }, + { + "epoch": 1.2860811504879301, + "grad_norm": 2.497950553894043, + "learning_rate": 3.203273696231443e-06, + "loss": 1.4412, + "step": 1252 + }, + { + "epoch": 1.2871083718541345, + "grad_norm": 2.4371297359466553, + "learning_rate": 3.2013703844689764e-06, + "loss": 1.3754, + "step": 1253 + }, + { + "epoch": 1.288135593220339, + "grad_norm": 2.4336745738983154, + "learning_rate": 3.1994670727065098e-06, + "loss": 1.4258, + "step": 1254 + }, + { + "epoch": 1.2891628145865435, + "grad_norm": 2.5165536403656006, + "learning_rate": 3.197563760944043e-06, + "loss": 1.4414, + "step": 1255 + }, + { + "epoch": 1.2901900359527478, + "grad_norm": 2.5060458183288574, + "learning_rate": 3.195660449181576e-06, + "loss": 1.4336, + "step": 1256 + }, + { + "epoch": 1.2912172573189522, + "grad_norm": 2.584284543991089, + "learning_rate": 3.1937571374191094e-06, + "loss": 1.3768, + "step": 1257 + }, + { + "epoch": 1.2922444786851566, + "grad_norm": 2.615629196166992, + "learning_rate": 3.191853825656643e-06, + "loss": 1.401, + "step": 1258 + }, + { + "epoch": 1.2932717000513612, + "grad_norm": 2.5989596843719482, + "learning_rate": 3.189950513894176e-06, + "loss": 1.3915, + "step": 1259 + }, + { + "epoch": 1.2942989214175655, + "grad_norm": 2.5507843494415283, + "learning_rate": 3.1880472021317095e-06, + "loss": 1.3979, + "step": 1260 + }, + { + "epoch": 1.2942989214175655, + "eval_loss": 1.7715612649917603, + "eval_runtime": 37.9859, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1260 + }, + { + "epoch": 1.29532614278377, + "grad_norm": 2.512140989303589, + "learning_rate": 3.1861438903692425e-06, + "loss": 1.3523, + "step": 1261 + }, + { + "epoch": 1.2963533641499743, + "grad_norm": 2.4311885833740234, + "learning_rate": 3.184240578606776e-06, + "loss": 1.3828, + "step": 1262 + }, + { + "epoch": 1.2973805855161786, + "grad_norm": 2.439265489578247, + "learning_rate": 3.182337266844309e-06, + "loss": 1.402, + "step": 1263 + }, + { + "epoch": 1.2984078068823832, + "grad_norm": 2.5773651599884033, + "learning_rate": 3.1804339550818426e-06, + "loss": 1.4362, + "step": 1264 + }, + { + "epoch": 1.2994350282485876, + "grad_norm": 2.456021547317505, + "learning_rate": 3.178530643319376e-06, + "loss": 1.3918, + "step": 1265 + }, + { + "epoch": 1.300462249614792, + "grad_norm": 2.44404673576355, + "learning_rate": 3.176627331556909e-06, + "loss": 1.3662, + "step": 1266 + }, + { + "epoch": 1.3014894709809965, + "grad_norm": 2.778027057647705, + "learning_rate": 3.1747240197944422e-06, + "loss": 1.4112, + "step": 1267 + }, + { + "epoch": 1.3025166923472007, + "grad_norm": 2.476496934890747, + "learning_rate": 3.172820708031976e-06, + "loss": 1.3654, + "step": 1268 + }, + { + "epoch": 1.3035439137134053, + "grad_norm": 2.6415998935699463, + "learning_rate": 3.1709173962695094e-06, + "loss": 1.4441, + "step": 1269 + }, + { + "epoch": 1.3045711350796096, + "grad_norm": 2.5046234130859375, + "learning_rate": 3.1690140845070427e-06, + "loss": 1.4316, + "step": 1270 + }, + { + "epoch": 1.3045711350796096, + "eval_loss": 1.7736135721206665, + "eval_runtime": 37.9392, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1270 + }, + { + "epoch": 1.305598356445814, + "grad_norm": 2.6234474182128906, + "learning_rate": 3.167110772744576e-06, + "loss": 1.4, + "step": 1271 + }, + { + "epoch": 1.3066255778120186, + "grad_norm": 2.488297700881958, + "learning_rate": 3.1652074609821095e-06, + "loss": 1.3908, + "step": 1272 + }, + { + "epoch": 1.307652799178223, + "grad_norm": 2.3759512901306152, + "learning_rate": 3.1633041492196424e-06, + "loss": 1.3908, + "step": 1273 + }, + { + "epoch": 1.3086800205444273, + "grad_norm": 2.487281084060669, + "learning_rate": 3.1614008374571758e-06, + "loss": 1.4217, + "step": 1274 + }, + { + "epoch": 1.3097072419106317, + "grad_norm": 2.5366780757904053, + "learning_rate": 3.159497525694709e-06, + "loss": 1.4229, + "step": 1275 + }, + { + "epoch": 1.310734463276836, + "grad_norm": 2.416141986846924, + "learning_rate": 3.1575942139322425e-06, + "loss": 1.3865, + "step": 1276 + }, + { + "epoch": 1.3117616846430407, + "grad_norm": 2.537964344024658, + "learning_rate": 3.155690902169776e-06, + "loss": 1.4141, + "step": 1277 + }, + { + "epoch": 1.312788906009245, + "grad_norm": 2.481692314147949, + "learning_rate": 3.1537875904073088e-06, + "loss": 1.383, + "step": 1278 + }, + { + "epoch": 1.3138161273754494, + "grad_norm": 2.4898593425750732, + "learning_rate": 3.151884278644842e-06, + "loss": 1.4078, + "step": 1279 + }, + { + "epoch": 1.3148433487416538, + "grad_norm": 2.4692323207855225, + "learning_rate": 3.1499809668823755e-06, + "loss": 1.3818, + "step": 1280 + }, + { + "epoch": 1.3148433487416538, + "eval_loss": 1.7660051584243774, + "eval_runtime": 37.8922, + "eval_samples_per_second": 26.391, + "eval_steps_per_second": 1.108, + "step": 1280 + }, + { + "epoch": 1.3158705701078581, + "grad_norm": 2.475862503051758, + "learning_rate": 3.148077655119909e-06, + "loss": 1.4384, + "step": 1281 + }, + { + "epoch": 1.3168977914740627, + "grad_norm": 2.2839648723602295, + "learning_rate": 3.1461743433574422e-06, + "loss": 1.3722, + "step": 1282 + }, + { + "epoch": 1.317925012840267, + "grad_norm": 2.3965585231781006, + "learning_rate": 3.144271031594975e-06, + "loss": 1.4262, + "step": 1283 + }, + { + "epoch": 1.3189522342064715, + "grad_norm": 2.4107236862182617, + "learning_rate": 3.1423677198325085e-06, + "loss": 1.4216, + "step": 1284 + }, + { + "epoch": 1.319979455572676, + "grad_norm": 2.438899040222168, + "learning_rate": 3.140464408070042e-06, + "loss": 1.383, + "step": 1285 + }, + { + "epoch": 1.3210066769388804, + "grad_norm": 2.4146060943603516, + "learning_rate": 3.1385610963075753e-06, + "loss": 1.3827, + "step": 1286 + }, + { + "epoch": 1.3220338983050848, + "grad_norm": 2.4193496704101562, + "learning_rate": 3.1366577845451086e-06, + "loss": 1.3707, + "step": 1287 + }, + { + "epoch": 1.3230611196712891, + "grad_norm": 2.5284323692321777, + "learning_rate": 3.134754472782642e-06, + "loss": 1.3623, + "step": 1288 + }, + { + "epoch": 1.3240883410374935, + "grad_norm": 2.421509265899658, + "learning_rate": 3.1328511610201758e-06, + "loss": 1.35, + "step": 1289 + }, + { + "epoch": 1.325115562403698, + "grad_norm": 2.3964431285858154, + "learning_rate": 3.1309478492577087e-06, + "loss": 1.3486, + "step": 1290 + }, + { + "epoch": 1.325115562403698, + "eval_loss": 1.7699168920516968, + "eval_runtime": 37.9003, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 1290 + }, + { + "epoch": 1.3261427837699025, + "grad_norm": 2.4371817111968994, + "learning_rate": 3.129044537495242e-06, + "loss": 1.3958, + "step": 1291 + }, + { + "epoch": 1.3271700051361068, + "grad_norm": 2.45233154296875, + "learning_rate": 3.1271412257327754e-06, + "loss": 1.4175, + "step": 1292 + }, + { + "epoch": 1.3281972265023112, + "grad_norm": 2.314384937286377, + "learning_rate": 3.125237913970309e-06, + "loss": 1.3701, + "step": 1293 + }, + { + "epoch": 1.3292244478685156, + "grad_norm": 2.5603694915771484, + "learning_rate": 3.123334602207842e-06, + "loss": 1.3489, + "step": 1294 + }, + { + "epoch": 1.3302516692347202, + "grad_norm": 2.5919623374938965, + "learning_rate": 3.121431290445375e-06, + "loss": 1.4472, + "step": 1295 + }, + { + "epoch": 1.3312788906009245, + "grad_norm": 2.6912758350372314, + "learning_rate": 3.1195279786829085e-06, + "loss": 1.4917, + "step": 1296 + }, + { + "epoch": 1.332306111967129, + "grad_norm": 2.439089298248291, + "learning_rate": 3.117624666920442e-06, + "loss": 1.3776, + "step": 1297 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 2.5207953453063965, + "learning_rate": 3.115721355157975e-06, + "loss": 1.3784, + "step": 1298 + }, + { + "epoch": 1.3343605546995376, + "grad_norm": 2.507791519165039, + "learning_rate": 3.1138180433955085e-06, + "loss": 1.3346, + "step": 1299 + }, + { + "epoch": 1.3353877760657422, + "grad_norm": 2.450531482696533, + "learning_rate": 3.1119147316330415e-06, + "loss": 1.4151, + "step": 1300 + }, + { + "epoch": 1.3353877760657422, + "eval_loss": 1.7761608362197876, + "eval_runtime": 37.9041, + "eval_samples_per_second": 26.382, + "eval_steps_per_second": 1.108, + "step": 1300 + }, + { + "epoch": 1.3364149974319466, + "grad_norm": 2.488255262374878, + "learning_rate": 3.110011419870575e-06, + "loss": 1.4475, + "step": 1301 + }, + { + "epoch": 1.337442218798151, + "grad_norm": 2.5059807300567627, + "learning_rate": 3.1081081081081082e-06, + "loss": 1.3812, + "step": 1302 + }, + { + "epoch": 1.3384694401643555, + "grad_norm": 2.5579488277435303, + "learning_rate": 3.1062047963456416e-06, + "loss": 1.4134, + "step": 1303 + }, + { + "epoch": 1.33949666153056, + "grad_norm": 2.6170148849487305, + "learning_rate": 3.104301484583175e-06, + "loss": 1.459, + "step": 1304 + }, + { + "epoch": 1.3405238828967643, + "grad_norm": 2.5572125911712646, + "learning_rate": 3.1023981728207083e-06, + "loss": 1.4354, + "step": 1305 + }, + { + "epoch": 1.3415511042629686, + "grad_norm": 2.396883487701416, + "learning_rate": 3.1004948610582412e-06, + "loss": 1.4016, + "step": 1306 + }, + { + "epoch": 1.342578325629173, + "grad_norm": 2.5869390964508057, + "learning_rate": 3.0985915492957746e-06, + "loss": 1.3945, + "step": 1307 + }, + { + "epoch": 1.3436055469953776, + "grad_norm": 2.4271085262298584, + "learning_rate": 3.096688237533308e-06, + "loss": 1.3506, + "step": 1308 + }, + { + "epoch": 1.344632768361582, + "grad_norm": 2.398407220840454, + "learning_rate": 3.0947849257708418e-06, + "loss": 1.4475, + "step": 1309 + }, + { + "epoch": 1.3456599897277863, + "grad_norm": 2.441934823989868, + "learning_rate": 3.092881614008375e-06, + "loss": 1.3972, + "step": 1310 + }, + { + "epoch": 1.3456599897277863, + "eval_loss": 1.774910569190979, + "eval_runtime": 37.6105, + "eval_samples_per_second": 26.588, + "eval_steps_per_second": 1.117, + "step": 1310 + }, + { + "epoch": 1.3466872110939907, + "grad_norm": 2.5015370845794678, + "learning_rate": 3.0909783022459085e-06, + "loss": 1.4418, + "step": 1311 + }, + { + "epoch": 1.347714432460195, + "grad_norm": 2.4616997241973877, + "learning_rate": 3.0890749904834414e-06, + "loss": 1.3833, + "step": 1312 + }, + { + "epoch": 1.3487416538263997, + "grad_norm": 2.5993902683258057, + "learning_rate": 3.0871716787209748e-06, + "loss": 1.4385, + "step": 1313 + }, + { + "epoch": 1.349768875192604, + "grad_norm": 2.518195867538452, + "learning_rate": 3.085268366958508e-06, + "loss": 1.4127, + "step": 1314 + }, + { + "epoch": 1.3507960965588084, + "grad_norm": 2.545189380645752, + "learning_rate": 3.0833650551960415e-06, + "loss": 1.42, + "step": 1315 + }, + { + "epoch": 1.3518233179250128, + "grad_norm": 2.461818218231201, + "learning_rate": 3.081461743433575e-06, + "loss": 1.3796, + "step": 1316 + }, + { + "epoch": 1.3528505392912171, + "grad_norm": 2.4081408977508545, + "learning_rate": 3.079558431671108e-06, + "loss": 1.3939, + "step": 1317 + }, + { + "epoch": 1.3538777606574217, + "grad_norm": 2.2885539531707764, + "learning_rate": 3.077655119908641e-06, + "loss": 1.3645, + "step": 1318 + }, + { + "epoch": 1.354904982023626, + "grad_norm": 2.453791618347168, + "learning_rate": 3.0757518081461745e-06, + "loss": 1.3906, + "step": 1319 + }, + { + "epoch": 1.3559322033898304, + "grad_norm": 2.4538352489471436, + "learning_rate": 3.073848496383708e-06, + "loss": 1.4127, + "step": 1320 + }, + { + "epoch": 1.3559322033898304, + "eval_loss": 1.7821755409240723, + "eval_runtime": 37.7585, + "eval_samples_per_second": 26.484, + "eval_steps_per_second": 1.112, + "step": 1320 + }, + { + "epoch": 1.356959424756035, + "grad_norm": 2.438425064086914, + "learning_rate": 3.0719451846212413e-06, + "loss": 1.3763, + "step": 1321 + }, + { + "epoch": 1.3579866461222394, + "grad_norm": 2.4398279190063477, + "learning_rate": 3.0700418728587746e-06, + "loss": 1.3963, + "step": 1322 + }, + { + "epoch": 1.3590138674884438, + "grad_norm": 2.6763861179351807, + "learning_rate": 3.0681385610963076e-06, + "loss": 1.3837, + "step": 1323 + }, + { + "epoch": 1.3600410888546481, + "grad_norm": 2.685056209564209, + "learning_rate": 3.066235249333841e-06, + "loss": 1.4158, + "step": 1324 + }, + { + "epoch": 1.3610683102208525, + "grad_norm": 2.6663355827331543, + "learning_rate": 3.0643319375713743e-06, + "loss": 1.4533, + "step": 1325 + }, + { + "epoch": 1.362095531587057, + "grad_norm": 2.4613428115844727, + "learning_rate": 3.0624286258089076e-06, + "loss": 1.3927, + "step": 1326 + }, + { + "epoch": 1.3631227529532615, + "grad_norm": 2.348334789276123, + "learning_rate": 3.060525314046441e-06, + "loss": 1.3782, + "step": 1327 + }, + { + "epoch": 1.3641499743194658, + "grad_norm": 2.5011332035064697, + "learning_rate": 3.058622002283974e-06, + "loss": 1.4175, + "step": 1328 + }, + { + "epoch": 1.3651771956856702, + "grad_norm": 2.4692256450653076, + "learning_rate": 3.0567186905215077e-06, + "loss": 1.4319, + "step": 1329 + }, + { + "epoch": 1.3662044170518746, + "grad_norm": 2.616990089416504, + "learning_rate": 3.054815378759041e-06, + "loss": 1.4264, + "step": 1330 + }, + { + "epoch": 1.3662044170518746, + "eval_loss": 1.781583309173584, + "eval_runtime": 37.9291, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 1330 + }, + { + "epoch": 1.3672316384180792, + "grad_norm": 2.5416364669799805, + "learning_rate": 3.0529120669965745e-06, + "loss": 1.4249, + "step": 1331 + }, + { + "epoch": 1.3682588597842835, + "grad_norm": 2.5346689224243164, + "learning_rate": 3.051008755234108e-06, + "loss": 1.3647, + "step": 1332 + }, + { + "epoch": 1.3692860811504879, + "grad_norm": 2.585714340209961, + "learning_rate": 3.049105443471641e-06, + "loss": 1.3714, + "step": 1333 + }, + { + "epoch": 1.3703133025166925, + "grad_norm": 2.5958940982818604, + "learning_rate": 3.0472021317091745e-06, + "loss": 1.5041, + "step": 1334 + }, + { + "epoch": 1.3713405238828968, + "grad_norm": 2.4795925617218018, + "learning_rate": 3.0452988199467075e-06, + "loss": 1.4317, + "step": 1335 + }, + { + "epoch": 1.3723677452491012, + "grad_norm": 2.4608285427093506, + "learning_rate": 3.043395508184241e-06, + "loss": 1.4259, + "step": 1336 + }, + { + "epoch": 1.3733949666153056, + "grad_norm": 2.5970380306243896, + "learning_rate": 3.041492196421774e-06, + "loss": 1.3859, + "step": 1337 + }, + { + "epoch": 1.37442218798151, + "grad_norm": 2.517047166824341, + "learning_rate": 3.0395888846593076e-06, + "loss": 1.4086, + "step": 1338 + }, + { + "epoch": 1.3754494093477145, + "grad_norm": 2.4541358947753906, + "learning_rate": 3.037685572896841e-06, + "loss": 1.3959, + "step": 1339 + }, + { + "epoch": 1.376476630713919, + "grad_norm": 2.6143250465393066, + "learning_rate": 3.035782261134374e-06, + "loss": 1.4088, + "step": 1340 + }, + { + "epoch": 1.376476630713919, + "eval_loss": 1.7629026174545288, + "eval_runtime": 38.0156, + "eval_samples_per_second": 26.305, + "eval_steps_per_second": 1.105, + "step": 1340 + }, + { + "epoch": 1.3775038520801233, + "grad_norm": 2.763033151626587, + "learning_rate": 3.0338789493719072e-06, + "loss": 1.4104, + "step": 1341 + }, + { + "epoch": 1.3785310734463276, + "grad_norm": 2.523888349533081, + "learning_rate": 3.0319756376094406e-06, + "loss": 1.4073, + "step": 1342 + }, + { + "epoch": 1.379558294812532, + "grad_norm": 2.5393309593200684, + "learning_rate": 3.030072325846974e-06, + "loss": 1.4145, + "step": 1343 + }, + { + "epoch": 1.3805855161787366, + "grad_norm": 2.455733060836792, + "learning_rate": 3.0281690140845073e-06, + "loss": 1.3782, + "step": 1344 + }, + { + "epoch": 1.381612737544941, + "grad_norm": 2.600311517715454, + "learning_rate": 3.0262657023220403e-06, + "loss": 1.4158, + "step": 1345 + }, + { + "epoch": 1.3826399589111453, + "grad_norm": 2.453249931335449, + "learning_rate": 3.0243623905595736e-06, + "loss": 1.414, + "step": 1346 + }, + { + "epoch": 1.3836671802773497, + "grad_norm": 2.603515625, + "learning_rate": 3.022459078797107e-06, + "loss": 1.4102, + "step": 1347 + }, + { + "epoch": 1.384694401643554, + "grad_norm": 2.6515872478485107, + "learning_rate": 3.0205557670346404e-06, + "loss": 1.4063, + "step": 1348 + }, + { + "epoch": 1.3857216230097587, + "grad_norm": 2.412846565246582, + "learning_rate": 3.018652455272174e-06, + "loss": 1.3739, + "step": 1349 + }, + { + "epoch": 1.386748844375963, + "grad_norm": 2.748563051223755, + "learning_rate": 3.0167491435097075e-06, + "loss": 1.4909, + "step": 1350 + }, + { + "epoch": 1.386748844375963, + "eval_loss": 1.7653924226760864, + "eval_runtime": 37.9855, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1350 + }, + { + "epoch": 1.3877760657421674, + "grad_norm": 2.7861363887786865, + "learning_rate": 3.014845831747241e-06, + "loss": 1.4028, + "step": 1351 + }, + { + "epoch": 1.388803287108372, + "grad_norm": 2.5898189544677734, + "learning_rate": 3.012942519984774e-06, + "loss": 1.3793, + "step": 1352 + }, + { + "epoch": 1.3898305084745763, + "grad_norm": 2.4240355491638184, + "learning_rate": 3.011039208222307e-06, + "loss": 1.3889, + "step": 1353 + }, + { + "epoch": 1.3908577298407807, + "grad_norm": 2.4492504596710205, + "learning_rate": 3.0091358964598405e-06, + "loss": 1.3464, + "step": 1354 + }, + { + "epoch": 1.391884951206985, + "grad_norm": 2.4425718784332275, + "learning_rate": 3.007232584697374e-06, + "loss": 1.4073, + "step": 1355 + }, + { + "epoch": 1.3929121725731894, + "grad_norm": 2.5057594776153564, + "learning_rate": 3.0053292729349073e-06, + "loss": 1.4053, + "step": 1356 + }, + { + "epoch": 1.393939393939394, + "grad_norm": 2.5386273860931396, + "learning_rate": 3.00342596117244e-06, + "loss": 1.3991, + "step": 1357 + }, + { + "epoch": 1.3949666153055984, + "grad_norm": 2.6289730072021484, + "learning_rate": 3.0015226494099736e-06, + "loss": 1.3971, + "step": 1358 + }, + { + "epoch": 1.3959938366718028, + "grad_norm": 2.5091774463653564, + "learning_rate": 2.999619337647507e-06, + "loss": 1.3815, + "step": 1359 + }, + { + "epoch": 1.3970210580380071, + "grad_norm": 2.567659616470337, + "learning_rate": 2.9977160258850403e-06, + "loss": 1.3786, + "step": 1360 + }, + { + "epoch": 1.3970210580380071, + "eval_loss": 1.7706127166748047, + "eval_runtime": 37.9897, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1360 + }, + { + "epoch": 1.3980482794042115, + "grad_norm": 2.5508553981781006, + "learning_rate": 2.9958127141225736e-06, + "loss": 1.3629, + "step": 1361 + }, + { + "epoch": 1.399075500770416, + "grad_norm": 2.624800205230713, + "learning_rate": 2.9939094023601066e-06, + "loss": 1.4355, + "step": 1362 + }, + { + "epoch": 1.4001027221366205, + "grad_norm": 2.432114839553833, + "learning_rate": 2.99200609059764e-06, + "loss": 1.3995, + "step": 1363 + }, + { + "epoch": 1.4011299435028248, + "grad_norm": 2.4633607864379883, + "learning_rate": 2.9901027788351733e-06, + "loss": 1.3499, + "step": 1364 + }, + { + "epoch": 1.4021571648690292, + "grad_norm": 2.557415246963501, + "learning_rate": 2.9881994670727067e-06, + "loss": 1.3614, + "step": 1365 + }, + { + "epoch": 1.4031843862352336, + "grad_norm": 2.411295175552368, + "learning_rate": 2.98629615531024e-06, + "loss": 1.3857, + "step": 1366 + }, + { + "epoch": 1.4042116076014381, + "grad_norm": 2.429797410964966, + "learning_rate": 2.984392843547773e-06, + "loss": 1.3701, + "step": 1367 + }, + { + "epoch": 1.4052388289676425, + "grad_norm": 2.6080522537231445, + "learning_rate": 2.9824895317853063e-06, + "loss": 1.3996, + "step": 1368 + }, + { + "epoch": 1.4062660503338469, + "grad_norm": 2.5128798484802246, + "learning_rate": 2.98058622002284e-06, + "loss": 1.3973, + "step": 1369 + }, + { + "epoch": 1.4072932717000515, + "grad_norm": 2.5906126499176025, + "learning_rate": 2.9786829082603735e-06, + "loss": 1.4412, + "step": 1370 + }, + { + "epoch": 1.4072932717000515, + "eval_loss": 1.7700779438018799, + "eval_runtime": 38.017, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 1370 + }, + { + "epoch": 1.4083204930662558, + "grad_norm": 2.4890248775482178, + "learning_rate": 2.976779596497907e-06, + "loss": 1.4251, + "step": 1371 + }, + { + "epoch": 1.4093477144324602, + "grad_norm": 2.3803439140319824, + "learning_rate": 2.97487628473544e-06, + "loss": 1.4363, + "step": 1372 + }, + { + "epoch": 1.4103749357986646, + "grad_norm": 2.461958646774292, + "learning_rate": 2.9729729729729736e-06, + "loss": 1.3948, + "step": 1373 + }, + { + "epoch": 1.411402157164869, + "grad_norm": 2.419793128967285, + "learning_rate": 2.9710696612105065e-06, + "loss": 1.4367, + "step": 1374 + }, + { + "epoch": 1.4124293785310735, + "grad_norm": 2.5748846530914307, + "learning_rate": 2.96916634944804e-06, + "loss": 1.3873, + "step": 1375 + }, + { + "epoch": 1.413456599897278, + "grad_norm": 2.564558506011963, + "learning_rate": 2.9672630376855732e-06, + "loss": 1.4395, + "step": 1376 + }, + { + "epoch": 1.4144838212634823, + "grad_norm": 2.596085548400879, + "learning_rate": 2.9653597259231066e-06, + "loss": 1.4091, + "step": 1377 + }, + { + "epoch": 1.4155110426296866, + "grad_norm": 2.49029278755188, + "learning_rate": 2.96345641416064e-06, + "loss": 1.3639, + "step": 1378 + }, + { + "epoch": 1.416538263995891, + "grad_norm": 2.506887912750244, + "learning_rate": 2.961553102398173e-06, + "loss": 1.3065, + "step": 1379 + }, + { + "epoch": 1.4175654853620956, + "grad_norm": 2.517609119415283, + "learning_rate": 2.9596497906357063e-06, + "loss": 1.4365, + "step": 1380 + }, + { + "epoch": 1.4175654853620956, + "eval_loss": 1.7673696279525757, + "eval_runtime": 37.9941, + "eval_samples_per_second": 26.32, + "eval_steps_per_second": 1.105, + "step": 1380 + }, + { + "epoch": 1.4185927067283, + "grad_norm": 2.6870028972625732, + "learning_rate": 2.9577464788732396e-06, + "loss": 1.4386, + "step": 1381 + }, + { + "epoch": 1.4196199280945043, + "grad_norm": 2.3195550441741943, + "learning_rate": 2.955843167110773e-06, + "loss": 1.3771, + "step": 1382 + }, + { + "epoch": 1.420647149460709, + "grad_norm": 2.4306302070617676, + "learning_rate": 2.9539398553483064e-06, + "loss": 1.3963, + "step": 1383 + }, + { + "epoch": 1.4216743708269133, + "grad_norm": 2.4227781295776367, + "learning_rate": 2.9520365435858393e-06, + "loss": 1.394, + "step": 1384 + }, + { + "epoch": 1.4227015921931176, + "grad_norm": 2.521425485610962, + "learning_rate": 2.9501332318233727e-06, + "loss": 1.4664, + "step": 1385 + }, + { + "epoch": 1.423728813559322, + "grad_norm": 2.533363103866577, + "learning_rate": 2.948229920060906e-06, + "loss": 1.4072, + "step": 1386 + }, + { + "epoch": 1.4247560349255264, + "grad_norm": 2.378007173538208, + "learning_rate": 2.9463266082984394e-06, + "loss": 1.38, + "step": 1387 + }, + { + "epoch": 1.425783256291731, + "grad_norm": 2.476451873779297, + "learning_rate": 2.9444232965359727e-06, + "loss": 1.3864, + "step": 1388 + }, + { + "epoch": 1.4268104776579353, + "grad_norm": 2.428365707397461, + "learning_rate": 2.9425199847735057e-06, + "loss": 1.4081, + "step": 1389 + }, + { + "epoch": 1.4278376990241397, + "grad_norm": 2.4382991790771484, + "learning_rate": 2.94061667301104e-06, + "loss": 1.3806, + "step": 1390 + }, + { + "epoch": 1.4278376990241397, + "eval_loss": 1.7905043363571167, + "eval_runtime": 38.0264, + "eval_samples_per_second": 26.298, + "eval_steps_per_second": 1.104, + "step": 1390 + }, + { + "epoch": 1.428864920390344, + "grad_norm": 2.414433240890503, + "learning_rate": 2.938713361248573e-06, + "loss": 1.3113, + "step": 1391 + }, + { + "epoch": 1.4298921417565484, + "grad_norm": 2.450570821762085, + "learning_rate": 2.936810049486106e-06, + "loss": 1.4023, + "step": 1392 + }, + { + "epoch": 1.430919363122753, + "grad_norm": 2.4372148513793945, + "learning_rate": 2.9349067377236396e-06, + "loss": 1.4057, + "step": 1393 + }, + { + "epoch": 1.4319465844889574, + "grad_norm": 2.4278311729431152, + "learning_rate": 2.933003425961173e-06, + "loss": 1.3574, + "step": 1394 + }, + { + "epoch": 1.4329738058551618, + "grad_norm": 2.448517322540283, + "learning_rate": 2.9311001141987063e-06, + "loss": 1.3722, + "step": 1395 + }, + { + "epoch": 1.4340010272213661, + "grad_norm": 2.354310989379883, + "learning_rate": 2.9291968024362392e-06, + "loss": 1.3682, + "step": 1396 + }, + { + "epoch": 1.4350282485875705, + "grad_norm": 2.3318262100219727, + "learning_rate": 2.9272934906737726e-06, + "loss": 1.3604, + "step": 1397 + }, + { + "epoch": 1.436055469953775, + "grad_norm": 2.377561092376709, + "learning_rate": 2.925390178911306e-06, + "loss": 1.372, + "step": 1398 + }, + { + "epoch": 1.4370826913199795, + "grad_norm": 2.5543324947357178, + "learning_rate": 2.9234868671488393e-06, + "loss": 1.3846, + "step": 1399 + }, + { + "epoch": 1.4381099126861838, + "grad_norm": 2.6063973903656006, + "learning_rate": 2.9215835553863727e-06, + "loss": 1.3989, + "step": 1400 + }, + { + "epoch": 1.4381099126861838, + "eval_loss": 1.7730437517166138, + "eval_runtime": 38.0407, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 1400 + } + ], + "logging_steps": 1, + "max_steps": 2919, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 4.364233173106688e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..491d7aa6873ac345ad7d2d8238b3751df06748aa --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc77a0f3d732b76deec0b9f9ee77f2ad22c4c8a301af0167ff41a26d22d5beb3 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..0c1653c04de59e53b4d05673f04ed4b2c7039c73 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a2973146fd35e5e1a515f6332b8a2b33728c4360853ce2c03e1a7a90b79c2cc +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..4cf5bfa682e62dbe5c2e496536a84208dc8433d0 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d1658c107e9585fdf02f4eb4b846349794600f9e1c2f5151d959824c6021559 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..a3bc6e3b27283a03d7a973d49132444c5c2b8c03 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca21c6fccfe6f3e5961c1c2133919126cadcedbd519989dfde1a82374aa9c676 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1500/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..5509f0e1ece25901639e38b63337f0f28e33b98d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0123f1b841cbb90ad2cc1a10c38027bc63872dc02a831b18d254a6186f0ffff0 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..bdaddc4981c690246dfa665b10da40b38a679bc3 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86e82dbca3ee98b4aff4d87464271202c99a718b90530a7719851b8b6a191547 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..fd1ca9202bb0004e7be63108d043e05db918daf3 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fdc0a29e8caac820aa227d61191352d70074e7bc59c8656d34721d88d568ccb2 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..e2b7cd514696897148ef3dd45094ad3ec9e5060f --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cdfdecd5113b997bf1e3875b86693f06ede5dc3946dbb8e7c7e412180a03d47f +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1600/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1800/zero_to_fp32.py b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1800/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1800/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/generation_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/latest b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/latest new file mode 100644 index 0000000000000000000000000000000000000000..987364c74ed1983839a73116dad31651fadca561 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/latest @@ -0,0 +1 @@ +global_step1900 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/model.safetensors.index.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..abb168e98e42120baf8687307c412b6fa634e499 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57f176d3a980f67e95fd5a0692c1fe3a51c8aa772d429a141c1b7edffe66d794 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..d68e58486b1e6fbdc119f0904ce5ff91c752d7dd --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e31adeb682e8c4fa9e569fcaed25bff4e532b23d5c335f8e8ca96942ee81bc5e +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..7fcdcd8ef7aec1446fb95658c67ee1c412dc4760 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d20e1950654e165087114ba4a2ca396b7f4f6760619194a98f851bce1a6cdc0 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..d45ed7fefe7310564fcee65e294451d1b1a04578 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1d21ae4dcf31decd56f50288185409ae4dae50c7bf6974b104345305c5ebcd78 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/special_tokens_map.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/tokenizer_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0b782d4602b361f7577abb52b710a8c0e088588d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/tokenizer_config.json @@ -0,0 +1,2078 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128257": { + "content": "🅁", + "lstrip": true, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/trainer_state.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..6b1e586ea50fbcf8c63e8cf623afe7645f9ede89 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/trainer_state.json @@ -0,0 +1,14853 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.9517205957883923, + "eval_steps": 10, + "global_step": 1900, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7525, + "step": 1 + }, + { + "epoch": 0.002054442732408834, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7605, + "step": 2 + }, + { + "epoch": 0.0030816640986132513, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8008, + "step": 3 + }, + { + "epoch": 0.004108885464817668, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.809, + "step": 4 + }, + { + "epoch": 0.005136106831022085, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7841, + "step": 5 + }, + { + "epoch": 0.0061633281972265025, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7924, + "step": 6 + }, + { + "epoch": 0.007190549563430919, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7732, + "step": 7 + }, + { + "epoch": 0.008217770929635337, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7953, + "step": 8 + }, + { + "epoch": 0.009244992295839754, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7747, + "step": 9 + }, + { + "epoch": 0.01027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7909, + "step": 10 + }, + { + "epoch": 0.01027221366204417, + "eval_loss": 2.813098192214966, + "eval_runtime": 38.1187, + "eval_samples_per_second": 26.234, + "eval_steps_per_second": 1.102, + "step": 10 + }, + { + "epoch": 0.011299435028248588, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7592, + "step": 11 + }, + { + "epoch": 0.012326656394453005, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7394, + "step": 12 + }, + { + "epoch": 0.01335387776065742, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8111, + "step": 13 + }, + { + "epoch": 0.014381099126861838, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7786, + "step": 14 + }, + { + "epoch": 0.015408320493066256, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7711, + "step": 15 + }, + { + "epoch": 0.016435541859270673, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7622, + "step": 16 + }, + { + "epoch": 0.01746276322547509, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7291, + "step": 17 + }, + { + "epoch": 0.01848998459167951, + "grad_norm": 6.8249125480651855, + "learning_rate": 3.424657534246575e-08, + "loss": 2.7995, + "step": 18 + }, + { + "epoch": 0.019517205957883924, + "grad_norm": 5.581949710845947, + "learning_rate": 5.136986301369863e-08, + "loss": 2.5998, + "step": 19 + }, + { + "epoch": 0.02054442732408834, + "grad_norm": 3.0888617038726807, + "learning_rate": 6.84931506849315e-08, + "loss": 2.6031, + "step": 20 + }, + { + "epoch": 0.02054442732408834, + "eval_loss": 2.60642147064209, + "eval_runtime": 38.2226, + "eval_samples_per_second": 26.163, + "eval_steps_per_second": 1.099, + "step": 20 + }, + { + "epoch": 0.02157164869029276, + "grad_norm": 4.42941427230835, + "learning_rate": 8.561643835616439e-08, + "loss": 2.5816, + "step": 21 + }, + { + "epoch": 0.022598870056497175, + "grad_norm": 2.359813690185547, + "learning_rate": 1.0273972602739726e-07, + "loss": 2.5352, + "step": 22 + }, + { + "epoch": 0.02362609142270159, + "grad_norm": 2.9022927284240723, + "learning_rate": 1.1986301369863014e-07, + "loss": 2.4641, + "step": 23 + }, + { + "epoch": 0.02465331278890601, + "grad_norm": 1.6895997524261475, + "learning_rate": 1.36986301369863e-07, + "loss": 2.4582, + "step": 24 + }, + { + "epoch": 0.025680534155110426, + "grad_norm": 2.4555776119232178, + "learning_rate": 1.541095890410959e-07, + "loss": 2.4171, + "step": 25 + }, + { + "epoch": 0.02670775552131484, + "grad_norm": 1.9735779762268066, + "learning_rate": 1.7123287671232878e-07, + "loss": 2.4249, + "step": 26 + }, + { + "epoch": 0.02773497688751926, + "grad_norm": 2.71161150932312, + "learning_rate": 1.8835616438356165e-07, + "loss": 2.3791, + "step": 27 + }, + { + "epoch": 0.028762198253723677, + "grad_norm": 1.8482733964920044, + "learning_rate": 2.0547945205479452e-07, + "loss": 2.374, + "step": 28 + }, + { + "epoch": 0.029789419619928096, + "grad_norm": 2.191246747970581, + "learning_rate": 2.226027397260274e-07, + "loss": 2.3557, + "step": 29 + }, + { + "epoch": 0.030816640986132512, + "grad_norm": 2.0575971603393555, + "learning_rate": 2.397260273972603e-07, + "loss": 2.3315, + "step": 30 + }, + { + "epoch": 0.030816640986132512, + "eval_loss": 2.399639844894409, + "eval_runtime": 37.737, + "eval_samples_per_second": 26.499, + "eval_steps_per_second": 1.113, + "step": 30 + }, + { + "epoch": 0.03184386235233693, + "grad_norm": 2.0346758365631104, + "learning_rate": 2.568493150684932e-07, + "loss": 2.2791, + "step": 31 + }, + { + "epoch": 0.03287108371854135, + "grad_norm": 1.8710602521896362, + "learning_rate": 2.73972602739726e-07, + "loss": 2.3219, + "step": 32 + }, + { + "epoch": 0.03389830508474576, + "grad_norm": 1.9982675313949585, + "learning_rate": 2.910958904109589e-07, + "loss": 2.273, + "step": 33 + }, + { + "epoch": 0.03492552645095018, + "grad_norm": 2.092076301574707, + "learning_rate": 3.082191780821918e-07, + "loss": 2.2555, + "step": 34 + }, + { + "epoch": 0.035952747817154594, + "grad_norm": 1.870373249053955, + "learning_rate": 3.2534246575342466e-07, + "loss": 2.2314, + "step": 35 + }, + { + "epoch": 0.03697996918335902, + "grad_norm": 1.8378052711486816, + "learning_rate": 3.4246575342465755e-07, + "loss": 2.1244, + "step": 36 + }, + { + "epoch": 0.03800719054956343, + "grad_norm": 2.0156054496765137, + "learning_rate": 3.595890410958904e-07, + "loss": 2.2502, + "step": 37 + }, + { + "epoch": 0.03903441191576785, + "grad_norm": 1.6175312995910645, + "learning_rate": 3.767123287671233e-07, + "loss": 2.1566, + "step": 38 + }, + { + "epoch": 0.040061633281972264, + "grad_norm": 1.9051450490951538, + "learning_rate": 3.938356164383562e-07, + "loss": 2.167, + "step": 39 + }, + { + "epoch": 0.04108885464817668, + "grad_norm": 2.383063793182373, + "learning_rate": 4.1095890410958903e-07, + "loss": 2.2099, + "step": 40 + }, + { + "epoch": 0.04108885464817668, + "eval_loss": 2.2763354778289795, + "eval_runtime": 37.8085, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 40 + }, + { + "epoch": 0.042116076014381096, + "grad_norm": 2.1310720443725586, + "learning_rate": 4.2808219178082193e-07, + "loss": 2.1714, + "step": 41 + }, + { + "epoch": 0.04314329738058552, + "grad_norm": 1.771941900253296, + "learning_rate": 4.452054794520548e-07, + "loss": 2.0895, + "step": 42 + }, + { + "epoch": 0.044170518746789934, + "grad_norm": 2.3784172534942627, + "learning_rate": 4.6232876712328767e-07, + "loss": 2.1555, + "step": 43 + }, + { + "epoch": 0.04519774011299435, + "grad_norm": 1.6692813634872437, + "learning_rate": 4.794520547945206e-07, + "loss": 2.11, + "step": 44 + }, + { + "epoch": 0.046224961479198766, + "grad_norm": 2.1780643463134766, + "learning_rate": 4.965753424657534e-07, + "loss": 2.1256, + "step": 45 + }, + { + "epoch": 0.04725218284540318, + "grad_norm": 1.9390568733215332, + "learning_rate": 5.136986301369864e-07, + "loss": 2.0971, + "step": 46 + }, + { + "epoch": 0.048279404211607604, + "grad_norm": 2.276216506958008, + "learning_rate": 5.308219178082192e-07, + "loss": 2.1479, + "step": 47 + }, + { + "epoch": 0.04930662557781202, + "grad_norm": 2.148573160171509, + "learning_rate": 5.47945205479452e-07, + "loss": 2.0808, + "step": 48 + }, + { + "epoch": 0.050333846944016436, + "grad_norm": 2.259422540664673, + "learning_rate": 5.65068493150685e-07, + "loss": 2.0823, + "step": 49 + }, + { + "epoch": 0.05136106831022085, + "grad_norm": 1.7703359127044678, + "learning_rate": 5.821917808219178e-07, + "loss": 2.0236, + "step": 50 + }, + { + "epoch": 0.05136106831022085, + "eval_loss": 2.2037274837493896, + "eval_runtime": 37.8385, + "eval_samples_per_second": 26.428, + "eval_steps_per_second": 1.11, + "step": 50 + }, + { + "epoch": 0.05238828967642527, + "grad_norm": 2.56268310546875, + "learning_rate": 5.993150684931507e-07, + "loss": 2.0228, + "step": 51 + }, + { + "epoch": 0.05341551104262968, + "grad_norm": 2.5256848335266113, + "learning_rate": 6.164383561643836e-07, + "loss": 2.1401, + "step": 52 + }, + { + "epoch": 0.054442732408834106, + "grad_norm": 2.6329638957977295, + "learning_rate": 6.335616438356165e-07, + "loss": 2.0107, + "step": 53 + }, + { + "epoch": 0.05546995377503852, + "grad_norm": 2.477762222290039, + "learning_rate": 6.506849315068493e-07, + "loss": 2.0648, + "step": 54 + }, + { + "epoch": 0.05649717514124294, + "grad_norm": 1.8869198560714722, + "learning_rate": 6.678082191780823e-07, + "loss": 2.0214, + "step": 55 + }, + { + "epoch": 0.05752439650744735, + "grad_norm": 2.707345485687256, + "learning_rate": 6.849315068493151e-07, + "loss": 1.9893, + "step": 56 + }, + { + "epoch": 0.05855161787365177, + "grad_norm": 2.067230701446533, + "learning_rate": 7.020547945205481e-07, + "loss": 2.0308, + "step": 57 + }, + { + "epoch": 0.05957883923985619, + "grad_norm": 2.1961982250213623, + "learning_rate": 7.191780821917808e-07, + "loss": 1.948, + "step": 58 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 2.1032280921936035, + "learning_rate": 7.363013698630137e-07, + "loss": 2.0098, + "step": 59 + }, + { + "epoch": 0.061633281972265024, + "grad_norm": 2.5191144943237305, + "learning_rate": 7.534246575342466e-07, + "loss": 2.0183, + "step": 60 + }, + { + "epoch": 0.061633281972265024, + "eval_loss": 2.1474344730377197, + "eval_runtime": 37.8631, + "eval_samples_per_second": 26.411, + "eval_steps_per_second": 1.109, + "step": 60 + }, + { + "epoch": 0.06266050333846944, + "grad_norm": 2.177739143371582, + "learning_rate": 7.705479452054795e-07, + "loss": 1.9702, + "step": 61 + }, + { + "epoch": 0.06368772470467386, + "grad_norm": 2.499784231185913, + "learning_rate": 7.876712328767124e-07, + "loss": 1.9825, + "step": 62 + }, + { + "epoch": 0.06471494607087827, + "grad_norm": 2.181591749191284, + "learning_rate": 8.047945205479453e-07, + "loss": 2.0414, + "step": 63 + }, + { + "epoch": 0.0657421674370827, + "grad_norm": 1.9575999975204468, + "learning_rate": 8.219178082191781e-07, + "loss": 2.0115, + "step": 64 + }, + { + "epoch": 0.0667693888032871, + "grad_norm": 2.3431336879730225, + "learning_rate": 8.39041095890411e-07, + "loss": 1.8781, + "step": 65 + }, + { + "epoch": 0.06779661016949153, + "grad_norm": 2.042747974395752, + "learning_rate": 8.561643835616439e-07, + "loss": 1.9798, + "step": 66 + }, + { + "epoch": 0.06882383153569595, + "grad_norm": 2.784104585647583, + "learning_rate": 8.732876712328768e-07, + "loss": 1.9274, + "step": 67 + }, + { + "epoch": 0.06985105290190036, + "grad_norm": 1.9781445264816284, + "learning_rate": 8.904109589041097e-07, + "loss": 2.0036, + "step": 68 + }, + { + "epoch": 0.07087827426810478, + "grad_norm": 2.631537437438965, + "learning_rate": 9.075342465753426e-07, + "loss": 1.938, + "step": 69 + }, + { + "epoch": 0.07190549563430919, + "grad_norm": 1.902701497077942, + "learning_rate": 9.246575342465753e-07, + "loss": 2.0545, + "step": 70 + }, + { + "epoch": 0.07190549563430919, + "eval_loss": 2.115269422531128, + "eval_runtime": 37.887, + "eval_samples_per_second": 26.394, + "eval_steps_per_second": 1.109, + "step": 70 + }, + { + "epoch": 0.07293271700051361, + "grad_norm": 1.9323904514312744, + "learning_rate": 9.417808219178083e-07, + "loss": 1.9463, + "step": 71 + }, + { + "epoch": 0.07395993836671803, + "grad_norm": 2.1631956100463867, + "learning_rate": 9.589041095890411e-07, + "loss": 1.8982, + "step": 72 + }, + { + "epoch": 0.07498715973292244, + "grad_norm": 1.959518313407898, + "learning_rate": 9.76027397260274e-07, + "loss": 1.9538, + "step": 73 + }, + { + "epoch": 0.07601438109912687, + "grad_norm": 1.9834569692611694, + "learning_rate": 9.931506849315068e-07, + "loss": 1.9257, + "step": 74 + }, + { + "epoch": 0.07704160246533127, + "grad_norm": 2.2818732261657715, + "learning_rate": 1.0102739726027399e-06, + "loss": 1.9318, + "step": 75 + }, + { + "epoch": 0.0780688238315357, + "grad_norm": 1.7696406841278076, + "learning_rate": 1.0273972602739727e-06, + "loss": 1.912, + "step": 76 + }, + { + "epoch": 0.07909604519774012, + "grad_norm": 2.396303176879883, + "learning_rate": 1.0445205479452056e-06, + "loss": 1.9482, + "step": 77 + }, + { + "epoch": 0.08012326656394453, + "grad_norm": 2.2371881008148193, + "learning_rate": 1.0616438356164384e-06, + "loss": 1.9448, + "step": 78 + }, + { + "epoch": 0.08115048793014895, + "grad_norm": 2.40077543258667, + "learning_rate": 1.0787671232876712e-06, + "loss": 1.8764, + "step": 79 + }, + { + "epoch": 0.08217770929635336, + "grad_norm": 2.267916679382324, + "learning_rate": 1.095890410958904e-06, + "loss": 1.9478, + "step": 80 + }, + { + "epoch": 0.08217770929635336, + "eval_loss": 2.0851893424987793, + "eval_runtime": 37.9176, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 80 + }, + { + "epoch": 0.08320493066255778, + "grad_norm": 2.286259651184082, + "learning_rate": 1.1130136986301371e-06, + "loss": 1.8856, + "step": 81 + }, + { + "epoch": 0.08423215202876219, + "grad_norm": 2.160848379135132, + "learning_rate": 1.13013698630137e-06, + "loss": 1.9196, + "step": 82 + }, + { + "epoch": 0.08525937339496661, + "grad_norm": 2.3618931770324707, + "learning_rate": 1.1472602739726028e-06, + "loss": 1.9229, + "step": 83 + }, + { + "epoch": 0.08628659476117104, + "grad_norm": 2.1963961124420166, + "learning_rate": 1.1643835616438357e-06, + "loss": 1.8998, + "step": 84 + }, + { + "epoch": 0.08731381612737545, + "grad_norm": 2.197981357574463, + "learning_rate": 1.1815068493150685e-06, + "loss": 1.8716, + "step": 85 + }, + { + "epoch": 0.08834103749357987, + "grad_norm": 2.3562612533569336, + "learning_rate": 1.1986301369863014e-06, + "loss": 1.9336, + "step": 86 + }, + { + "epoch": 0.08936825885978428, + "grad_norm": 2.49712872505188, + "learning_rate": 1.2157534246575344e-06, + "loss": 1.8827, + "step": 87 + }, + { + "epoch": 0.0903954802259887, + "grad_norm": 2.315244436264038, + "learning_rate": 1.2328767123287673e-06, + "loss": 1.8801, + "step": 88 + }, + { + "epoch": 0.09142270159219312, + "grad_norm": 2.774587869644165, + "learning_rate": 1.25e-06, + "loss": 1.8972, + "step": 89 + }, + { + "epoch": 0.09244992295839753, + "grad_norm": 2.262364387512207, + "learning_rate": 1.267123287671233e-06, + "loss": 1.8426, + "step": 90 + }, + { + "epoch": 0.09244992295839753, + "eval_loss": 2.069462299346924, + "eval_runtime": 37.9319, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 90 + }, + { + "epoch": 0.09347714432460195, + "grad_norm": 2.6195597648620605, + "learning_rate": 1.284246575342466e-06, + "loss": 1.871, + "step": 91 + }, + { + "epoch": 0.09450436569080636, + "grad_norm": 2.051847219467163, + "learning_rate": 1.3013698630136986e-06, + "loss": 1.8922, + "step": 92 + }, + { + "epoch": 0.09553158705701079, + "grad_norm": 2.5930886268615723, + "learning_rate": 1.3184931506849317e-06, + "loss": 1.8818, + "step": 93 + }, + { + "epoch": 0.09655880842321521, + "grad_norm": 2.304164171218872, + "learning_rate": 1.3356164383561645e-06, + "loss": 1.8779, + "step": 94 + }, + { + "epoch": 0.09758602978941962, + "grad_norm": 2.187354564666748, + "learning_rate": 1.3527397260273976e-06, + "loss": 1.9272, + "step": 95 + }, + { + "epoch": 0.09861325115562404, + "grad_norm": 2.2492458820343018, + "learning_rate": 1.3698630136986302e-06, + "loss": 1.805, + "step": 96 + }, + { + "epoch": 0.09964047252182845, + "grad_norm": 2.3325352668762207, + "learning_rate": 1.386986301369863e-06, + "loss": 1.8817, + "step": 97 + }, + { + "epoch": 0.10066769388803287, + "grad_norm": 2.1708617210388184, + "learning_rate": 1.4041095890410961e-06, + "loss": 1.7956, + "step": 98 + }, + { + "epoch": 0.1016949152542373, + "grad_norm": 2.4267706871032715, + "learning_rate": 1.421232876712329e-06, + "loss": 1.8116, + "step": 99 + }, + { + "epoch": 0.1027221366204417, + "grad_norm": 2.3286805152893066, + "learning_rate": 1.4383561643835616e-06, + "loss": 1.8299, + "step": 100 + }, + { + "epoch": 0.1027221366204417, + "eval_loss": 2.0549135208129883, + "eval_runtime": 37.935, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 100 + }, + { + "epoch": 0.10374935798664613, + "grad_norm": 2.5391643047332764, + "learning_rate": 1.4554794520547946e-06, + "loss": 1.8521, + "step": 101 + }, + { + "epoch": 0.10477657935285054, + "grad_norm": 2.018380880355835, + "learning_rate": 1.4726027397260275e-06, + "loss": 1.8209, + "step": 102 + }, + { + "epoch": 0.10580380071905496, + "grad_norm": 2.3706746101379395, + "learning_rate": 1.4897260273972605e-06, + "loss": 1.79, + "step": 103 + }, + { + "epoch": 0.10683102208525937, + "grad_norm": 2.1290764808654785, + "learning_rate": 1.5068493150684932e-06, + "loss": 1.8141, + "step": 104 + }, + { + "epoch": 0.10785824345146379, + "grad_norm": 2.2768120765686035, + "learning_rate": 1.5239726027397262e-06, + "loss": 1.8826, + "step": 105 + }, + { + "epoch": 0.10888546481766821, + "grad_norm": 2.316584348678589, + "learning_rate": 1.541095890410959e-06, + "loss": 1.8604, + "step": 106 + }, + { + "epoch": 0.10991268618387262, + "grad_norm": 2.14679217338562, + "learning_rate": 1.5582191780821921e-06, + "loss": 1.865, + "step": 107 + }, + { + "epoch": 0.11093990755007704, + "grad_norm": 2.3288021087646484, + "learning_rate": 1.5753424657534248e-06, + "loss": 1.8512, + "step": 108 + }, + { + "epoch": 0.11196712891628145, + "grad_norm": 2.300482988357544, + "learning_rate": 1.5924657534246576e-06, + "loss": 1.7455, + "step": 109 + }, + { + "epoch": 0.11299435028248588, + "grad_norm": 2.4226229190826416, + "learning_rate": 1.6095890410958907e-06, + "loss": 1.8058, + "step": 110 + }, + { + "epoch": 0.11299435028248588, + "eval_loss": 2.025261640548706, + "eval_runtime": 37.6779, + "eval_samples_per_second": 26.541, + "eval_steps_per_second": 1.115, + "step": 110 + }, + { + "epoch": 0.1140215716486903, + "grad_norm": 2.0930111408233643, + "learning_rate": 1.6267123287671235e-06, + "loss": 1.7375, + "step": 111 + }, + { + "epoch": 0.1150487930148947, + "grad_norm": 2.4973645210266113, + "learning_rate": 1.6438356164383561e-06, + "loss": 1.8268, + "step": 112 + }, + { + "epoch": 0.11607601438109913, + "grad_norm": 2.3295114040374756, + "learning_rate": 1.6609589041095892e-06, + "loss": 1.7882, + "step": 113 + }, + { + "epoch": 0.11710323574730354, + "grad_norm": 2.3693318367004395, + "learning_rate": 1.678082191780822e-06, + "loss": 1.8192, + "step": 114 + }, + { + "epoch": 0.11813045711350796, + "grad_norm": 2.553637742996216, + "learning_rate": 1.695205479452055e-06, + "loss": 1.7491, + "step": 115 + }, + { + "epoch": 0.11915767847971238, + "grad_norm": 2.303945541381836, + "learning_rate": 1.7123287671232877e-06, + "loss": 1.8737, + "step": 116 + }, + { + "epoch": 0.12018489984591679, + "grad_norm": 2.250356435775757, + "learning_rate": 1.7294520547945206e-06, + "loss": 1.8039, + "step": 117 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 2.3854403495788574, + "learning_rate": 1.7465753424657536e-06, + "loss": 1.8357, + "step": 118 + }, + { + "epoch": 0.12223934257832562, + "grad_norm": 2.4049572944641113, + "learning_rate": 1.7636986301369865e-06, + "loss": 1.8209, + "step": 119 + }, + { + "epoch": 0.12326656394453005, + "grad_norm": 2.313551425933838, + "learning_rate": 1.7808219178082193e-06, + "loss": 1.9026, + "step": 120 + }, + { + "epoch": 0.12326656394453005, + "eval_loss": 2.012171745300293, + "eval_runtime": 37.7908, + "eval_samples_per_second": 26.461, + "eval_steps_per_second": 1.111, + "step": 120 + }, + { + "epoch": 0.12429378531073447, + "grad_norm": 2.518320322036743, + "learning_rate": 1.7979452054794521e-06, + "loss": 1.8241, + "step": 121 + }, + { + "epoch": 0.12532100667693888, + "grad_norm": 2.6504592895507812, + "learning_rate": 1.8150684931506852e-06, + "loss": 1.8091, + "step": 122 + }, + { + "epoch": 0.1263482280431433, + "grad_norm": 2.556610107421875, + "learning_rate": 1.832191780821918e-06, + "loss": 1.7547, + "step": 123 + }, + { + "epoch": 0.12737544940934772, + "grad_norm": 2.247633695602417, + "learning_rate": 1.8493150684931507e-06, + "loss": 1.7923, + "step": 124 + }, + { + "epoch": 0.12840267077555212, + "grad_norm": 2.195596933364868, + "learning_rate": 1.8664383561643837e-06, + "loss": 1.747, + "step": 125 + }, + { + "epoch": 0.12942989214175654, + "grad_norm": 2.121396780014038, + "learning_rate": 1.8835616438356166e-06, + "loss": 1.8088, + "step": 126 + }, + { + "epoch": 0.13045711350796096, + "grad_norm": 2.4572501182556152, + "learning_rate": 1.9006849315068496e-06, + "loss": 1.7285, + "step": 127 + }, + { + "epoch": 0.1314843348741654, + "grad_norm": 2.2590177059173584, + "learning_rate": 1.9178082191780823e-06, + "loss": 1.7929, + "step": 128 + }, + { + "epoch": 0.1325115562403698, + "grad_norm": 2.5099234580993652, + "learning_rate": 1.9349315068493153e-06, + "loss": 1.8365, + "step": 129 + }, + { + "epoch": 0.1335387776065742, + "grad_norm": 2.3064465522766113, + "learning_rate": 1.952054794520548e-06, + "loss": 1.8351, + "step": 130 + }, + { + "epoch": 0.1335387776065742, + "eval_loss": 2.0043866634368896, + "eval_runtime": 37.8661, + "eval_samples_per_second": 26.409, + "eval_steps_per_second": 1.109, + "step": 130 + }, + { + "epoch": 0.13456599897277863, + "grad_norm": 2.5977718830108643, + "learning_rate": 1.969178082191781e-06, + "loss": 1.7603, + "step": 131 + }, + { + "epoch": 0.13559322033898305, + "grad_norm": 2.23221755027771, + "learning_rate": 1.9863013698630136e-06, + "loss": 1.8237, + "step": 132 + }, + { + "epoch": 0.13662044170518747, + "grad_norm": 2.247044324874878, + "learning_rate": 2.0034246575342467e-06, + "loss": 1.8006, + "step": 133 + }, + { + "epoch": 0.1376476630713919, + "grad_norm": 2.1796107292175293, + "learning_rate": 2.0205479452054797e-06, + "loss": 1.7354, + "step": 134 + }, + { + "epoch": 0.1386748844375963, + "grad_norm": 2.3269693851470947, + "learning_rate": 2.037671232876713e-06, + "loss": 1.783, + "step": 135 + }, + { + "epoch": 0.1397021058038007, + "grad_norm": 2.2903053760528564, + "learning_rate": 2.0547945205479454e-06, + "loss": 1.8072, + "step": 136 + }, + { + "epoch": 0.14072932717000514, + "grad_norm": 2.634150981903076, + "learning_rate": 2.071917808219178e-06, + "loss": 1.7251, + "step": 137 + }, + { + "epoch": 0.14175654853620956, + "grad_norm": 2.42579984664917, + "learning_rate": 2.089041095890411e-06, + "loss": 1.7997, + "step": 138 + }, + { + "epoch": 0.14278376990241398, + "grad_norm": 2.615828514099121, + "learning_rate": 2.106164383561644e-06, + "loss": 1.766, + "step": 139 + }, + { + "epoch": 0.14381099126861838, + "grad_norm": 2.312714099884033, + "learning_rate": 2.123287671232877e-06, + "loss": 1.7538, + "step": 140 + }, + { + "epoch": 0.14381099126861838, + "eval_loss": 1.9918214082717896, + "eval_runtime": 37.8698, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 140 + }, + { + "epoch": 0.1448382126348228, + "grad_norm": 2.7540030479431152, + "learning_rate": 2.14041095890411e-06, + "loss": 1.7856, + "step": 141 + }, + { + "epoch": 0.14586543400102722, + "grad_norm": 2.289191246032715, + "learning_rate": 2.1575342465753425e-06, + "loss": 1.7288, + "step": 142 + }, + { + "epoch": 0.14689265536723164, + "grad_norm": 2.788001775741577, + "learning_rate": 2.1746575342465755e-06, + "loss": 1.8703, + "step": 143 + }, + { + "epoch": 0.14791987673343607, + "grad_norm": 2.4128854274749756, + "learning_rate": 2.191780821917808e-06, + "loss": 1.7461, + "step": 144 + }, + { + "epoch": 0.14894709809964046, + "grad_norm": 2.4108099937438965, + "learning_rate": 2.2089041095890412e-06, + "loss": 1.73, + "step": 145 + }, + { + "epoch": 0.14997431946584489, + "grad_norm": 2.4315948486328125, + "learning_rate": 2.2260273972602743e-06, + "loss": 1.7303, + "step": 146 + }, + { + "epoch": 0.1510015408320493, + "grad_norm": 2.494056224822998, + "learning_rate": 2.243150684931507e-06, + "loss": 1.7616, + "step": 147 + }, + { + "epoch": 0.15202876219825373, + "grad_norm": 2.4557394981384277, + "learning_rate": 2.26027397260274e-06, + "loss": 1.7655, + "step": 148 + }, + { + "epoch": 0.15305598356445815, + "grad_norm": 2.4114091396331787, + "learning_rate": 2.2773972602739726e-06, + "loss": 1.8538, + "step": 149 + }, + { + "epoch": 0.15408320493066255, + "grad_norm": 2.609219789505005, + "learning_rate": 2.2945205479452057e-06, + "loss": 1.7767, + "step": 150 + }, + { + "epoch": 0.15408320493066255, + "eval_loss": 1.981367588043213, + "eval_runtime": 37.9562, + "eval_samples_per_second": 26.346, + "eval_steps_per_second": 1.107, + "step": 150 + }, + { + "epoch": 0.15511042629686697, + "grad_norm": 2.213622808456421, + "learning_rate": 2.3116438356164387e-06, + "loss": 1.687, + "step": 151 + }, + { + "epoch": 0.1561376476630714, + "grad_norm": 2.6517975330352783, + "learning_rate": 2.3287671232876713e-06, + "loss": 1.7311, + "step": 152 + }, + { + "epoch": 0.15716486902927582, + "grad_norm": 2.5174458026885986, + "learning_rate": 2.3458904109589044e-06, + "loss": 1.7835, + "step": 153 + }, + { + "epoch": 0.15819209039548024, + "grad_norm": 2.485480308532715, + "learning_rate": 2.363013698630137e-06, + "loss": 1.7749, + "step": 154 + }, + { + "epoch": 0.15921931176168463, + "grad_norm": 2.5386061668395996, + "learning_rate": 2.38013698630137e-06, + "loss": 1.7223, + "step": 155 + }, + { + "epoch": 0.16024653312788906, + "grad_norm": 2.2895941734313965, + "learning_rate": 2.3972602739726027e-06, + "loss": 1.6845, + "step": 156 + }, + { + "epoch": 0.16127375449409348, + "grad_norm": 2.5453789234161377, + "learning_rate": 2.4143835616438358e-06, + "loss": 1.7627, + "step": 157 + }, + { + "epoch": 0.1623009758602979, + "grad_norm": 2.5322303771972656, + "learning_rate": 2.431506849315069e-06, + "loss": 1.764, + "step": 158 + }, + { + "epoch": 0.1633281972265023, + "grad_norm": 2.2669899463653564, + "learning_rate": 2.4486301369863015e-06, + "loss": 1.7487, + "step": 159 + }, + { + "epoch": 0.16435541859270672, + "grad_norm": 2.4577853679656982, + "learning_rate": 2.4657534246575345e-06, + "loss": 1.7635, + "step": 160 + }, + { + "epoch": 0.16435541859270672, + "eval_loss": 1.9793355464935303, + "eval_runtime": 37.8576, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 160 + }, + { + "epoch": 0.16538263995891114, + "grad_norm": 2.354893922805786, + "learning_rate": 2.482876712328767e-06, + "loss": 1.7538, + "step": 161 + }, + { + "epoch": 0.16640986132511557, + "grad_norm": 2.3255856037139893, + "learning_rate": 2.5e-06, + "loss": 1.7277, + "step": 162 + }, + { + "epoch": 0.16743708269132, + "grad_norm": 2.3501994609832764, + "learning_rate": 2.5171232876712333e-06, + "loss": 1.6985, + "step": 163 + }, + { + "epoch": 0.16846430405752438, + "grad_norm": 2.386627674102783, + "learning_rate": 2.534246575342466e-06, + "loss": 1.7552, + "step": 164 + }, + { + "epoch": 0.1694915254237288, + "grad_norm": 2.2238457202911377, + "learning_rate": 2.551369863013699e-06, + "loss": 1.7805, + "step": 165 + }, + { + "epoch": 0.17051874678993323, + "grad_norm": 2.2653403282165527, + "learning_rate": 2.568493150684932e-06, + "loss": 1.7656, + "step": 166 + }, + { + "epoch": 0.17154596815613765, + "grad_norm": 2.233132839202881, + "learning_rate": 2.585616438356164e-06, + "loss": 1.6997, + "step": 167 + }, + { + "epoch": 0.17257318952234207, + "grad_norm": 2.1696043014526367, + "learning_rate": 2.6027397260273973e-06, + "loss": 1.7487, + "step": 168 + }, + { + "epoch": 0.17360041088854647, + "grad_norm": 2.4479899406433105, + "learning_rate": 2.6198630136986303e-06, + "loss": 1.7957, + "step": 169 + }, + { + "epoch": 0.1746276322547509, + "grad_norm": 2.3054702281951904, + "learning_rate": 2.6369863013698634e-06, + "loss": 1.766, + "step": 170 + }, + { + "epoch": 0.1746276322547509, + "eval_loss": 1.9543790817260742, + "eval_runtime": 37.8837, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 170 + }, + { + "epoch": 0.17565485362095531, + "grad_norm": 2.3591110706329346, + "learning_rate": 2.654109589041096e-06, + "loss": 1.7149, + "step": 171 + }, + { + "epoch": 0.17668207498715974, + "grad_norm": 2.450526237487793, + "learning_rate": 2.671232876712329e-06, + "loss": 1.7074, + "step": 172 + }, + { + "epoch": 0.17770929635336416, + "grad_norm": 2.1511754989624023, + "learning_rate": 2.688356164383562e-06, + "loss": 1.7036, + "step": 173 + }, + { + "epoch": 0.17873651771956856, + "grad_norm": 2.5944275856018066, + "learning_rate": 2.705479452054795e-06, + "loss": 1.6566, + "step": 174 + }, + { + "epoch": 0.17976373908577298, + "grad_norm": 2.4301369190216064, + "learning_rate": 2.7226027397260274e-06, + "loss": 1.7977, + "step": 175 + }, + { + "epoch": 0.1807909604519774, + "grad_norm": 2.287665367126465, + "learning_rate": 2.7397260273972604e-06, + "loss": 1.7526, + "step": 176 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.380032777786255, + "learning_rate": 2.7568493150684935e-06, + "loss": 1.6753, + "step": 177 + }, + { + "epoch": 0.18284540318438625, + "grad_norm": 2.4040510654449463, + "learning_rate": 2.773972602739726e-06, + "loss": 1.758, + "step": 178 + }, + { + "epoch": 0.18387262455059064, + "grad_norm": 2.5074045658111572, + "learning_rate": 2.791095890410959e-06, + "loss": 1.6498, + "step": 179 + }, + { + "epoch": 0.18489984591679506, + "grad_norm": 2.3926949501037598, + "learning_rate": 2.8082191780821922e-06, + "loss": 1.8006, + "step": 180 + }, + { + "epoch": 0.18489984591679506, + "eval_loss": 1.9446762800216675, + "eval_runtime": 37.8572, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 180 + }, + { + "epoch": 0.1859270672829995, + "grad_norm": 2.4991633892059326, + "learning_rate": 2.8253424657534253e-06, + "loss": 1.7649, + "step": 181 + }, + { + "epoch": 0.1869542886492039, + "grad_norm": 2.2558324337005615, + "learning_rate": 2.842465753424658e-06, + "loss": 1.7743, + "step": 182 + }, + { + "epoch": 0.18798151001540833, + "grad_norm": 2.219717025756836, + "learning_rate": 2.8595890410958905e-06, + "loss": 1.7807, + "step": 183 + }, + { + "epoch": 0.18900873138161273, + "grad_norm": 2.406562566757202, + "learning_rate": 2.876712328767123e-06, + "loss": 1.6289, + "step": 184 + }, + { + "epoch": 0.19003595274781715, + "grad_norm": 2.356579303741455, + "learning_rate": 2.8938356164383562e-06, + "loss": 1.6747, + "step": 185 + }, + { + "epoch": 0.19106317411402157, + "grad_norm": 2.4844322204589844, + "learning_rate": 2.9109589041095893e-06, + "loss": 1.6731, + "step": 186 + }, + { + "epoch": 0.192090395480226, + "grad_norm": 2.8179149627685547, + "learning_rate": 2.9280821917808223e-06, + "loss": 1.6786, + "step": 187 + }, + { + "epoch": 0.19311761684643042, + "grad_norm": 2.395416259765625, + "learning_rate": 2.945205479452055e-06, + "loss": 1.7313, + "step": 188 + }, + { + "epoch": 0.1941448382126348, + "grad_norm": 2.848740816116333, + "learning_rate": 2.962328767123288e-06, + "loss": 1.7258, + "step": 189 + }, + { + "epoch": 0.19517205957883924, + "grad_norm": 2.427395820617676, + "learning_rate": 2.979452054794521e-06, + "loss": 1.6633, + "step": 190 + }, + { + "epoch": 0.19517205957883924, + "eval_loss": 1.9443999528884888, + "eval_runtime": 37.8579, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 190 + }, + { + "epoch": 0.19619928094504366, + "grad_norm": 2.5347256660461426, + "learning_rate": 2.9965753424657533e-06, + "loss": 1.7342, + "step": 191 + }, + { + "epoch": 0.19722650231124808, + "grad_norm": 2.8696866035461426, + "learning_rate": 3.0136986301369864e-06, + "loss": 1.7037, + "step": 192 + }, + { + "epoch": 0.1982537236774525, + "grad_norm": 2.427468776702881, + "learning_rate": 3.0308219178082194e-06, + "loss": 1.672, + "step": 193 + }, + { + "epoch": 0.1992809450436569, + "grad_norm": 2.4253737926483154, + "learning_rate": 3.0479452054794525e-06, + "loss": 1.7252, + "step": 194 + }, + { + "epoch": 0.20030816640986132, + "grad_norm": 2.801668643951416, + "learning_rate": 3.065068493150685e-06, + "loss": 1.6694, + "step": 195 + }, + { + "epoch": 0.20133538777606574, + "grad_norm": 2.442263603210449, + "learning_rate": 3.082191780821918e-06, + "loss": 1.6866, + "step": 196 + }, + { + "epoch": 0.20236260914227017, + "grad_norm": 2.777463674545288, + "learning_rate": 3.099315068493151e-06, + "loss": 1.686, + "step": 197 + }, + { + "epoch": 0.2033898305084746, + "grad_norm": 2.491227388381958, + "learning_rate": 3.1164383561643843e-06, + "loss": 1.6414, + "step": 198 + }, + { + "epoch": 0.20441705187467898, + "grad_norm": 2.33585262298584, + "learning_rate": 3.1335616438356165e-06, + "loss": 1.7808, + "step": 199 + }, + { + "epoch": 0.2054442732408834, + "grad_norm": 2.518160581588745, + "learning_rate": 3.1506849315068495e-06, + "loss": 1.6592, + "step": 200 + }, + { + "epoch": 0.2054442732408834, + "eval_loss": 1.9371097087860107, + "eval_runtime": 37.8602, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 200 + }, + { + "epoch": 0.20647149460708783, + "grad_norm": 2.460204839706421, + "learning_rate": 3.167808219178082e-06, + "loss": 1.6668, + "step": 201 + }, + { + "epoch": 0.20749871597329225, + "grad_norm": 2.377824068069458, + "learning_rate": 3.184931506849315e-06, + "loss": 1.6462, + "step": 202 + }, + { + "epoch": 0.20852593733949668, + "grad_norm": 2.426177740097046, + "learning_rate": 3.2020547945205483e-06, + "loss": 1.6272, + "step": 203 + }, + { + "epoch": 0.20955315870570107, + "grad_norm": 2.4532034397125244, + "learning_rate": 3.2191780821917813e-06, + "loss": 1.6169, + "step": 204 + }, + { + "epoch": 0.2105803800719055, + "grad_norm": 2.517059564590454, + "learning_rate": 3.236301369863014e-06, + "loss": 1.5991, + "step": 205 + }, + { + "epoch": 0.21160760143810992, + "grad_norm": 2.22658109664917, + "learning_rate": 3.253424657534247e-06, + "loss": 1.6754, + "step": 206 + }, + { + "epoch": 0.21263482280431434, + "grad_norm": 2.344799518585205, + "learning_rate": 3.2705479452054796e-06, + "loss": 1.6648, + "step": 207 + }, + { + "epoch": 0.21366204417051873, + "grad_norm": 2.331875801086426, + "learning_rate": 3.2876712328767123e-06, + "loss": 1.6112, + "step": 208 + }, + { + "epoch": 0.21468926553672316, + "grad_norm": 2.262343645095825, + "learning_rate": 3.3047945205479453e-06, + "loss": 1.6479, + "step": 209 + }, + { + "epoch": 0.21571648690292758, + "grad_norm": 2.282278537750244, + "learning_rate": 3.3219178082191784e-06, + "loss": 1.6753, + "step": 210 + }, + { + "epoch": 0.21571648690292758, + "eval_loss": 1.9338418245315552, + "eval_runtime": 37.6083, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 210 + }, + { + "epoch": 0.216743708269132, + "grad_norm": 2.578476905822754, + "learning_rate": 3.3390410958904114e-06, + "loss": 1.7043, + "step": 211 + }, + { + "epoch": 0.21777092963533642, + "grad_norm": 2.3118813037872314, + "learning_rate": 3.356164383561644e-06, + "loss": 1.7541, + "step": 212 + }, + { + "epoch": 0.21879815100154082, + "grad_norm": 2.2418692111968994, + "learning_rate": 3.373287671232877e-06, + "loss": 1.6886, + "step": 213 + }, + { + "epoch": 0.21982537236774524, + "grad_norm": 2.3583438396453857, + "learning_rate": 3.39041095890411e-06, + "loss": 1.7241, + "step": 214 + }, + { + "epoch": 0.22085259373394966, + "grad_norm": 2.345090866088867, + "learning_rate": 3.4075342465753424e-06, + "loss": 1.6194, + "step": 215 + }, + { + "epoch": 0.2218798151001541, + "grad_norm": 2.3996200561523438, + "learning_rate": 3.4246575342465754e-06, + "loss": 1.6889, + "step": 216 + }, + { + "epoch": 0.2229070364663585, + "grad_norm": 2.3552374839782715, + "learning_rate": 3.4417808219178085e-06, + "loss": 1.6713, + "step": 217 + }, + { + "epoch": 0.2239342578325629, + "grad_norm": 2.2653744220733643, + "learning_rate": 3.458904109589041e-06, + "loss": 1.673, + "step": 218 + }, + { + "epoch": 0.22496147919876733, + "grad_norm": 2.263288974761963, + "learning_rate": 3.476027397260274e-06, + "loss": 1.673, + "step": 219 + }, + { + "epoch": 0.22598870056497175, + "grad_norm": 2.3173880577087402, + "learning_rate": 3.4931506849315072e-06, + "loss": 1.6803, + "step": 220 + }, + { + "epoch": 0.22598870056497175, + "eval_loss": 1.9166476726531982, + "eval_runtime": 37.7659, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 220 + }, + { + "epoch": 0.22701592193117617, + "grad_norm": 2.4005839824676514, + "learning_rate": 3.5102739726027403e-06, + "loss": 1.6809, + "step": 221 + }, + { + "epoch": 0.2280431432973806, + "grad_norm": 2.5032525062561035, + "learning_rate": 3.527397260273973e-06, + "loss": 1.6411, + "step": 222 + }, + { + "epoch": 0.229070364663585, + "grad_norm": 2.533583879470825, + "learning_rate": 3.5445205479452056e-06, + "loss": 1.6608, + "step": 223 + }, + { + "epoch": 0.2300975860297894, + "grad_norm": 2.3575572967529297, + "learning_rate": 3.5616438356164386e-06, + "loss": 1.7236, + "step": 224 + }, + { + "epoch": 0.23112480739599384, + "grad_norm": 2.3408942222595215, + "learning_rate": 3.5787671232876712e-06, + "loss": 1.7101, + "step": 225 + }, + { + "epoch": 0.23215202876219826, + "grad_norm": 2.3819596767425537, + "learning_rate": 3.5958904109589043e-06, + "loss": 1.6414, + "step": 226 + }, + { + "epoch": 0.23317925012840268, + "grad_norm": 2.2874338626861572, + "learning_rate": 3.6130136986301373e-06, + "loss": 1.6767, + "step": 227 + }, + { + "epoch": 0.23420647149460708, + "grad_norm": 2.3523948192596436, + "learning_rate": 3.6301369863013704e-06, + "loss": 1.6932, + "step": 228 + }, + { + "epoch": 0.2352336928608115, + "grad_norm": 2.476128578186035, + "learning_rate": 3.647260273972603e-06, + "loss": 1.7125, + "step": 229 + }, + { + "epoch": 0.23626091422701592, + "grad_norm": 2.3841593265533447, + "learning_rate": 3.664383561643836e-06, + "loss": 1.6469, + "step": 230 + }, + { + "epoch": 0.23626091422701592, + "eval_loss": 1.925522804260254, + "eval_runtime": 37.8581, + "eval_samples_per_second": 26.414, + "eval_steps_per_second": 1.109, + "step": 230 + }, + { + "epoch": 0.23728813559322035, + "grad_norm": 2.4401519298553467, + "learning_rate": 3.6815068493150687e-06, + "loss": 1.6726, + "step": 231 + }, + { + "epoch": 0.23831535695942477, + "grad_norm": 2.329674005508423, + "learning_rate": 3.6986301369863014e-06, + "loss": 1.6349, + "step": 232 + }, + { + "epoch": 0.23934257832562916, + "grad_norm": 2.654290199279785, + "learning_rate": 3.7157534246575344e-06, + "loss": 1.7309, + "step": 233 + }, + { + "epoch": 0.24036979969183359, + "grad_norm": 2.1599373817443848, + "learning_rate": 3.7328767123287675e-06, + "loss": 1.6437, + "step": 234 + }, + { + "epoch": 0.241397021058038, + "grad_norm": 2.6170263290405273, + "learning_rate": 3.7500000000000005e-06, + "loss": 1.6862, + "step": 235 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 2.2836415767669678, + "learning_rate": 3.767123287671233e-06, + "loss": 1.7073, + "step": 236 + }, + { + "epoch": 0.24345146379044685, + "grad_norm": 2.3001456260681152, + "learning_rate": 3.784246575342466e-06, + "loss": 1.6455, + "step": 237 + }, + { + "epoch": 0.24447868515665125, + "grad_norm": 2.4549131393432617, + "learning_rate": 3.8013698630136993e-06, + "loss": 1.6566, + "step": 238 + }, + { + "epoch": 0.24550590652285567, + "grad_norm": 2.4387831687927246, + "learning_rate": 3.818493150684932e-06, + "loss": 1.627, + "step": 239 + }, + { + "epoch": 0.2465331278890601, + "grad_norm": 2.136474847793579, + "learning_rate": 3.8356164383561645e-06, + "loss": 1.7286, + "step": 240 + }, + { + "epoch": 0.2465331278890601, + "eval_loss": 1.9188542366027832, + "eval_runtime": 37.9079, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 240 + }, + { + "epoch": 0.24756034925526452, + "grad_norm": 2.430297374725342, + "learning_rate": 3.852739726027397e-06, + "loss": 1.7014, + "step": 241 + }, + { + "epoch": 0.24858757062146894, + "grad_norm": 2.302480697631836, + "learning_rate": 3.869863013698631e-06, + "loss": 1.6412, + "step": 242 + }, + { + "epoch": 0.24961479198767333, + "grad_norm": 2.2905426025390625, + "learning_rate": 3.886986301369863e-06, + "loss": 1.6445, + "step": 243 + }, + { + "epoch": 0.25064201335387776, + "grad_norm": 2.55385684967041, + "learning_rate": 3.904109589041096e-06, + "loss": 1.6265, + "step": 244 + }, + { + "epoch": 0.2516692347200822, + "grad_norm": 2.3053576946258545, + "learning_rate": 3.921232876712329e-06, + "loss": 1.6338, + "step": 245 + }, + { + "epoch": 0.2526964560862866, + "grad_norm": 2.4894707202911377, + "learning_rate": 3.938356164383562e-06, + "loss": 1.5966, + "step": 246 + }, + { + "epoch": 0.253723677452491, + "grad_norm": 2.6198930740356445, + "learning_rate": 3.9554794520547955e-06, + "loss": 1.6935, + "step": 247 + }, + { + "epoch": 0.25475089881869545, + "grad_norm": 2.3548998832702637, + "learning_rate": 3.972602739726027e-06, + "loss": 1.6227, + "step": 248 + }, + { + "epoch": 0.25577812018489987, + "grad_norm": 2.535460948944092, + "learning_rate": 3.989726027397261e-06, + "loss": 1.6699, + "step": 249 + }, + { + "epoch": 0.25680534155110424, + "grad_norm": 2.3680567741394043, + "learning_rate": 4.006849315068493e-06, + "loss": 1.6666, + "step": 250 + }, + { + "epoch": 0.25680534155110424, + "eval_loss": 1.9149967432022095, + "eval_runtime": 37.9423, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 250 + }, + { + "epoch": 0.25783256291730866, + "grad_norm": 2.3914358615875244, + "learning_rate": 4.023972602739726e-06, + "loss": 1.6633, + "step": 251 + }, + { + "epoch": 0.2588597842835131, + "grad_norm": 2.4293038845062256, + "learning_rate": 4.0410958904109595e-06, + "loss": 1.6573, + "step": 252 + }, + { + "epoch": 0.2598870056497175, + "grad_norm": 2.3409900665283203, + "learning_rate": 4.058219178082192e-06, + "loss": 1.6815, + "step": 253 + }, + { + "epoch": 0.26091422701592193, + "grad_norm": 2.5807456970214844, + "learning_rate": 4.075342465753426e-06, + "loss": 1.6412, + "step": 254 + }, + { + "epoch": 0.26194144838212635, + "grad_norm": 2.317861795425415, + "learning_rate": 4.092465753424658e-06, + "loss": 1.6311, + "step": 255 + }, + { + "epoch": 0.2629686697483308, + "grad_norm": 2.231682777404785, + "learning_rate": 4.109589041095891e-06, + "loss": 1.7021, + "step": 256 + }, + { + "epoch": 0.2639958911145352, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.1267123287671235e-06, + "loss": 1.6446, + "step": 257 + }, + { + "epoch": 0.2650231124807396, + "grad_norm": 2.2619071006774902, + "learning_rate": 4.143835616438356e-06, + "loss": 1.7101, + "step": 258 + }, + { + "epoch": 0.26605033384694404, + "grad_norm": 2.461945056915283, + "learning_rate": 4.16095890410959e-06, + "loss": 1.6901, + "step": 259 + }, + { + "epoch": 0.2670775552131484, + "grad_norm": 2.516348123550415, + "learning_rate": 4.178082191780822e-06, + "loss": 1.644, + "step": 260 + }, + { + "epoch": 0.2670775552131484, + "eval_loss": 1.904897928237915, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 260 + }, + { + "epoch": 0.26810477657935283, + "grad_norm": 2.2311813831329346, + "learning_rate": 4.195205479452055e-06, + "loss": 1.6301, + "step": 261 + }, + { + "epoch": 0.26913199794555726, + "grad_norm": 2.605109214782715, + "learning_rate": 4.212328767123288e-06, + "loss": 1.6393, + "step": 262 + }, + { + "epoch": 0.2701592193117617, + "grad_norm": 2.6103134155273438, + "learning_rate": 4.229452054794521e-06, + "loss": 1.6468, + "step": 263 + }, + { + "epoch": 0.2711864406779661, + "grad_norm": 2.5538523197174072, + "learning_rate": 4.246575342465754e-06, + "loss": 1.6294, + "step": 264 + }, + { + "epoch": 0.2722136620441705, + "grad_norm": 2.5841965675354004, + "learning_rate": 4.263698630136986e-06, + "loss": 1.5837, + "step": 265 + }, + { + "epoch": 0.27324088341037495, + "grad_norm": 2.4604523181915283, + "learning_rate": 4.28082191780822e-06, + "loss": 1.7473, + "step": 266 + }, + { + "epoch": 0.27426810477657937, + "grad_norm": 2.6701765060424805, + "learning_rate": 4.297945205479452e-06, + "loss": 1.6539, + "step": 267 + }, + { + "epoch": 0.2752953261427838, + "grad_norm": 2.360069990158081, + "learning_rate": 4.315068493150685e-06, + "loss": 1.6385, + "step": 268 + }, + { + "epoch": 0.2763225475089882, + "grad_norm": 2.4922797679901123, + "learning_rate": 4.3321917808219185e-06, + "loss": 1.6586, + "step": 269 + }, + { + "epoch": 0.2773497688751926, + "grad_norm": 2.5443379878997803, + "learning_rate": 4.349315068493151e-06, + "loss": 1.6318, + "step": 270 + }, + { + "epoch": 0.2773497688751926, + "eval_loss": 1.9062609672546387, + "eval_runtime": 37.92, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 270 + }, + { + "epoch": 0.278376990241397, + "grad_norm": 2.4064481258392334, + "learning_rate": 4.3664383561643846e-06, + "loss": 1.6178, + "step": 271 + }, + { + "epoch": 0.2794042116076014, + "grad_norm": 2.681637763977051, + "learning_rate": 4.383561643835616e-06, + "loss": 1.7071, + "step": 272 + }, + { + "epoch": 0.28043143297380585, + "grad_norm": 2.5805184841156006, + "learning_rate": 4.40068493150685e-06, + "loss": 1.6553, + "step": 273 + }, + { + "epoch": 0.2814586543400103, + "grad_norm": 2.4169883728027344, + "learning_rate": 4.4178082191780825e-06, + "loss": 1.6249, + "step": 274 + }, + { + "epoch": 0.2824858757062147, + "grad_norm": 2.557570457458496, + "learning_rate": 4.434931506849315e-06, + "loss": 1.5932, + "step": 275 + }, + { + "epoch": 0.2835130970724191, + "grad_norm": 2.3808107376098633, + "learning_rate": 4.4520547945205486e-06, + "loss": 1.6377, + "step": 276 + }, + { + "epoch": 0.28454031843862354, + "grad_norm": 2.402404546737671, + "learning_rate": 4.469178082191781e-06, + "loss": 1.6285, + "step": 277 + }, + { + "epoch": 0.28556753980482796, + "grad_norm": 2.4676456451416016, + "learning_rate": 4.486301369863014e-06, + "loss": 1.6838, + "step": 278 + }, + { + "epoch": 0.28659476117103233, + "grad_norm": 2.659083366394043, + "learning_rate": 4.503424657534247e-06, + "loss": 1.5872, + "step": 279 + }, + { + "epoch": 0.28762198253723675, + "grad_norm": 2.326733350753784, + "learning_rate": 4.52054794520548e-06, + "loss": 1.6142, + "step": 280 + }, + { + "epoch": 0.28762198253723675, + "eval_loss": 1.904759407043457, + "eval_runtime": 37.9651, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 280 + }, + { + "epoch": 0.2886492039034412, + "grad_norm": 2.299504518508911, + "learning_rate": 4.537671232876713e-06, + "loss": 1.6673, + "step": 281 + }, + { + "epoch": 0.2896764252696456, + "grad_norm": 2.4932825565338135, + "learning_rate": 4.554794520547945e-06, + "loss": 1.6586, + "step": 282 + }, + { + "epoch": 0.29070364663585, + "grad_norm": 2.2599050998687744, + "learning_rate": 4.571917808219179e-06, + "loss": 1.6568, + "step": 283 + }, + { + "epoch": 0.29173086800205444, + "grad_norm": 2.435098886489868, + "learning_rate": 4.589041095890411e-06, + "loss": 1.6113, + "step": 284 + }, + { + "epoch": 0.29275808936825887, + "grad_norm": 2.539246082305908, + "learning_rate": 4.606164383561644e-06, + "loss": 1.6306, + "step": 285 + }, + { + "epoch": 0.2937853107344633, + "grad_norm": 2.596198320388794, + "learning_rate": 4.6232876712328774e-06, + "loss": 1.6473, + "step": 286 + }, + { + "epoch": 0.2948125321006677, + "grad_norm": 2.349133014678955, + "learning_rate": 4.64041095890411e-06, + "loss": 1.6042, + "step": 287 + }, + { + "epoch": 0.29583975346687214, + "grad_norm": 2.3572661876678467, + "learning_rate": 4.657534246575343e-06, + "loss": 1.6635, + "step": 288 + }, + { + "epoch": 0.2968669748330765, + "grad_norm": 2.4210307598114014, + "learning_rate": 4.674657534246575e-06, + "loss": 1.624, + "step": 289 + }, + { + "epoch": 0.2978941961992809, + "grad_norm": 2.427403688430786, + "learning_rate": 4.691780821917809e-06, + "loss": 1.6824, + "step": 290 + }, + { + "epoch": 0.2978941961992809, + "eval_loss": 1.8888994455337524, + "eval_runtime": 37.9777, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 290 + }, + { + "epoch": 0.29892141756548535, + "grad_norm": 2.534029722213745, + "learning_rate": 4.7089041095890414e-06, + "loss": 1.6456, + "step": 291 + }, + { + "epoch": 0.29994863893168977, + "grad_norm": 2.329305648803711, + "learning_rate": 4.726027397260274e-06, + "loss": 1.646, + "step": 292 + }, + { + "epoch": 0.3009758602978942, + "grad_norm": 2.561548948287964, + "learning_rate": 4.7431506849315075e-06, + "loss": 1.6511, + "step": 293 + }, + { + "epoch": 0.3020030816640986, + "grad_norm": 2.707712411880493, + "learning_rate": 4.76027397260274e-06, + "loss": 1.5984, + "step": 294 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 2.403089761734009, + "learning_rate": 4.777397260273973e-06, + "loss": 1.6164, + "step": 295 + }, + { + "epoch": 0.30405752439650746, + "grad_norm": 2.424910306930542, + "learning_rate": 4.7945205479452054e-06, + "loss": 1.6291, + "step": 296 + }, + { + "epoch": 0.3050847457627119, + "grad_norm": 2.3474302291870117, + "learning_rate": 4.811643835616439e-06, + "loss": 1.553, + "step": 297 + }, + { + "epoch": 0.3061119671289163, + "grad_norm": 2.2211227416992188, + "learning_rate": 4.8287671232876716e-06, + "loss": 1.6108, + "step": 298 + }, + { + "epoch": 0.3071391884951207, + "grad_norm": 2.495983600616455, + "learning_rate": 4.845890410958904e-06, + "loss": 1.6135, + "step": 299 + }, + { + "epoch": 0.3081664098613251, + "grad_norm": 2.525495767593384, + "learning_rate": 4.863013698630138e-06, + "loss": 1.6061, + "step": 300 + }, + { + "epoch": 0.3081664098613251, + "eval_loss": 1.8982168436050415, + "eval_runtime": 37.9848, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 300 + }, + { + "epoch": 0.3091936312275295, + "grad_norm": 2.449064254760742, + "learning_rate": 4.88013698630137e-06, + "loss": 1.5827, + "step": 301 + }, + { + "epoch": 0.31022085259373394, + "grad_norm": 2.288095712661743, + "learning_rate": 4.897260273972603e-06, + "loss": 1.704, + "step": 302 + }, + { + "epoch": 0.31124807395993837, + "grad_norm": 2.5301458835601807, + "learning_rate": 4.914383561643836e-06, + "loss": 1.6342, + "step": 303 + }, + { + "epoch": 0.3122752953261428, + "grad_norm": 2.2793986797332764, + "learning_rate": 4.931506849315069e-06, + "loss": 1.6688, + "step": 304 + }, + { + "epoch": 0.3133025166923472, + "grad_norm": 2.38783597946167, + "learning_rate": 4.948630136986302e-06, + "loss": 1.7014, + "step": 305 + }, + { + "epoch": 0.31432973805855163, + "grad_norm": 2.5224838256835938, + "learning_rate": 4.965753424657534e-06, + "loss": 1.5812, + "step": 306 + }, + { + "epoch": 0.31535695942475606, + "grad_norm": 2.434432029724121, + "learning_rate": 4.982876712328768e-06, + "loss": 1.6186, + "step": 307 + }, + { + "epoch": 0.3163841807909605, + "grad_norm": 2.3708460330963135, + "learning_rate": 5e-06, + "loss": 1.5669, + "step": 308 + }, + { + "epoch": 0.31741140215716485, + "grad_norm": 2.5524990558624268, + "learning_rate": 4.998096688237534e-06, + "loss": 1.6688, + "step": 309 + }, + { + "epoch": 0.31843862352336927, + "grad_norm": 2.322441577911377, + "learning_rate": 4.996193376475067e-06, + "loss": 1.5571, + "step": 310 + }, + { + "epoch": 0.31843862352336927, + "eval_loss": 1.8914265632629395, + "eval_runtime": 37.6505, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 310 + }, + { + "epoch": 0.3194658448895737, + "grad_norm": 2.5248687267303467, + "learning_rate": 4.9942900647126005e-06, + "loss": 1.5585, + "step": 311 + }, + { + "epoch": 0.3204930662557781, + "grad_norm": 2.2464914321899414, + "learning_rate": 4.992386752950134e-06, + "loss": 1.5921, + "step": 312 + }, + { + "epoch": 0.32152028762198254, + "grad_norm": 2.5355331897735596, + "learning_rate": 4.990483441187667e-06, + "loss": 1.6263, + "step": 313 + }, + { + "epoch": 0.32254750898818696, + "grad_norm": 2.5611143112182617, + "learning_rate": 4.9885801294252e-06, + "loss": 1.6267, + "step": 314 + }, + { + "epoch": 0.3235747303543914, + "grad_norm": 2.364356517791748, + "learning_rate": 4.986676817662733e-06, + "loss": 1.5847, + "step": 315 + }, + { + "epoch": 0.3246019517205958, + "grad_norm": 2.455165147781372, + "learning_rate": 4.9847735059002665e-06, + "loss": 1.661, + "step": 316 + }, + { + "epoch": 0.3256291730868002, + "grad_norm": 2.4312195777893066, + "learning_rate": 4.9828701941378e-06, + "loss": 1.5995, + "step": 317 + }, + { + "epoch": 0.3266563944530046, + "grad_norm": 2.4081499576568604, + "learning_rate": 4.980966882375333e-06, + "loss": 1.5756, + "step": 318 + }, + { + "epoch": 0.327683615819209, + "grad_norm": 2.2608511447906494, + "learning_rate": 4.979063570612867e-06, + "loss": 1.543, + "step": 319 + }, + { + "epoch": 0.32871083718541344, + "grad_norm": 2.6551144123077393, + "learning_rate": 4.9771602588504e-06, + "loss": 1.6467, + "step": 320 + }, + { + "epoch": 0.32871083718541344, + "eval_loss": 1.89058518409729, + "eval_runtime": 37.8192, + "eval_samples_per_second": 26.442, + "eval_steps_per_second": 1.111, + "step": 320 + }, + { + "epoch": 0.32973805855161786, + "grad_norm": 2.3897273540496826, + "learning_rate": 4.975256947087933e-06, + "loss": 1.5607, + "step": 321 + }, + { + "epoch": 0.3307652799178223, + "grad_norm": 2.3869469165802, + "learning_rate": 4.973353635325467e-06, + "loss": 1.6079, + "step": 322 + }, + { + "epoch": 0.3317925012840267, + "grad_norm": 2.159193515777588, + "learning_rate": 4.971450323563e-06, + "loss": 1.5475, + "step": 323 + }, + { + "epoch": 0.33281972265023113, + "grad_norm": 2.3069653511047363, + "learning_rate": 4.969547011800533e-06, + "loss": 1.616, + "step": 324 + }, + { + "epoch": 0.33384694401643555, + "grad_norm": 2.3131422996520996, + "learning_rate": 4.967643700038067e-06, + "loss": 1.6106, + "step": 325 + }, + { + "epoch": 0.33487416538264, + "grad_norm": 2.4398200511932373, + "learning_rate": 4.9657403882756e-06, + "loss": 1.5779, + "step": 326 + }, + { + "epoch": 0.3359013867488444, + "grad_norm": 2.278024435043335, + "learning_rate": 4.9638370765131335e-06, + "loss": 1.5724, + "step": 327 + }, + { + "epoch": 0.33692860811504877, + "grad_norm": 2.4989173412323, + "learning_rate": 4.961933764750667e-06, + "loss": 1.6337, + "step": 328 + }, + { + "epoch": 0.3379558294812532, + "grad_norm": 2.46319580078125, + "learning_rate": 4.9600304529882e-06, + "loss": 1.6519, + "step": 329 + }, + { + "epoch": 0.3389830508474576, + "grad_norm": 2.2703049182891846, + "learning_rate": 4.9581271412257335e-06, + "loss": 1.6081, + "step": 330 + }, + { + "epoch": 0.3389830508474576, + "eval_loss": 1.8755545616149902, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 330 + }, + { + "epoch": 0.34001027221366203, + "grad_norm": 2.502196788787842, + "learning_rate": 4.956223829463266e-06, + "loss": 1.5221, + "step": 331 + }, + { + "epoch": 0.34103749357986646, + "grad_norm": 2.284280776977539, + "learning_rate": 4.954320517700799e-06, + "loss": 1.6268, + "step": 332 + }, + { + "epoch": 0.3420647149460709, + "grad_norm": 2.3551881313323975, + "learning_rate": 4.952417205938333e-06, + "loss": 1.6078, + "step": 333 + }, + { + "epoch": 0.3430919363122753, + "grad_norm": 2.343780755996704, + "learning_rate": 4.950513894175866e-06, + "loss": 1.591, + "step": 334 + }, + { + "epoch": 0.3441191576784797, + "grad_norm": 2.3524727821350098, + "learning_rate": 4.9486105824133995e-06, + "loss": 1.605, + "step": 335 + }, + { + "epoch": 0.34514637904468415, + "grad_norm": 2.349896192550659, + "learning_rate": 4.946707270650933e-06, + "loss": 1.6106, + "step": 336 + }, + { + "epoch": 0.34617360041088857, + "grad_norm": 2.5154407024383545, + "learning_rate": 4.944803958888466e-06, + "loss": 1.6429, + "step": 337 + }, + { + "epoch": 0.34720082177709294, + "grad_norm": 2.2692947387695312, + "learning_rate": 4.942900647126e-06, + "loss": 1.5865, + "step": 338 + }, + { + "epoch": 0.34822804314329736, + "grad_norm": 2.2672805786132812, + "learning_rate": 4.940997335363533e-06, + "loss": 1.5464, + "step": 339 + }, + { + "epoch": 0.3492552645095018, + "grad_norm": 2.5147039890289307, + "learning_rate": 4.939094023601066e-06, + "loss": 1.5837, + "step": 340 + }, + { + "epoch": 0.3492552645095018, + "eval_loss": 1.8672455549240112, + "eval_runtime": 37.9644, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 340 + }, + { + "epoch": 0.3502824858757062, + "grad_norm": 2.3416433334350586, + "learning_rate": 4.9371907118386e-06, + "loss": 1.5937, + "step": 341 + }, + { + "epoch": 0.35130970724191063, + "grad_norm": 2.3044960498809814, + "learning_rate": 4.935287400076133e-06, + "loss": 1.6132, + "step": 342 + }, + { + "epoch": 0.35233692860811505, + "grad_norm": 2.4360430240631104, + "learning_rate": 4.933384088313666e-06, + "loss": 1.5731, + "step": 343 + }, + { + "epoch": 0.3533641499743195, + "grad_norm": 2.392050266265869, + "learning_rate": 4.9314807765512e-06, + "loss": 1.5098, + "step": 344 + }, + { + "epoch": 0.3543913713405239, + "grad_norm": 2.3815808296203613, + "learning_rate": 4.929577464788733e-06, + "loss": 1.5827, + "step": 345 + }, + { + "epoch": 0.3554185927067283, + "grad_norm": 2.423077344894409, + "learning_rate": 4.9276741530262665e-06, + "loss": 1.5857, + "step": 346 + }, + { + "epoch": 0.35644581407293274, + "grad_norm": 2.5535898208618164, + "learning_rate": 4.9257708412638e-06, + "loss": 1.5884, + "step": 347 + }, + { + "epoch": 0.3574730354391371, + "grad_norm": 2.2665445804595947, + "learning_rate": 4.923867529501332e-06, + "loss": 1.5561, + "step": 348 + }, + { + "epoch": 0.35850025680534153, + "grad_norm": 2.3357481956481934, + "learning_rate": 4.921964217738866e-06, + "loss": 1.6444, + "step": 349 + }, + { + "epoch": 0.35952747817154596, + "grad_norm": 2.4537885189056396, + "learning_rate": 4.920060905976399e-06, + "loss": 1.6262, + "step": 350 + }, + { + "epoch": 0.35952747817154596, + "eval_loss": 1.8621584177017212, + "eval_runtime": 38.0027, + "eval_samples_per_second": 26.314, + "eval_steps_per_second": 1.105, + "step": 350 + }, + { + "epoch": 0.3605546995377504, + "grad_norm": 2.379754066467285, + "learning_rate": 4.9181575942139325e-06, + "loss": 1.6227, + "step": 351 + }, + { + "epoch": 0.3615819209039548, + "grad_norm": 2.432905435562134, + "learning_rate": 4.916254282451466e-06, + "loss": 1.6554, + "step": 352 + }, + { + "epoch": 0.3626091422701592, + "grad_norm": 2.34236216545105, + "learning_rate": 4.914350970688999e-06, + "loss": 1.5613, + "step": 353 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.3019444942474365, + "learning_rate": 4.9124476589265325e-06, + "loss": 1.5535, + "step": 354 + }, + { + "epoch": 0.36466358500256807, + "grad_norm": 2.3354644775390625, + "learning_rate": 4.910544347164066e-06, + "loss": 1.6384, + "step": 355 + }, + { + "epoch": 0.3656908063687725, + "grad_norm": 2.5154788494110107, + "learning_rate": 4.908641035401599e-06, + "loss": 1.6986, + "step": 356 + }, + { + "epoch": 0.3667180277349769, + "grad_norm": 2.2707178592681885, + "learning_rate": 4.906737723639133e-06, + "loss": 1.5705, + "step": 357 + }, + { + "epoch": 0.3677452491011813, + "grad_norm": 2.3058297634124756, + "learning_rate": 4.904834411876665e-06, + "loss": 1.6033, + "step": 358 + }, + { + "epoch": 0.3687724704673857, + "grad_norm": 2.3923370838165283, + "learning_rate": 4.902931100114199e-06, + "loss": 1.5333, + "step": 359 + }, + { + "epoch": 0.3697996918335901, + "grad_norm": 2.4023196697235107, + "learning_rate": 4.901027788351733e-06, + "loss": 1.5775, + "step": 360 + }, + { + "epoch": 0.3697996918335901, + "eval_loss": 1.867856740951538, + "eval_runtime": 38.0125, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 360 + }, + { + "epoch": 0.37082691319979455, + "grad_norm": 2.4903581142425537, + "learning_rate": 4.899124476589266e-06, + "loss": 1.5491, + "step": 361 + }, + { + "epoch": 0.371854134565999, + "grad_norm": 2.556683301925659, + "learning_rate": 4.8972211648267994e-06, + "loss": 1.5967, + "step": 362 + }, + { + "epoch": 0.3728813559322034, + "grad_norm": 2.3017377853393555, + "learning_rate": 4.895317853064333e-06, + "loss": 1.5661, + "step": 363 + }, + { + "epoch": 0.3739085772984078, + "grad_norm": 2.286191940307617, + "learning_rate": 4.893414541301866e-06, + "loss": 1.5679, + "step": 364 + }, + { + "epoch": 0.37493579866461224, + "grad_norm": 2.260812520980835, + "learning_rate": 4.891511229539399e-06, + "loss": 1.5965, + "step": 365 + }, + { + "epoch": 0.37596302003081666, + "grad_norm": 2.4569408893585205, + "learning_rate": 4.889607917776932e-06, + "loss": 1.5923, + "step": 366 + }, + { + "epoch": 0.37699024139702103, + "grad_norm": 2.571552038192749, + "learning_rate": 4.887704606014465e-06, + "loss": 1.6021, + "step": 367 + }, + { + "epoch": 0.37801746276322545, + "grad_norm": 2.287754774093628, + "learning_rate": 4.885801294251999e-06, + "loss": 1.5304, + "step": 368 + }, + { + "epoch": 0.3790446841294299, + "grad_norm": 2.317784070968628, + "learning_rate": 4.883897982489532e-06, + "loss": 1.6233, + "step": 369 + }, + { + "epoch": 0.3800719054956343, + "grad_norm": 2.290601968765259, + "learning_rate": 4.8819946707270655e-06, + "loss": 1.624, + "step": 370 + }, + { + "epoch": 0.3800719054956343, + "eval_loss": 1.8604528903961182, + "eval_runtime": 38.0176, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 370 + }, + { + "epoch": 0.3810991268618387, + "grad_norm": 2.28519868850708, + "learning_rate": 4.880091358964599e-06, + "loss": 1.5286, + "step": 371 + }, + { + "epoch": 0.38212634822804314, + "grad_norm": 2.271810293197632, + "learning_rate": 4.878188047202132e-06, + "loss": 1.6879, + "step": 372 + }, + { + "epoch": 0.38315356959424757, + "grad_norm": 2.391848564147949, + "learning_rate": 4.876284735439666e-06, + "loss": 1.6458, + "step": 373 + }, + { + "epoch": 0.384180790960452, + "grad_norm": 2.3787031173706055, + "learning_rate": 4.874381423677199e-06, + "loss": 1.5738, + "step": 374 + }, + { + "epoch": 0.3852080123266564, + "grad_norm": 2.404219150543213, + "learning_rate": 4.8724781119147315e-06, + "loss": 1.6184, + "step": 375 + }, + { + "epoch": 0.38623523369286084, + "grad_norm": 2.372661590576172, + "learning_rate": 4.870574800152265e-06, + "loss": 1.5316, + "step": 376 + }, + { + "epoch": 0.3872624550590652, + "grad_norm": 2.4397103786468506, + "learning_rate": 4.868671488389798e-06, + "loss": 1.5605, + "step": 377 + }, + { + "epoch": 0.3882896764252696, + "grad_norm": 2.371006727218628, + "learning_rate": 4.8667681766273316e-06, + "loss": 1.5184, + "step": 378 + }, + { + "epoch": 0.38931689779147405, + "grad_norm": 2.2621910572052, + "learning_rate": 4.864864864864866e-06, + "loss": 1.5788, + "step": 379 + }, + { + "epoch": 0.39034411915767847, + "grad_norm": 2.2888615131378174, + "learning_rate": 4.862961553102399e-06, + "loss": 1.6244, + "step": 380 + }, + { + "epoch": 0.39034411915767847, + "eval_loss": 1.8586912155151367, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 380 + }, + { + "epoch": 0.3913713405238829, + "grad_norm": 2.4439220428466797, + "learning_rate": 4.8610582413399325e-06, + "loss": 1.5749, + "step": 381 + }, + { + "epoch": 0.3923985618900873, + "grad_norm": 2.2423763275146484, + "learning_rate": 4.859154929577465e-06, + "loss": 1.5738, + "step": 382 + }, + { + "epoch": 0.39342578325629174, + "grad_norm": 2.267467737197876, + "learning_rate": 4.857251617814998e-06, + "loss": 1.6212, + "step": 383 + }, + { + "epoch": 0.39445300462249616, + "grad_norm": 2.3680620193481445, + "learning_rate": 4.855348306052532e-06, + "loss": 1.6129, + "step": 384 + }, + { + "epoch": 0.3954802259887006, + "grad_norm": 2.3782172203063965, + "learning_rate": 4.853444994290065e-06, + "loss": 1.5605, + "step": 385 + }, + { + "epoch": 0.396507447354905, + "grad_norm": 2.5745720863342285, + "learning_rate": 4.8515416825275985e-06, + "loss": 1.634, + "step": 386 + }, + { + "epoch": 0.3975346687211094, + "grad_norm": 2.38212513923645, + "learning_rate": 4.849638370765132e-06, + "loss": 1.5671, + "step": 387 + }, + { + "epoch": 0.3985618900873138, + "grad_norm": 2.3395450115203857, + "learning_rate": 4.847735059002665e-06, + "loss": 1.5088, + "step": 388 + }, + { + "epoch": 0.3995891114535182, + "grad_norm": 2.474440574645996, + "learning_rate": 4.8458317472401985e-06, + "loss": 1.5997, + "step": 389 + }, + { + "epoch": 0.40061633281972264, + "grad_norm": 2.429133176803589, + "learning_rate": 4.843928435477732e-06, + "loss": 1.5611, + "step": 390 + }, + { + "epoch": 0.40061633281972264, + "eval_loss": 1.8613603115081787, + "eval_runtime": 37.9862, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 390 + }, + { + "epoch": 0.40164355418592707, + "grad_norm": 2.403207778930664, + "learning_rate": 4.842025123715265e-06, + "loss": 1.5976, + "step": 391 + }, + { + "epoch": 0.4026707755521315, + "grad_norm": 2.383018732070923, + "learning_rate": 4.840121811952798e-06, + "loss": 1.5791, + "step": 392 + }, + { + "epoch": 0.4036979969183359, + "grad_norm": 2.5303306579589844, + "learning_rate": 4.838218500190331e-06, + "loss": 1.6035, + "step": 393 + }, + { + "epoch": 0.40472521828454033, + "grad_norm": 2.391075372695923, + "learning_rate": 4.8363151884278645e-06, + "loss": 1.5795, + "step": 394 + }, + { + "epoch": 0.40575243965074476, + "grad_norm": 2.396934747695923, + "learning_rate": 4.834411876665398e-06, + "loss": 1.5965, + "step": 395 + }, + { + "epoch": 0.4067796610169492, + "grad_norm": 2.4631423950195312, + "learning_rate": 4.832508564902931e-06, + "loss": 1.5372, + "step": 396 + }, + { + "epoch": 0.40780688238315355, + "grad_norm": 2.3457326889038086, + "learning_rate": 4.830605253140465e-06, + "loss": 1.5729, + "step": 397 + }, + { + "epoch": 0.40883410374935797, + "grad_norm": 2.3763277530670166, + "learning_rate": 4.828701941377998e-06, + "loss": 1.5309, + "step": 398 + }, + { + "epoch": 0.4098613251155624, + "grad_norm": 2.442439556121826, + "learning_rate": 4.826798629615531e-06, + "loss": 1.5465, + "step": 399 + }, + { + "epoch": 0.4108885464817668, + "grad_norm": 2.3464581966400146, + "learning_rate": 4.824895317853065e-06, + "loss": 1.5665, + "step": 400 + }, + { + "epoch": 0.4108885464817668, + "eval_loss": 1.8671027421951294, + "eval_runtime": 38.0166, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 400 + }, + { + "epoch": 0.41191576784797124, + "grad_norm": 2.5441887378692627, + "learning_rate": 4.822992006090598e-06, + "loss": 1.5232, + "step": 401 + }, + { + "epoch": 0.41294298921417566, + "grad_norm": 2.4234976768493652, + "learning_rate": 4.821088694328131e-06, + "loss": 1.5569, + "step": 402 + }, + { + "epoch": 0.4139702105803801, + "grad_norm": 2.3603928089141846, + "learning_rate": 4.819185382565665e-06, + "loss": 1.612, + "step": 403 + }, + { + "epoch": 0.4149974319465845, + "grad_norm": 2.337862491607666, + "learning_rate": 4.817282070803198e-06, + "loss": 1.5882, + "step": 404 + }, + { + "epoch": 0.41602465331278893, + "grad_norm": 2.3805742263793945, + "learning_rate": 4.8153787590407315e-06, + "loss": 1.5204, + "step": 405 + }, + { + "epoch": 0.41705187467899335, + "grad_norm": 2.3335204124450684, + "learning_rate": 4.813475447278265e-06, + "loss": 1.6041, + "step": 406 + }, + { + "epoch": 0.4180790960451977, + "grad_norm": 2.406137466430664, + "learning_rate": 4.811572135515798e-06, + "loss": 1.5348, + "step": 407 + }, + { + "epoch": 0.41910631741140214, + "grad_norm": 2.2823917865753174, + "learning_rate": 4.809668823753332e-06, + "loss": 1.5351, + "step": 408 + }, + { + "epoch": 0.42013353877760656, + "grad_norm": 2.3577091693878174, + "learning_rate": 4.807765511990864e-06, + "loss": 1.5733, + "step": 409 + }, + { + "epoch": 0.421160760143811, + "grad_norm": 2.536691665649414, + "learning_rate": 4.8058622002283975e-06, + "loss": 1.6153, + "step": 410 + }, + { + "epoch": 0.421160760143811, + "eval_loss": 1.8613901138305664, + "eval_runtime": 37.6689, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 410 + }, + { + "epoch": 0.4221879815100154, + "grad_norm": 2.4020118713378906, + "learning_rate": 4.803958888465931e-06, + "loss": 1.5722, + "step": 411 + }, + { + "epoch": 0.42321520287621983, + "grad_norm": 2.5284810066223145, + "learning_rate": 4.802055576703464e-06, + "loss": 1.6546, + "step": 412 + }, + { + "epoch": 0.42424242424242425, + "grad_norm": 2.270094156265259, + "learning_rate": 4.8001522649409976e-06, + "loss": 1.5345, + "step": 413 + }, + { + "epoch": 0.4252696456086287, + "grad_norm": 2.241417407989502, + "learning_rate": 4.798248953178531e-06, + "loss": 1.5167, + "step": 414 + }, + { + "epoch": 0.4262968669748331, + "grad_norm": 2.3733017444610596, + "learning_rate": 4.796345641416064e-06, + "loss": 1.5532, + "step": 415 + }, + { + "epoch": 0.42732408834103747, + "grad_norm": 2.452587842941284, + "learning_rate": 4.794442329653598e-06, + "loss": 1.6136, + "step": 416 + }, + { + "epoch": 0.4283513097072419, + "grad_norm": 2.417217969894409, + "learning_rate": 4.792539017891131e-06, + "loss": 1.5398, + "step": 417 + }, + { + "epoch": 0.4293785310734463, + "grad_norm": 2.4502453804016113, + "learning_rate": 4.790635706128664e-06, + "loss": 1.5741, + "step": 418 + }, + { + "epoch": 0.43040575243965074, + "grad_norm": 2.2996811866760254, + "learning_rate": 4.788732394366197e-06, + "loss": 1.511, + "step": 419 + }, + { + "epoch": 0.43143297380585516, + "grad_norm": 2.3573555946350098, + "learning_rate": 4.786829082603731e-06, + "loss": 1.5117, + "step": 420 + }, + { + "epoch": 0.43143297380585516, + "eval_loss": 1.8690649271011353, + "eval_runtime": 37.7813, + "eval_samples_per_second": 26.468, + "eval_steps_per_second": 1.112, + "step": 420 + }, + { + "epoch": 0.4324601951720596, + "grad_norm": 2.37233567237854, + "learning_rate": 4.7849257708412645e-06, + "loss": 1.5873, + "step": 421 + }, + { + "epoch": 0.433487416538264, + "grad_norm": 2.146909713745117, + "learning_rate": 4.783022459078798e-06, + "loss": 1.5291, + "step": 422 + }, + { + "epoch": 0.4345146379044684, + "grad_norm": 2.5237205028533936, + "learning_rate": 4.781119147316331e-06, + "loss": 1.5545, + "step": 423 + }, + { + "epoch": 0.43554185927067285, + "grad_norm": 2.356842279434204, + "learning_rate": 4.7792158355538645e-06, + "loss": 1.5635, + "step": 424 + }, + { + "epoch": 0.43656908063687727, + "grad_norm": 2.1963610649108887, + "learning_rate": 4.777312523791398e-06, + "loss": 1.5355, + "step": 425 + }, + { + "epoch": 0.43759630200308164, + "grad_norm": 2.6425886154174805, + "learning_rate": 4.77540921202893e-06, + "loss": 1.5803, + "step": 426 + }, + { + "epoch": 0.43862352336928606, + "grad_norm": 2.4072272777557373, + "learning_rate": 4.773505900266464e-06, + "loss": 1.5649, + "step": 427 + }, + { + "epoch": 0.4396507447354905, + "grad_norm": 2.592374801635742, + "learning_rate": 4.771602588503997e-06, + "loss": 1.5693, + "step": 428 + }, + { + "epoch": 0.4406779661016949, + "grad_norm": 2.4534225463867188, + "learning_rate": 4.7696992767415305e-06, + "loss": 1.591, + "step": 429 + }, + { + "epoch": 0.44170518746789933, + "grad_norm": 2.434457778930664, + "learning_rate": 4.767795964979064e-06, + "loss": 1.5991, + "step": 430 + }, + { + "epoch": 0.44170518746789933, + "eval_loss": 1.8584965467453003, + "eval_runtime": 37.8854, + "eval_samples_per_second": 26.395, + "eval_steps_per_second": 1.109, + "step": 430 + }, + { + "epoch": 0.44273240883410375, + "grad_norm": 2.3150746822357178, + "learning_rate": 4.765892653216597e-06, + "loss": 1.6032, + "step": 431 + }, + { + "epoch": 0.4437596302003082, + "grad_norm": 2.3040223121643066, + "learning_rate": 4.763989341454131e-06, + "loss": 1.5115, + "step": 432 + }, + { + "epoch": 0.4447868515665126, + "grad_norm": 2.413007974624634, + "learning_rate": 4.762086029691664e-06, + "loss": 1.5305, + "step": 433 + }, + { + "epoch": 0.445814072932717, + "grad_norm": 2.526289463043213, + "learning_rate": 4.760182717929197e-06, + "loss": 1.564, + "step": 434 + }, + { + "epoch": 0.44684129429892144, + "grad_norm": 2.4122776985168457, + "learning_rate": 4.758279406166731e-06, + "loss": 1.4918, + "step": 435 + }, + { + "epoch": 0.4478685156651258, + "grad_norm": 2.474274158477783, + "learning_rate": 4.756376094404263e-06, + "loss": 1.5583, + "step": 436 + }, + { + "epoch": 0.44889573703133023, + "grad_norm": 2.5322892665863037, + "learning_rate": 4.7544727826417966e-06, + "loss": 1.6081, + "step": 437 + }, + { + "epoch": 0.44992295839753466, + "grad_norm": 2.391660690307617, + "learning_rate": 4.75256947087933e-06, + "loss": 1.5628, + "step": 438 + }, + { + "epoch": 0.4509501797637391, + "grad_norm": 2.426671028137207, + "learning_rate": 4.750666159116863e-06, + "loss": 1.531, + "step": 439 + }, + { + "epoch": 0.4519774011299435, + "grad_norm": 2.7095694541931152, + "learning_rate": 4.7487628473543975e-06, + "loss": 1.4811, + "step": 440 + }, + { + "epoch": 0.4519774011299435, + "eval_loss": 1.8348888158798218, + "eval_runtime": 37.9114, + "eval_samples_per_second": 26.377, + "eval_steps_per_second": 1.108, + "step": 440 + }, + { + "epoch": 0.4530046224961479, + "grad_norm": 2.5425240993499756, + "learning_rate": 4.746859535591931e-06, + "loss": 1.5453, + "step": 441 + }, + { + "epoch": 0.45403184386235235, + "grad_norm": 2.6848669052124023, + "learning_rate": 4.744956223829464e-06, + "loss": 1.526, + "step": 442 + }, + { + "epoch": 0.45505906522855677, + "grad_norm": 2.634068489074707, + "learning_rate": 4.743052912066997e-06, + "loss": 1.5748, + "step": 443 + }, + { + "epoch": 0.4560862865947612, + "grad_norm": 2.492030382156372, + "learning_rate": 4.74114960030453e-06, + "loss": 1.5858, + "step": 444 + }, + { + "epoch": 0.4571135079609656, + "grad_norm": 2.787740707397461, + "learning_rate": 4.7392462885420635e-06, + "loss": 1.4774, + "step": 445 + }, + { + "epoch": 0.45814072932717, + "grad_norm": 2.395914077758789, + "learning_rate": 4.737342976779597e-06, + "loss": 1.558, + "step": 446 + }, + { + "epoch": 0.4591679506933744, + "grad_norm": 2.6642563343048096, + "learning_rate": 4.73543966501713e-06, + "loss": 1.5558, + "step": 447 + }, + { + "epoch": 0.4601951720595788, + "grad_norm": 2.619943857192993, + "learning_rate": 4.7335363532546636e-06, + "loss": 1.5442, + "step": 448 + }, + { + "epoch": 0.46122239342578325, + "grad_norm": 2.355133056640625, + "learning_rate": 4.731633041492197e-06, + "loss": 1.5613, + "step": 449 + }, + { + "epoch": 0.4622496147919877, + "grad_norm": 2.6572225093841553, + "learning_rate": 4.72972972972973e-06, + "loss": 1.5951, + "step": 450 + }, + { + "epoch": 0.4622496147919877, + "eval_loss": 1.83199143409729, + "eval_runtime": 37.8996, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 450 + }, + { + "epoch": 0.4632768361581921, + "grad_norm": 2.692873001098633, + "learning_rate": 4.727826417967264e-06, + "loss": 1.5531, + "step": 451 + }, + { + "epoch": 0.4643040575243965, + "grad_norm": 2.3903489112854004, + "learning_rate": 4.725923106204797e-06, + "loss": 1.5411, + "step": 452 + }, + { + "epoch": 0.46533127889060094, + "grad_norm": 2.453263521194458, + "learning_rate": 4.72401979444233e-06, + "loss": 1.5632, + "step": 453 + }, + { + "epoch": 0.46635850025680536, + "grad_norm": 2.5637760162353516, + "learning_rate": 4.722116482679863e-06, + "loss": 1.5629, + "step": 454 + }, + { + "epoch": 0.46738572162300973, + "grad_norm": 2.387033700942993, + "learning_rate": 4.720213170917396e-06, + "loss": 1.5786, + "step": 455 + }, + { + "epoch": 0.46841294298921415, + "grad_norm": 2.4563071727752686, + "learning_rate": 4.71830985915493e-06, + "loss": 1.5321, + "step": 456 + }, + { + "epoch": 0.4694401643554186, + "grad_norm": 2.607451915740967, + "learning_rate": 4.716406547392463e-06, + "loss": 1.5958, + "step": 457 + }, + { + "epoch": 0.470467385721623, + "grad_norm": 2.4846854209899902, + "learning_rate": 4.714503235629996e-06, + "loss": 1.5205, + "step": 458 + }, + { + "epoch": 0.4714946070878274, + "grad_norm": 2.377629041671753, + "learning_rate": 4.71259992386753e-06, + "loss": 1.5403, + "step": 459 + }, + { + "epoch": 0.47252182845403184, + "grad_norm": 2.4470160007476807, + "learning_rate": 4.710696612105063e-06, + "loss": 1.5456, + "step": 460 + }, + { + "epoch": 0.47252182845403184, + "eval_loss": 1.83836829662323, + "eval_runtime": 37.88, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 460 + }, + { + "epoch": 0.47354904982023627, + "grad_norm": 2.383418560028076, + "learning_rate": 4.708793300342596e-06, + "loss": 1.5468, + "step": 461 + }, + { + "epoch": 0.4745762711864407, + "grad_norm": 2.3672895431518555, + "learning_rate": 4.70688998858013e-06, + "loss": 1.5313, + "step": 462 + }, + { + "epoch": 0.4756034925526451, + "grad_norm": 2.415717124938965, + "learning_rate": 4.704986676817663e-06, + "loss": 1.5645, + "step": 463 + }, + { + "epoch": 0.47663071391884954, + "grad_norm": 2.4690067768096924, + "learning_rate": 4.7030833650551965e-06, + "loss": 1.5285, + "step": 464 + }, + { + "epoch": 0.4776579352850539, + "grad_norm": 2.3845672607421875, + "learning_rate": 4.70118005329273e-06, + "loss": 1.596, + "step": 465 + }, + { + "epoch": 0.4786851566512583, + "grad_norm": 2.2801802158355713, + "learning_rate": 4.699276741530263e-06, + "loss": 1.5777, + "step": 466 + }, + { + "epoch": 0.47971237801746275, + "grad_norm": 2.4668991565704346, + "learning_rate": 4.697373429767797e-06, + "loss": 1.5566, + "step": 467 + }, + { + "epoch": 0.48073959938366717, + "grad_norm": 2.446078062057495, + "learning_rate": 4.69547011800533e-06, + "loss": 1.5664, + "step": 468 + }, + { + "epoch": 0.4817668207498716, + "grad_norm": 2.2310681343078613, + "learning_rate": 4.693566806242863e-06, + "loss": 1.5425, + "step": 469 + }, + { + "epoch": 0.482794042116076, + "grad_norm": 2.3938229084014893, + "learning_rate": 4.691663494480397e-06, + "loss": 1.576, + "step": 470 + }, + { + "epoch": 0.482794042116076, + "eval_loss": 1.83778977394104, + "eval_runtime": 37.9276, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 470 + }, + { + "epoch": 0.48382126348228044, + "grad_norm": 2.23827862739563, + "learning_rate": 4.689760182717929e-06, + "loss": 1.4763, + "step": 471 + }, + { + "epoch": 0.48484848484848486, + "grad_norm": 2.3148930072784424, + "learning_rate": 4.6878568709554626e-06, + "loss": 1.5386, + "step": 472 + }, + { + "epoch": 0.4858757062146893, + "grad_norm": 2.407036781311035, + "learning_rate": 4.685953559192996e-06, + "loss": 1.5576, + "step": 473 + }, + { + "epoch": 0.4869029275808937, + "grad_norm": 2.4675374031066895, + "learning_rate": 4.684050247430529e-06, + "loss": 1.4931, + "step": 474 + }, + { + "epoch": 0.4879301489470981, + "grad_norm": 2.4593775272369385, + "learning_rate": 4.682146935668063e-06, + "loss": 1.5348, + "step": 475 + }, + { + "epoch": 0.4889573703133025, + "grad_norm": 2.4169375896453857, + "learning_rate": 4.680243623905596e-06, + "loss": 1.5614, + "step": 476 + }, + { + "epoch": 0.4899845916795069, + "grad_norm": 2.6177303791046143, + "learning_rate": 4.678340312143129e-06, + "loss": 1.5138, + "step": 477 + }, + { + "epoch": 0.49101181304571134, + "grad_norm": 2.6418113708496094, + "learning_rate": 4.676437000380663e-06, + "loss": 1.6377, + "step": 478 + }, + { + "epoch": 0.49203903441191577, + "grad_norm": 2.338491439819336, + "learning_rate": 4.674533688618196e-06, + "loss": 1.5682, + "step": 479 + }, + { + "epoch": 0.4930662557781202, + "grad_norm": 2.5901622772216797, + "learning_rate": 4.6726303768557295e-06, + "loss": 1.4899, + "step": 480 + }, + { + "epoch": 0.4930662557781202, + "eval_loss": 1.8238886594772339, + "eval_runtime": 37.9605, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 480 + }, + { + "epoch": 0.4940934771443246, + "grad_norm": 2.470827102661133, + "learning_rate": 4.670727065093263e-06, + "loss": 1.5016, + "step": 481 + }, + { + "epoch": 0.49512069851052903, + "grad_norm": 2.371201992034912, + "learning_rate": 4.668823753330796e-06, + "loss": 1.5399, + "step": 482 + }, + { + "epoch": 0.49614791987673346, + "grad_norm": 2.532802104949951, + "learning_rate": 4.6669204415683295e-06, + "loss": 1.5985, + "step": 483 + }, + { + "epoch": 0.4971751412429379, + "grad_norm": 2.6294727325439453, + "learning_rate": 4.665017129805863e-06, + "loss": 1.5802, + "step": 484 + }, + { + "epoch": 0.49820236260914225, + "grad_norm": 2.2164878845214844, + "learning_rate": 4.663113818043396e-06, + "loss": 1.5384, + "step": 485 + }, + { + "epoch": 0.49922958397534667, + "grad_norm": 2.311232805252075, + "learning_rate": 4.66121050628093e-06, + "loss": 1.5785, + "step": 486 + }, + { + "epoch": 0.5002568053415511, + "grad_norm": 2.2916736602783203, + "learning_rate": 4.659307194518463e-06, + "loss": 1.5401, + "step": 487 + }, + { + "epoch": 0.5012840267077555, + "grad_norm": 2.3978800773620605, + "learning_rate": 4.6574038827559955e-06, + "loss": 1.5619, + "step": 488 + }, + { + "epoch": 0.50231124807396, + "grad_norm": 2.438917398452759, + "learning_rate": 4.655500570993529e-06, + "loss": 1.5824, + "step": 489 + }, + { + "epoch": 0.5033384694401644, + "grad_norm": 2.528862237930298, + "learning_rate": 4.653597259231062e-06, + "loss": 1.4802, + "step": 490 + }, + { + "epoch": 0.5033384694401644, + "eval_loss": 1.83504056930542, + "eval_runtime": 37.9357, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 490 + }, + { + "epoch": 0.5043656908063687, + "grad_norm": 2.54522442817688, + "learning_rate": 4.651693947468596e-06, + "loss": 1.6186, + "step": 491 + }, + { + "epoch": 0.5053929121725732, + "grad_norm": 2.5037031173706055, + "learning_rate": 4.649790635706129e-06, + "loss": 1.5159, + "step": 492 + }, + { + "epoch": 0.5064201335387776, + "grad_norm": 2.4600465297698975, + "learning_rate": 4.647887323943662e-06, + "loss": 1.5378, + "step": 493 + }, + { + "epoch": 0.507447354904982, + "grad_norm": 2.393695831298828, + "learning_rate": 4.645984012181196e-06, + "loss": 1.4485, + "step": 494 + }, + { + "epoch": 0.5084745762711864, + "grad_norm": 2.488626718521118, + "learning_rate": 4.644080700418729e-06, + "loss": 1.57, + "step": 495 + }, + { + "epoch": 0.5095017976373909, + "grad_norm": 2.409858465194702, + "learning_rate": 4.642177388656262e-06, + "loss": 1.5343, + "step": 496 + }, + { + "epoch": 0.5105290190035953, + "grad_norm": 2.5607573986053467, + "learning_rate": 4.640274076893796e-06, + "loss": 1.5825, + "step": 497 + }, + { + "epoch": 0.5115562403697997, + "grad_norm": 2.5953354835510254, + "learning_rate": 4.638370765131328e-06, + "loss": 1.5129, + "step": 498 + }, + { + "epoch": 0.5125834617360041, + "grad_norm": 2.3321897983551025, + "learning_rate": 4.636467453368862e-06, + "loss": 1.4972, + "step": 499 + }, + { + "epoch": 0.5136106831022085, + "grad_norm": 2.332052707672119, + "learning_rate": 4.634564141606396e-06, + "loss": 1.567, + "step": 500 + }, + { + "epoch": 0.5136106831022085, + "eval_loss": 1.824558138847351, + "eval_runtime": 37.9157, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 500 + }, + { + "epoch": 0.514637904468413, + "grad_norm": 2.431809902191162, + "learning_rate": 4.632660829843929e-06, + "loss": 1.494, + "step": 501 + }, + { + "epoch": 0.5156651258346173, + "grad_norm": 2.7118983268737793, + "learning_rate": 4.630757518081463e-06, + "loss": 1.6289, + "step": 502 + }, + { + "epoch": 0.5166923472008218, + "grad_norm": 2.3549370765686035, + "learning_rate": 4.628854206318996e-06, + "loss": 1.5363, + "step": 503 + }, + { + "epoch": 0.5177195685670262, + "grad_norm": 2.3343000411987305, + "learning_rate": 4.626950894556529e-06, + "loss": 1.5749, + "step": 504 + }, + { + "epoch": 0.5187467899332306, + "grad_norm": 2.4281466007232666, + "learning_rate": 4.625047582794062e-06, + "loss": 1.4958, + "step": 505 + }, + { + "epoch": 0.519774011299435, + "grad_norm": 2.3858845233917236, + "learning_rate": 4.623144271031595e-06, + "loss": 1.5031, + "step": 506 + }, + { + "epoch": 0.5208012326656395, + "grad_norm": 2.3714635372161865, + "learning_rate": 4.6212409592691286e-06, + "loss": 1.5808, + "step": 507 + }, + { + "epoch": 0.5218284540318439, + "grad_norm": 2.5290520191192627, + "learning_rate": 4.619337647506662e-06, + "loss": 1.541, + "step": 508 + }, + { + "epoch": 0.5228556753980482, + "grad_norm": 2.5521368980407715, + "learning_rate": 4.617434335744195e-06, + "loss": 1.4992, + "step": 509 + }, + { + "epoch": 0.5238828967642527, + "grad_norm": 2.436588764190674, + "learning_rate": 4.615531023981729e-06, + "loss": 1.5283, + "step": 510 + }, + { + "epoch": 0.5238828967642527, + "eval_loss": 1.8220165967941284, + "eval_runtime": 37.6123, + "eval_samples_per_second": 26.587, + "eval_steps_per_second": 1.117, + "step": 510 + }, + { + "epoch": 0.5249101181304571, + "grad_norm": 2.5240023136138916, + "learning_rate": 4.613627712219262e-06, + "loss": 1.5426, + "step": 511 + }, + { + "epoch": 0.5259373394966615, + "grad_norm": 2.5292775630950928, + "learning_rate": 4.611724400456795e-06, + "loss": 1.5411, + "step": 512 + }, + { + "epoch": 0.5269645608628659, + "grad_norm": 2.3622210025787354, + "learning_rate": 4.609821088694329e-06, + "loss": 1.5083, + "step": 513 + }, + { + "epoch": 0.5279917822290704, + "grad_norm": 2.545267105102539, + "learning_rate": 4.607917776931862e-06, + "loss": 1.5541, + "step": 514 + }, + { + "epoch": 0.5290190035952748, + "grad_norm": 2.3417961597442627, + "learning_rate": 4.606014465169395e-06, + "loss": 1.5686, + "step": 515 + }, + { + "epoch": 0.5300462249614792, + "grad_norm": 2.2782580852508545, + "learning_rate": 4.604111153406928e-06, + "loss": 1.4761, + "step": 516 + }, + { + "epoch": 0.5310734463276836, + "grad_norm": 2.4437527656555176, + "learning_rate": 4.602207841644461e-06, + "loss": 1.5818, + "step": 517 + }, + { + "epoch": 0.5321006676938881, + "grad_norm": 2.3594319820404053, + "learning_rate": 4.600304529881995e-06, + "loss": 1.5556, + "step": 518 + }, + { + "epoch": 0.5331278890600925, + "grad_norm": 2.325979471206665, + "learning_rate": 4.598401218119528e-06, + "loss": 1.5726, + "step": 519 + }, + { + "epoch": 0.5341551104262968, + "grad_norm": 2.2765676975250244, + "learning_rate": 4.5964979063570614e-06, + "loss": 1.5151, + "step": 520 + }, + { + "epoch": 0.5341551104262968, + "eval_loss": 1.825669288635254, + "eval_runtime": 37.7491, + "eval_samples_per_second": 26.491, + "eval_steps_per_second": 1.113, + "step": 520 + }, + { + "epoch": 0.5351823317925013, + "grad_norm": 2.357647657394409, + "learning_rate": 4.594594594594596e-06, + "loss": 1.5505, + "step": 521 + }, + { + "epoch": 0.5362095531587057, + "grad_norm": 2.3652896881103516, + "learning_rate": 4.592691282832128e-06, + "loss": 1.5621, + "step": 522 + }, + { + "epoch": 0.5372367745249101, + "grad_norm": 2.234096050262451, + "learning_rate": 4.5907879710696615e-06, + "loss": 1.4851, + "step": 523 + }, + { + "epoch": 0.5382639958911145, + "grad_norm": 2.3414828777313232, + "learning_rate": 4.588884659307195e-06, + "loss": 1.5594, + "step": 524 + }, + { + "epoch": 0.539291217257319, + "grad_norm": 2.3934929370880127, + "learning_rate": 4.586981347544728e-06, + "loss": 1.4994, + "step": 525 + }, + { + "epoch": 0.5403184386235234, + "grad_norm": 2.400479555130005, + "learning_rate": 4.585078035782262e-06, + "loss": 1.5346, + "step": 526 + }, + { + "epoch": 0.5413456599897278, + "grad_norm": 2.3817873001098633, + "learning_rate": 4.583174724019795e-06, + "loss": 1.4803, + "step": 527 + }, + { + "epoch": 0.5423728813559322, + "grad_norm": 2.4503185749053955, + "learning_rate": 4.581271412257328e-06, + "loss": 1.4711, + "step": 528 + }, + { + "epoch": 0.5434001027221366, + "grad_norm": 2.452578067779541, + "learning_rate": 4.579368100494862e-06, + "loss": 1.5645, + "step": 529 + }, + { + "epoch": 0.544427324088341, + "grad_norm": 2.3278186321258545, + "learning_rate": 4.577464788732395e-06, + "loss": 1.5428, + "step": 530 + }, + { + "epoch": 0.544427324088341, + "eval_loss": 1.8242785930633545, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 530 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 2.5192365646362305, + "learning_rate": 4.575561476969928e-06, + "loss": 1.5896, + "step": 531 + }, + { + "epoch": 0.5464817668207499, + "grad_norm": 2.308701276779175, + "learning_rate": 4.573658165207461e-06, + "loss": 1.5518, + "step": 532 + }, + { + "epoch": 0.5475089881869543, + "grad_norm": 2.4221107959747314, + "learning_rate": 4.571754853444994e-06, + "loss": 1.5387, + "step": 533 + }, + { + "epoch": 0.5485362095531587, + "grad_norm": 2.457606077194214, + "learning_rate": 4.569851541682528e-06, + "loss": 1.4997, + "step": 534 + }, + { + "epoch": 0.5495634309193631, + "grad_norm": 2.3406503200531006, + "learning_rate": 4.567948229920061e-06, + "loss": 1.4761, + "step": 535 + }, + { + "epoch": 0.5505906522855676, + "grad_norm": 2.555712938308716, + "learning_rate": 4.566044918157594e-06, + "loss": 1.5114, + "step": 536 + }, + { + "epoch": 0.551617873651772, + "grad_norm": 2.373013734817505, + "learning_rate": 4.564141606395128e-06, + "loss": 1.5516, + "step": 537 + }, + { + "epoch": 0.5526450950179764, + "grad_norm": 2.3915255069732666, + "learning_rate": 4.562238294632661e-06, + "loss": 1.5145, + "step": 538 + }, + { + "epoch": 0.5536723163841808, + "grad_norm": 2.3272323608398438, + "learning_rate": 4.5603349828701945e-06, + "loss": 1.5472, + "step": 539 + }, + { + "epoch": 0.5546995377503852, + "grad_norm": 2.416072130203247, + "learning_rate": 4.558431671107728e-06, + "loss": 1.5842, + "step": 540 + }, + { + "epoch": 0.5546995377503852, + "eval_loss": 1.825427532196045, + "eval_runtime": 37.9905, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 540 + }, + { + "epoch": 0.5557267591165896, + "grad_norm": 2.3225271701812744, + "learning_rate": 4.556528359345261e-06, + "loss": 1.5188, + "step": 541 + }, + { + "epoch": 0.556753980482794, + "grad_norm": 2.2387030124664307, + "learning_rate": 4.5546250475827946e-06, + "loss": 1.4778, + "step": 542 + }, + { + "epoch": 0.5577812018489985, + "grad_norm": 2.3054494857788086, + "learning_rate": 4.552721735820328e-06, + "loss": 1.5229, + "step": 543 + }, + { + "epoch": 0.5588084232152029, + "grad_norm": 2.4956626892089844, + "learning_rate": 4.550818424057861e-06, + "loss": 1.5367, + "step": 544 + }, + { + "epoch": 0.5598356445814073, + "grad_norm": 2.454986810684204, + "learning_rate": 4.548915112295395e-06, + "loss": 1.5444, + "step": 545 + }, + { + "epoch": 0.5608628659476117, + "grad_norm": 2.466810703277588, + "learning_rate": 4.547011800532928e-06, + "loss": 1.4914, + "step": 546 + }, + { + "epoch": 0.5618900873138162, + "grad_norm": 2.51777982711792, + "learning_rate": 4.545108488770461e-06, + "loss": 1.5775, + "step": 547 + }, + { + "epoch": 0.5629173086800205, + "grad_norm": 2.340421199798584, + "learning_rate": 4.543205177007995e-06, + "loss": 1.5185, + "step": 548 + }, + { + "epoch": 0.5639445300462249, + "grad_norm": 2.376772880554199, + "learning_rate": 4.541301865245527e-06, + "loss": 1.5588, + "step": 549 + }, + { + "epoch": 0.5649717514124294, + "grad_norm": 2.4004335403442383, + "learning_rate": 4.539398553483061e-06, + "loss": 1.5283, + "step": 550 + }, + { + "epoch": 0.5649717514124294, + "eval_loss": 1.8294963836669922, + "eval_runtime": 37.9391, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 550 + }, + { + "epoch": 0.5659989727786338, + "grad_norm": 2.424459934234619, + "learning_rate": 4.537495241720594e-06, + "loss": 1.5576, + "step": 551 + }, + { + "epoch": 0.5670261941448382, + "grad_norm": 2.4413998126983643, + "learning_rate": 4.535591929958127e-06, + "loss": 1.5444, + "step": 552 + }, + { + "epoch": 0.5680534155110426, + "grad_norm": 2.408897876739502, + "learning_rate": 4.533688618195661e-06, + "loss": 1.5269, + "step": 553 + }, + { + "epoch": 0.5690806368772471, + "grad_norm": 2.471024990081787, + "learning_rate": 4.531785306433194e-06, + "loss": 1.5109, + "step": 554 + }, + { + "epoch": 0.5701078582434514, + "grad_norm": 2.374882459640503, + "learning_rate": 4.529881994670727e-06, + "loss": 1.5205, + "step": 555 + }, + { + "epoch": 0.5711350796096559, + "grad_norm": 2.4166030883789062, + "learning_rate": 4.527978682908261e-06, + "loss": 1.5276, + "step": 556 + }, + { + "epoch": 0.5721623009758603, + "grad_norm": 2.2681620121002197, + "learning_rate": 4.526075371145794e-06, + "loss": 1.5374, + "step": 557 + }, + { + "epoch": 0.5731895223420647, + "grad_norm": 2.3933286666870117, + "learning_rate": 4.5241720593833275e-06, + "loss": 1.5349, + "step": 558 + }, + { + "epoch": 0.5742167437082691, + "grad_norm": 2.4086451530456543, + "learning_rate": 4.52226874762086e-06, + "loss": 1.5512, + "step": 559 + }, + { + "epoch": 0.5752439650744735, + "grad_norm": 2.559021472930908, + "learning_rate": 4.520365435858393e-06, + "loss": 1.5789, + "step": 560 + }, + { + "epoch": 0.5752439650744735, + "eval_loss": 1.8243234157562256, + "eval_runtime": 37.934, + "eval_samples_per_second": 26.362, + "eval_steps_per_second": 1.107, + "step": 560 + }, + { + "epoch": 0.576271186440678, + "grad_norm": 2.483494520187378, + "learning_rate": 4.518462124095928e-06, + "loss": 1.5348, + "step": 561 + }, + { + "epoch": 0.5772984078068824, + "grad_norm": 2.4039993286132812, + "learning_rate": 4.516558812333461e-06, + "loss": 1.5201, + "step": 562 + }, + { + "epoch": 0.5783256291730868, + "grad_norm": 2.5080959796905518, + "learning_rate": 4.514655500570994e-06, + "loss": 1.5864, + "step": 563 + }, + { + "epoch": 0.5793528505392912, + "grad_norm": 2.620887279510498, + "learning_rate": 4.512752188808528e-06, + "loss": 1.5901, + "step": 564 + }, + { + "epoch": 0.5803800719054957, + "grad_norm": 2.5691096782684326, + "learning_rate": 4.510848877046061e-06, + "loss": 1.5506, + "step": 565 + }, + { + "epoch": 0.5814072932717, + "grad_norm": 2.3777835369110107, + "learning_rate": 4.5089455652835936e-06, + "loss": 1.5149, + "step": 566 + }, + { + "epoch": 0.5824345146379045, + "grad_norm": 2.2791666984558105, + "learning_rate": 4.507042253521127e-06, + "loss": 1.4771, + "step": 567 + }, + { + "epoch": 0.5834617360041089, + "grad_norm": 2.4953627586364746, + "learning_rate": 4.50513894175866e-06, + "loss": 1.556, + "step": 568 + }, + { + "epoch": 0.5844889573703133, + "grad_norm": 2.3563215732574463, + "learning_rate": 4.503235629996194e-06, + "loss": 1.549, + "step": 569 + }, + { + "epoch": 0.5855161787365177, + "grad_norm": 2.49676513671875, + "learning_rate": 4.501332318233727e-06, + "loss": 1.513, + "step": 570 + }, + { + "epoch": 0.5855161787365177, + "eval_loss": 1.8353043794631958, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 570 + }, + { + "epoch": 0.5865434001027221, + "grad_norm": 2.385275363922119, + "learning_rate": 4.49942900647126e-06, + "loss": 1.5622, + "step": 571 + }, + { + "epoch": 0.5875706214689266, + "grad_norm": 2.3089914321899414, + "learning_rate": 4.497525694708794e-06, + "loss": 1.5426, + "step": 572 + }, + { + "epoch": 0.588597842835131, + "grad_norm": 2.307323455810547, + "learning_rate": 4.495622382946327e-06, + "loss": 1.5088, + "step": 573 + }, + { + "epoch": 0.5896250642013354, + "grad_norm": 2.597288131713867, + "learning_rate": 4.4937190711838605e-06, + "loss": 1.5788, + "step": 574 + }, + { + "epoch": 0.5906522855675398, + "grad_norm": 2.3923256397247314, + "learning_rate": 4.491815759421394e-06, + "loss": 1.5038, + "step": 575 + }, + { + "epoch": 0.5916795069337443, + "grad_norm": 2.4199156761169434, + "learning_rate": 4.489912447658926e-06, + "loss": 1.5057, + "step": 576 + }, + { + "epoch": 0.5927067282999486, + "grad_norm": 2.5643393993377686, + "learning_rate": 4.48800913589646e-06, + "loss": 1.5274, + "step": 577 + }, + { + "epoch": 0.593733949666153, + "grad_norm": 2.5854485034942627, + "learning_rate": 4.486105824133993e-06, + "loss": 1.5341, + "step": 578 + }, + { + "epoch": 0.5947611710323575, + "grad_norm": 2.5363597869873047, + "learning_rate": 4.4842025123715264e-06, + "loss": 1.4953, + "step": 579 + }, + { + "epoch": 0.5957883923985619, + "grad_norm": 2.3880693912506104, + "learning_rate": 4.48229920060906e-06, + "loss": 1.4868, + "step": 580 + }, + { + "epoch": 0.5957883923985619, + "eval_loss": 1.8251410722732544, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 580 + }, + { + "epoch": 0.5968156137647663, + "grad_norm": 2.5272583961486816, + "learning_rate": 4.480395888846594e-06, + "loss": 1.5851, + "step": 581 + }, + { + "epoch": 0.5978428351309707, + "grad_norm": 2.4496943950653076, + "learning_rate": 4.478492577084127e-06, + "loss": 1.4896, + "step": 582 + }, + { + "epoch": 0.5988700564971752, + "grad_norm": 2.3996055126190186, + "learning_rate": 4.47658926532166e-06, + "loss": 1.5253, + "step": 583 + }, + { + "epoch": 0.5998972778633795, + "grad_norm": 2.5964202880859375, + "learning_rate": 4.474685953559193e-06, + "loss": 1.6294, + "step": 584 + }, + { + "epoch": 0.600924499229584, + "grad_norm": 2.5567739009857178, + "learning_rate": 4.472782641796727e-06, + "loss": 1.4834, + "step": 585 + }, + { + "epoch": 0.6019517205957884, + "grad_norm": 2.440291166305542, + "learning_rate": 4.47087933003426e-06, + "loss": 1.4966, + "step": 586 + }, + { + "epoch": 0.6029789419619929, + "grad_norm": 2.525763988494873, + "learning_rate": 4.468976018271793e-06, + "loss": 1.5005, + "step": 587 + }, + { + "epoch": 0.6040061633281972, + "grad_norm": 2.4822962284088135, + "learning_rate": 4.467072706509327e-06, + "loss": 1.5358, + "step": 588 + }, + { + "epoch": 0.6050333846944016, + "grad_norm": 2.591149091720581, + "learning_rate": 4.46516939474686e-06, + "loss": 1.4975, + "step": 589 + }, + { + "epoch": 0.6060606060606061, + "grad_norm": 2.4884238243103027, + "learning_rate": 4.463266082984393e-06, + "loss": 1.6107, + "step": 590 + }, + { + "epoch": 0.6060606060606061, + "eval_loss": 1.8383466005325317, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 590 + }, + { + "epoch": 0.6070878274268104, + "grad_norm": 2.4965732097625732, + "learning_rate": 4.461362771221927e-06, + "loss": 1.4986, + "step": 591 + }, + { + "epoch": 0.6081150487930149, + "grad_norm": 2.544792413711548, + "learning_rate": 4.45945945945946e-06, + "loss": 1.5067, + "step": 592 + }, + { + "epoch": 0.6091422701592193, + "grad_norm": 2.5726170539855957, + "learning_rate": 4.457556147696993e-06, + "loss": 1.5861, + "step": 593 + }, + { + "epoch": 0.6101694915254238, + "grad_norm": 2.493027448654175, + "learning_rate": 4.455652835934526e-06, + "loss": 1.5857, + "step": 594 + }, + { + "epoch": 0.6111967128916281, + "grad_norm": 2.470771074295044, + "learning_rate": 4.453749524172059e-06, + "loss": 1.5419, + "step": 595 + }, + { + "epoch": 0.6122239342578326, + "grad_norm": 2.568052053451538, + "learning_rate": 4.451846212409593e-06, + "loss": 1.5391, + "step": 596 + }, + { + "epoch": 0.613251155624037, + "grad_norm": 2.4346213340759277, + "learning_rate": 4.449942900647126e-06, + "loss": 1.4761, + "step": 597 + }, + { + "epoch": 0.6142783769902413, + "grad_norm": 2.418133497238159, + "learning_rate": 4.4480395888846595e-06, + "loss": 1.5383, + "step": 598 + }, + { + "epoch": 0.6153055983564458, + "grad_norm": 2.4795567989349365, + "learning_rate": 4.446136277122193e-06, + "loss": 1.4777, + "step": 599 + }, + { + "epoch": 0.6163328197226502, + "grad_norm": 2.5122742652893066, + "learning_rate": 4.444232965359726e-06, + "loss": 1.5353, + "step": 600 + }, + { + "epoch": 0.6163328197226502, + "eval_loss": 1.8325241804122925, + "eval_runtime": 37.9546, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 600 + }, + { + "epoch": 0.6173600410888547, + "grad_norm": 2.604067325592041, + "learning_rate": 4.4423296535972596e-06, + "loss": 1.489, + "step": 601 + }, + { + "epoch": 0.618387262455059, + "grad_norm": 2.3741753101348877, + "learning_rate": 4.440426341834793e-06, + "loss": 1.532, + "step": 602 + }, + { + "epoch": 0.6194144838212635, + "grad_norm": 2.5373036861419678, + "learning_rate": 4.438523030072326e-06, + "loss": 1.561, + "step": 603 + }, + { + "epoch": 0.6204417051874679, + "grad_norm": 2.2958099842071533, + "learning_rate": 4.43661971830986e-06, + "loss": 1.5025, + "step": 604 + }, + { + "epoch": 0.6214689265536724, + "grad_norm": 2.417006015777588, + "learning_rate": 4.434716406547393e-06, + "loss": 1.4977, + "step": 605 + }, + { + "epoch": 0.6224961479198767, + "grad_norm": 2.335296392440796, + "learning_rate": 4.432813094784926e-06, + "loss": 1.5504, + "step": 606 + }, + { + "epoch": 0.6235233692860811, + "grad_norm": 2.4086828231811523, + "learning_rate": 4.43090978302246e-06, + "loss": 1.5217, + "step": 607 + }, + { + "epoch": 0.6245505906522856, + "grad_norm": 2.3816819190979004, + "learning_rate": 4.429006471259993e-06, + "loss": 1.5126, + "step": 608 + }, + { + "epoch": 0.6255778120184899, + "grad_norm": 2.307671546936035, + "learning_rate": 4.4271031594975265e-06, + "loss": 1.4876, + "step": 609 + }, + { + "epoch": 0.6266050333846944, + "grad_norm": 2.355837821960449, + "learning_rate": 4.425199847735059e-06, + "loss": 1.4827, + "step": 610 + }, + { + "epoch": 0.6266050333846944, + "eval_loss": 1.8269455432891846, + "eval_runtime": 37.6459, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 610 + }, + { + "epoch": 0.6276322547508988, + "grad_norm": 2.558631658554077, + "learning_rate": 4.423296535972592e-06, + "loss": 1.5172, + "step": 611 + }, + { + "epoch": 0.6286594761171033, + "grad_norm": 2.3170175552368164, + "learning_rate": 4.421393224210126e-06, + "loss": 1.4784, + "step": 612 + }, + { + "epoch": 0.6296866974833076, + "grad_norm": 2.4064524173736572, + "learning_rate": 4.419489912447659e-06, + "loss": 1.5616, + "step": 613 + }, + { + "epoch": 0.6307139188495121, + "grad_norm": 2.3719160556793213, + "learning_rate": 4.4175866006851924e-06, + "loss": 1.5292, + "step": 614 + }, + { + "epoch": 0.6317411402157165, + "grad_norm": 2.3808202743530273, + "learning_rate": 4.415683288922726e-06, + "loss": 1.5122, + "step": 615 + }, + { + "epoch": 0.632768361581921, + "grad_norm": 2.4208037853240967, + "learning_rate": 4.413779977160259e-06, + "loss": 1.5088, + "step": 616 + }, + { + "epoch": 0.6337955829481253, + "grad_norm": 2.5558979511260986, + "learning_rate": 4.4118766653977925e-06, + "loss": 1.5678, + "step": 617 + }, + { + "epoch": 0.6348228043143297, + "grad_norm": 2.443908452987671, + "learning_rate": 4.409973353635326e-06, + "loss": 1.5234, + "step": 618 + }, + { + "epoch": 0.6358500256805342, + "grad_norm": 2.420408010482788, + "learning_rate": 4.408070041872859e-06, + "loss": 1.4708, + "step": 619 + }, + { + "epoch": 0.6368772470467385, + "grad_norm": 2.3884904384613037, + "learning_rate": 4.406166730110393e-06, + "loss": 1.4821, + "step": 620 + }, + { + "epoch": 0.6368772470467385, + "eval_loss": 1.8217289447784424, + "eval_runtime": 37.7936, + "eval_samples_per_second": 26.459, + "eval_steps_per_second": 1.111, + "step": 620 + }, + { + "epoch": 0.637904468412943, + "grad_norm": 2.313720703125, + "learning_rate": 4.404263418347925e-06, + "loss": 1.4962, + "step": 621 + }, + { + "epoch": 0.6389316897791474, + "grad_norm": 2.3173866271972656, + "learning_rate": 4.402360106585459e-06, + "loss": 1.4857, + "step": 622 + }, + { + "epoch": 0.6399589111453519, + "grad_norm": 2.541757822036743, + "learning_rate": 4.400456794822993e-06, + "loss": 1.5177, + "step": 623 + }, + { + "epoch": 0.6409861325115562, + "grad_norm": 2.4382357597351074, + "learning_rate": 4.398553483060526e-06, + "loss": 1.4718, + "step": 624 + }, + { + "epoch": 0.6420133538777607, + "grad_norm": 2.2988059520721436, + "learning_rate": 4.396650171298059e-06, + "loss": 1.4883, + "step": 625 + }, + { + "epoch": 0.6430405752439651, + "grad_norm": 2.4817495346069336, + "learning_rate": 4.394746859535593e-06, + "loss": 1.4823, + "step": 626 + }, + { + "epoch": 0.6440677966101694, + "grad_norm": 2.4578585624694824, + "learning_rate": 4.392843547773125e-06, + "loss": 1.4882, + "step": 627 + }, + { + "epoch": 0.6450950179763739, + "grad_norm": 2.184972047805786, + "learning_rate": 4.390940236010659e-06, + "loss": 1.514, + "step": 628 + }, + { + "epoch": 0.6461222393425783, + "grad_norm": 2.4040279388427734, + "learning_rate": 4.389036924248192e-06, + "loss": 1.5343, + "step": 629 + }, + { + "epoch": 0.6471494607087828, + "grad_norm": 2.3676459789276123, + "learning_rate": 4.387133612485725e-06, + "loss": 1.5007, + "step": 630 + }, + { + "epoch": 0.6471494607087828, + "eval_loss": 1.8106164932250977, + "eval_runtime": 37.9383, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 630 + }, + { + "epoch": 0.6481766820749871, + "grad_norm": 2.234689235687256, + "learning_rate": 4.385230300723259e-06, + "loss": 1.5064, + "step": 631 + }, + { + "epoch": 0.6492039034411916, + "grad_norm": 2.281742572784424, + "learning_rate": 4.383326988960792e-06, + "loss": 1.4777, + "step": 632 + }, + { + "epoch": 0.650231124807396, + "grad_norm": 2.4314322471618652, + "learning_rate": 4.3814236771983255e-06, + "loss": 1.5331, + "step": 633 + }, + { + "epoch": 0.6512583461736005, + "grad_norm": 2.4092588424682617, + "learning_rate": 4.379520365435859e-06, + "loss": 1.5122, + "step": 634 + }, + { + "epoch": 0.6522855675398048, + "grad_norm": 2.517362117767334, + "learning_rate": 4.377617053673392e-06, + "loss": 1.52, + "step": 635 + }, + { + "epoch": 0.6533127889060092, + "grad_norm": 2.4784109592437744, + "learning_rate": 4.3757137419109256e-06, + "loss": 1.5771, + "step": 636 + }, + { + "epoch": 0.6543400102722137, + "grad_norm": 2.32181715965271, + "learning_rate": 4.373810430148459e-06, + "loss": 1.4718, + "step": 637 + }, + { + "epoch": 0.655367231638418, + "grad_norm": 2.5146260261535645, + "learning_rate": 4.3719071183859914e-06, + "loss": 1.5307, + "step": 638 + }, + { + "epoch": 0.6563944530046225, + "grad_norm": 2.2364494800567627, + "learning_rate": 4.370003806623525e-06, + "loss": 1.4803, + "step": 639 + }, + { + "epoch": 0.6574216743708269, + "grad_norm": 2.3735806941986084, + "learning_rate": 4.368100494861058e-06, + "loss": 1.5249, + "step": 640 + }, + { + "epoch": 0.6574216743708269, + "eval_loss": 1.8134880065917969, + "eval_runtime": 38.0376, + "eval_samples_per_second": 26.29, + "eval_steps_per_second": 1.104, + "step": 640 + }, + { + "epoch": 0.6584488957370314, + "grad_norm": 2.4600815773010254, + "learning_rate": 4.3661971830985915e-06, + "loss": 1.4893, + "step": 641 + }, + { + "epoch": 0.6594761171032357, + "grad_norm": 2.4106783866882324, + "learning_rate": 4.364293871336126e-06, + "loss": 1.5163, + "step": 642 + }, + { + "epoch": 0.6605033384694402, + "grad_norm": 2.276624917984009, + "learning_rate": 4.362390559573659e-06, + "loss": 1.482, + "step": 643 + }, + { + "epoch": 0.6615305598356446, + "grad_norm": 2.336341381072998, + "learning_rate": 4.3604872478111925e-06, + "loss": 1.4152, + "step": 644 + }, + { + "epoch": 0.662557781201849, + "grad_norm": 2.3329105377197266, + "learning_rate": 4.358583936048725e-06, + "loss": 1.5411, + "step": 645 + }, + { + "epoch": 0.6635850025680534, + "grad_norm": 2.2877161502838135, + "learning_rate": 4.356680624286258e-06, + "loss": 1.4997, + "step": 646 + }, + { + "epoch": 0.6646122239342578, + "grad_norm": 2.410820484161377, + "learning_rate": 4.354777312523792e-06, + "loss": 1.5446, + "step": 647 + }, + { + "epoch": 0.6656394453004623, + "grad_norm": 2.403038501739502, + "learning_rate": 4.352874000761325e-06, + "loss": 1.4916, + "step": 648 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 2.3261778354644775, + "learning_rate": 4.3509706889988584e-06, + "loss": 1.5561, + "step": 649 + }, + { + "epoch": 0.6676938880328711, + "grad_norm": 2.4512174129486084, + "learning_rate": 4.349067377236392e-06, + "loss": 1.556, + "step": 650 + }, + { + "epoch": 0.6676938880328711, + "eval_loss": 1.8238279819488525, + "eval_runtime": 37.9988, + "eval_samples_per_second": 26.317, + "eval_steps_per_second": 1.105, + "step": 650 + }, + { + "epoch": 0.6687211093990755, + "grad_norm": 2.4523823261260986, + "learning_rate": 4.347164065473925e-06, + "loss": 1.5676, + "step": 651 + }, + { + "epoch": 0.66974833076528, + "grad_norm": 2.3770365715026855, + "learning_rate": 4.3452607537114585e-06, + "loss": 1.4815, + "step": 652 + }, + { + "epoch": 0.6707755521314843, + "grad_norm": 2.2440760135650635, + "learning_rate": 4.343357441948992e-06, + "loss": 1.5141, + "step": 653 + }, + { + "epoch": 0.6718027734976888, + "grad_norm": 2.3341948986053467, + "learning_rate": 4.341454130186525e-06, + "loss": 1.4859, + "step": 654 + }, + { + "epoch": 0.6728299948638932, + "grad_norm": 2.507354497909546, + "learning_rate": 4.339550818424058e-06, + "loss": 1.4734, + "step": 655 + }, + { + "epoch": 0.6738572162300975, + "grad_norm": 2.395954132080078, + "learning_rate": 4.337647506661591e-06, + "loss": 1.5069, + "step": 656 + }, + { + "epoch": 0.674884437596302, + "grad_norm": 2.362211227416992, + "learning_rate": 4.3357441948991245e-06, + "loss": 1.5299, + "step": 657 + }, + { + "epoch": 0.6759116589625064, + "grad_norm": 2.460008144378662, + "learning_rate": 4.333840883136658e-06, + "loss": 1.4795, + "step": 658 + }, + { + "epoch": 0.6769388803287109, + "grad_norm": 2.3578217029571533, + "learning_rate": 4.331937571374191e-06, + "loss": 1.4906, + "step": 659 + }, + { + "epoch": 0.6779661016949152, + "grad_norm": 2.398843288421631, + "learning_rate": 4.3300342596117246e-06, + "loss": 1.4986, + "step": 660 + }, + { + "epoch": 0.6779661016949152, + "eval_loss": 1.8206881284713745, + "eval_runtime": 38.044, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 660 + }, + { + "epoch": 0.6789933230611197, + "grad_norm": 2.3587260246276855, + "learning_rate": 4.328130947849258e-06, + "loss": 1.461, + "step": 661 + }, + { + "epoch": 0.6800205444273241, + "grad_norm": 2.311952590942383, + "learning_rate": 4.326227636086791e-06, + "loss": 1.5041, + "step": 662 + }, + { + "epoch": 0.6810477657935285, + "grad_norm": 2.376188278198242, + "learning_rate": 4.324324324324325e-06, + "loss": 1.5042, + "step": 663 + }, + { + "epoch": 0.6820749871597329, + "grad_norm": 2.5807230472564697, + "learning_rate": 4.322421012561858e-06, + "loss": 1.4935, + "step": 664 + }, + { + "epoch": 0.6831022085259374, + "grad_norm": 2.368128538131714, + "learning_rate": 4.320517700799391e-06, + "loss": 1.5851, + "step": 665 + }, + { + "epoch": 0.6841294298921418, + "grad_norm": 2.4517524242401123, + "learning_rate": 4.318614389036925e-06, + "loss": 1.5528, + "step": 666 + }, + { + "epoch": 0.6851566512583461, + "grad_norm": 2.5436811447143555, + "learning_rate": 4.316711077274458e-06, + "loss": 1.4633, + "step": 667 + }, + { + "epoch": 0.6861838726245506, + "grad_norm": 2.4287915229797363, + "learning_rate": 4.3148077655119915e-06, + "loss": 1.5454, + "step": 668 + }, + { + "epoch": 0.687211093990755, + "grad_norm": 2.386528730392456, + "learning_rate": 4.312904453749525e-06, + "loss": 1.5116, + "step": 669 + }, + { + "epoch": 0.6882383153569595, + "grad_norm": 2.6602251529693604, + "learning_rate": 4.311001141987058e-06, + "loss": 1.5199, + "step": 670 + }, + { + "epoch": 0.6882383153569595, + "eval_loss": 1.8252663612365723, + "eval_runtime": 38.0791, + "eval_samples_per_second": 26.261, + "eval_steps_per_second": 1.103, + "step": 670 + }, + { + "epoch": 0.6892655367231638, + "grad_norm": 2.351128339767456, + "learning_rate": 4.3090978302245916e-06, + "loss": 1.507, + "step": 671 + }, + { + "epoch": 0.6902927580893683, + "grad_norm": 2.393441677093506, + "learning_rate": 4.307194518462124e-06, + "loss": 1.579, + "step": 672 + }, + { + "epoch": 0.6913199794555727, + "grad_norm": 2.3730578422546387, + "learning_rate": 4.3052912066996574e-06, + "loss": 1.5446, + "step": 673 + }, + { + "epoch": 0.6923472008217771, + "grad_norm": 2.2522945404052734, + "learning_rate": 4.303387894937191e-06, + "loss": 1.4705, + "step": 674 + }, + { + "epoch": 0.6933744221879815, + "grad_norm": 2.384563446044922, + "learning_rate": 4.301484583174724e-06, + "loss": 1.4636, + "step": 675 + }, + { + "epoch": 0.6944016435541859, + "grad_norm": 2.2879116535186768, + "learning_rate": 4.2995812714122575e-06, + "loss": 1.4914, + "step": 676 + }, + { + "epoch": 0.6954288649203904, + "grad_norm": 2.494788646697998, + "learning_rate": 4.297677959649791e-06, + "loss": 1.5126, + "step": 677 + }, + { + "epoch": 0.6964560862865947, + "grad_norm": 2.700012683868408, + "learning_rate": 4.295774647887324e-06, + "loss": 1.5362, + "step": 678 + }, + { + "epoch": 0.6974833076527992, + "grad_norm": 2.5130302906036377, + "learning_rate": 4.293871336124858e-06, + "loss": 1.5075, + "step": 679 + }, + { + "epoch": 0.6985105290190036, + "grad_norm": 2.486236572265625, + "learning_rate": 4.291968024362391e-06, + "loss": 1.5338, + "step": 680 + }, + { + "epoch": 0.6985105290190036, + "eval_loss": 1.8229657411575317, + "eval_runtime": 38.0642, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 680 + }, + { + "epoch": 0.699537750385208, + "grad_norm": 2.441683530807495, + "learning_rate": 4.290064712599924e-06, + "loss": 1.537, + "step": 681 + }, + { + "epoch": 0.7005649717514124, + "grad_norm": 2.396798610687256, + "learning_rate": 4.288161400837458e-06, + "loss": 1.4553, + "step": 682 + }, + { + "epoch": 0.7015921931176169, + "grad_norm": 2.3636221885681152, + "learning_rate": 4.286258089074991e-06, + "loss": 1.5138, + "step": 683 + }, + { + "epoch": 0.7026194144838213, + "grad_norm": 2.4584708213806152, + "learning_rate": 4.284354777312524e-06, + "loss": 1.4725, + "step": 684 + }, + { + "epoch": 0.7036466358500256, + "grad_norm": 2.850332498550415, + "learning_rate": 4.282451465550058e-06, + "loss": 1.5075, + "step": 685 + }, + { + "epoch": 0.7046738572162301, + "grad_norm": 2.6345887184143066, + "learning_rate": 4.280548153787591e-06, + "loss": 1.5054, + "step": 686 + }, + { + "epoch": 0.7057010785824345, + "grad_norm": 2.3535311222076416, + "learning_rate": 4.2786448420251245e-06, + "loss": 1.4607, + "step": 687 + }, + { + "epoch": 0.706728299948639, + "grad_norm": 2.490431547164917, + "learning_rate": 4.276741530262658e-06, + "loss": 1.4852, + "step": 688 + }, + { + "epoch": 0.7077555213148433, + "grad_norm": 2.4353036880493164, + "learning_rate": 4.27483821850019e-06, + "loss": 1.5503, + "step": 689 + }, + { + "epoch": 0.7087827426810478, + "grad_norm": 2.3356940746307373, + "learning_rate": 4.272934906737724e-06, + "loss": 1.5416, + "step": 690 + }, + { + "epoch": 0.7087827426810478, + "eval_loss": 1.8164218664169312, + "eval_runtime": 38.0592, + "eval_samples_per_second": 26.275, + "eval_steps_per_second": 1.104, + "step": 690 + }, + { + "epoch": 0.7098099640472522, + "grad_norm": 2.4772775173187256, + "learning_rate": 4.271031594975257e-06, + "loss": 1.5392, + "step": 691 + }, + { + "epoch": 0.7108371854134566, + "grad_norm": 2.2921438217163086, + "learning_rate": 4.2691282832127905e-06, + "loss": 1.4935, + "step": 692 + }, + { + "epoch": 0.711864406779661, + "grad_norm": 2.3176305294036865, + "learning_rate": 4.267224971450324e-06, + "loss": 1.4451, + "step": 693 + }, + { + "epoch": 0.7128916281458655, + "grad_norm": 2.6825685501098633, + "learning_rate": 4.265321659687857e-06, + "loss": 1.4878, + "step": 694 + }, + { + "epoch": 0.7139188495120699, + "grad_norm": 2.6141414642333984, + "learning_rate": 4.2634183479253906e-06, + "loss": 1.4695, + "step": 695 + }, + { + "epoch": 0.7149460708782742, + "grad_norm": 2.5179429054260254, + "learning_rate": 4.261515036162924e-06, + "loss": 1.492, + "step": 696 + }, + { + "epoch": 0.7159732922444787, + "grad_norm": 2.4744884967803955, + "learning_rate": 4.259611724400457e-06, + "loss": 1.4752, + "step": 697 + }, + { + "epoch": 0.7170005136106831, + "grad_norm": 2.4911048412323, + "learning_rate": 4.257708412637991e-06, + "loss": 1.423, + "step": 698 + }, + { + "epoch": 0.7180277349768875, + "grad_norm": 2.5116214752197266, + "learning_rate": 4.255805100875523e-06, + "loss": 1.4535, + "step": 699 + }, + { + "epoch": 0.7190549563430919, + "grad_norm": 2.476651430130005, + "learning_rate": 4.2539017891130565e-06, + "loss": 1.5349, + "step": 700 + }, + { + "epoch": 0.7190549563430919, + "eval_loss": 1.8069179058074951, + "eval_runtime": 38.0292, + "eval_samples_per_second": 26.296, + "eval_steps_per_second": 1.104, + "step": 700 + }, + { + "epoch": 0.7200821777092964, + "grad_norm": 2.5234575271606445, + "learning_rate": 4.25199847735059e-06, + "loss": 1.4596, + "step": 701 + }, + { + "epoch": 0.7211093990755008, + "grad_norm": 2.33290433883667, + "learning_rate": 4.250095165588123e-06, + "loss": 1.5147, + "step": 702 + }, + { + "epoch": 0.7221366204417052, + "grad_norm": 2.479883909225464, + "learning_rate": 4.2481918538256575e-06, + "loss": 1.47, + "step": 703 + }, + { + "epoch": 0.7231638418079096, + "grad_norm": 2.432253837585449, + "learning_rate": 4.246288542063191e-06, + "loss": 1.5152, + "step": 704 + }, + { + "epoch": 0.724191063174114, + "grad_norm": 2.358304738998413, + "learning_rate": 4.244385230300724e-06, + "loss": 1.4562, + "step": 705 + }, + { + "epoch": 0.7252182845403184, + "grad_norm": 2.480877161026001, + "learning_rate": 4.242481918538257e-06, + "loss": 1.4906, + "step": 706 + }, + { + "epoch": 0.7262455059065228, + "grad_norm": 2.4287192821502686, + "learning_rate": 4.24057860677579e-06, + "loss": 1.5294, + "step": 707 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 2.4782888889312744, + "learning_rate": 4.2386752950133234e-06, + "loss": 1.4767, + "step": 708 + }, + { + "epoch": 0.7282999486389317, + "grad_norm": 2.4948253631591797, + "learning_rate": 4.236771983250857e-06, + "loss": 1.4664, + "step": 709 + }, + { + "epoch": 0.7293271700051361, + "grad_norm": 2.4625205993652344, + "learning_rate": 4.23486867148839e-06, + "loss": 1.495, + "step": 710 + }, + { + "epoch": 0.7293271700051361, + "eval_loss": 1.8140079975128174, + "eval_runtime": 37.6964, + "eval_samples_per_second": 26.528, + "eval_steps_per_second": 1.114, + "step": 710 + }, + { + "epoch": 0.7303543913713405, + "grad_norm": 2.669027805328369, + "learning_rate": 4.2329653597259235e-06, + "loss": 1.5204, + "step": 711 + }, + { + "epoch": 0.731381612737545, + "grad_norm": 2.5024707317352295, + "learning_rate": 4.231062047963457e-06, + "loss": 1.5283, + "step": 712 + }, + { + "epoch": 0.7324088341037494, + "grad_norm": 2.498258590698242, + "learning_rate": 4.22915873620099e-06, + "loss": 1.5139, + "step": 713 + }, + { + "epoch": 0.7334360554699538, + "grad_norm": 2.4305667877197266, + "learning_rate": 4.227255424438524e-06, + "loss": 1.5337, + "step": 714 + }, + { + "epoch": 0.7344632768361582, + "grad_norm": 2.3153419494628906, + "learning_rate": 4.225352112676057e-06, + "loss": 1.4646, + "step": 715 + }, + { + "epoch": 0.7354904982023626, + "grad_norm": 2.2835583686828613, + "learning_rate": 4.2234488009135895e-06, + "loss": 1.4328, + "step": 716 + }, + { + "epoch": 0.736517719568567, + "grad_norm": 2.401418924331665, + "learning_rate": 4.221545489151123e-06, + "loss": 1.4685, + "step": 717 + }, + { + "epoch": 0.7375449409347714, + "grad_norm": 2.454864978790283, + "learning_rate": 4.219642177388656e-06, + "loss": 1.4925, + "step": 718 + }, + { + "epoch": 0.7385721623009759, + "grad_norm": 2.2332873344421387, + "learning_rate": 4.2177388656261896e-06, + "loss": 1.4481, + "step": 719 + }, + { + "epoch": 0.7395993836671803, + "grad_norm": 2.46628475189209, + "learning_rate": 4.215835553863723e-06, + "loss": 1.4173, + "step": 720 + }, + { + "epoch": 0.7395993836671803, + "eval_loss": 1.817592740058899, + "eval_runtime": 37.8086, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 720 + }, + { + "epoch": 0.7406266050333847, + "grad_norm": 2.346245765686035, + "learning_rate": 4.213932242101256e-06, + "loss": 1.4688, + "step": 721 + }, + { + "epoch": 0.7416538263995891, + "grad_norm": 2.2708640098571777, + "learning_rate": 4.21202893033879e-06, + "loss": 1.4937, + "step": 722 + }, + { + "epoch": 0.7426810477657936, + "grad_norm": 2.3716516494750977, + "learning_rate": 4.210125618576323e-06, + "loss": 1.4869, + "step": 723 + }, + { + "epoch": 0.743708269131998, + "grad_norm": 2.443058490753174, + "learning_rate": 4.208222306813856e-06, + "loss": 1.542, + "step": 724 + }, + { + "epoch": 0.7447354904982023, + "grad_norm": 2.327270984649658, + "learning_rate": 4.20631899505139e-06, + "loss": 1.3933, + "step": 725 + }, + { + "epoch": 0.7457627118644068, + "grad_norm": 2.3734512329101562, + "learning_rate": 4.204415683288923e-06, + "loss": 1.5109, + "step": 726 + }, + { + "epoch": 0.7467899332306112, + "grad_norm": 2.3631174564361572, + "learning_rate": 4.2025123715264565e-06, + "loss": 1.474, + "step": 727 + }, + { + "epoch": 0.7478171545968156, + "grad_norm": 2.306068181991577, + "learning_rate": 4.20060905976399e-06, + "loss": 1.42, + "step": 728 + }, + { + "epoch": 0.74884437596302, + "grad_norm": 2.391889810562134, + "learning_rate": 4.198705748001523e-06, + "loss": 1.5178, + "step": 729 + }, + { + "epoch": 0.7498715973292245, + "grad_norm": 2.3151705265045166, + "learning_rate": 4.1968024362390566e-06, + "loss": 1.4808, + "step": 730 + }, + { + "epoch": 0.7498715973292245, + "eval_loss": 1.8157070875167847, + "eval_runtime": 37.8799, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 730 + }, + { + "epoch": 0.7508988186954288, + "grad_norm": 2.336479902267456, + "learning_rate": 4.19489912447659e-06, + "loss": 1.5438, + "step": 731 + }, + { + "epoch": 0.7519260400616333, + "grad_norm": 2.338242292404175, + "learning_rate": 4.192995812714123e-06, + "loss": 1.4869, + "step": 732 + }, + { + "epoch": 0.7529532614278377, + "grad_norm": 2.268650770187378, + "learning_rate": 4.191092500951656e-06, + "loss": 1.4787, + "step": 733 + }, + { + "epoch": 0.7539804827940421, + "grad_norm": 2.3010005950927734, + "learning_rate": 4.189189189189189e-06, + "loss": 1.4711, + "step": 734 + }, + { + "epoch": 0.7550077041602465, + "grad_norm": 2.54770827293396, + "learning_rate": 4.1872858774267225e-06, + "loss": 1.5351, + "step": 735 + }, + { + "epoch": 0.7560349255264509, + "grad_norm": 2.3412787914276123, + "learning_rate": 4.185382565664256e-06, + "loss": 1.4655, + "step": 736 + }, + { + "epoch": 0.7570621468926554, + "grad_norm": 2.266413927078247, + "learning_rate": 4.183479253901789e-06, + "loss": 1.4402, + "step": 737 + }, + { + "epoch": 0.7580893682588598, + "grad_norm": 2.2785308361053467, + "learning_rate": 4.181575942139323e-06, + "loss": 1.4874, + "step": 738 + }, + { + "epoch": 0.7591165896250642, + "grad_norm": 2.2710752487182617, + "learning_rate": 4.179672630376856e-06, + "loss": 1.4644, + "step": 739 + }, + { + "epoch": 0.7601438109912686, + "grad_norm": 2.2460498809814453, + "learning_rate": 4.177769318614389e-06, + "loss": 1.5267, + "step": 740 + }, + { + "epoch": 0.7601438109912686, + "eval_loss": 1.8187947273254395, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 740 + }, + { + "epoch": 0.7611710323574731, + "grad_norm": 2.3589370250701904, + "learning_rate": 4.175866006851923e-06, + "loss": 1.5115, + "step": 741 + }, + { + "epoch": 0.7621982537236774, + "grad_norm": 2.2818758487701416, + "learning_rate": 4.173962695089456e-06, + "loss": 1.4945, + "step": 742 + }, + { + "epoch": 0.7632254750898819, + "grad_norm": 2.338085889816284, + "learning_rate": 4.1720593833269894e-06, + "loss": 1.5076, + "step": 743 + }, + { + "epoch": 0.7642526964560863, + "grad_norm": 2.3874809741973877, + "learning_rate": 4.170156071564523e-06, + "loss": 1.5665, + "step": 744 + }, + { + "epoch": 0.7652799178222907, + "grad_norm": 2.2418739795684814, + "learning_rate": 4.168252759802056e-06, + "loss": 1.4827, + "step": 745 + }, + { + "epoch": 0.7663071391884951, + "grad_norm": 2.441315174102783, + "learning_rate": 4.1663494480395895e-06, + "loss": 1.5171, + "step": 746 + }, + { + "epoch": 0.7673343605546995, + "grad_norm": 2.3201963901519775, + "learning_rate": 4.164446136277123e-06, + "loss": 1.5153, + "step": 747 + }, + { + "epoch": 0.768361581920904, + "grad_norm": 2.3395135402679443, + "learning_rate": 4.162542824514656e-06, + "loss": 1.4911, + "step": 748 + }, + { + "epoch": 0.7693888032871083, + "grad_norm": 2.3455941677093506, + "learning_rate": 4.16063951275219e-06, + "loss": 1.4493, + "step": 749 + }, + { + "epoch": 0.7704160246533128, + "grad_norm": 2.3737382888793945, + "learning_rate": 4.158736200989722e-06, + "loss": 1.4893, + "step": 750 + }, + { + "epoch": 0.7704160246533128, + "eval_loss": 1.8153141736984253, + "eval_runtime": 37.9216, + "eval_samples_per_second": 26.37, + "eval_steps_per_second": 1.108, + "step": 750 + }, + { + "epoch": 0.7714432460195172, + "grad_norm": 2.503803014755249, + "learning_rate": 4.1568328892272555e-06, + "loss": 1.499, + "step": 751 + }, + { + "epoch": 0.7724704673857217, + "grad_norm": 2.469083547592163, + "learning_rate": 4.154929577464789e-06, + "loss": 1.4984, + "step": 752 + }, + { + "epoch": 0.773497688751926, + "grad_norm": 2.387991428375244, + "learning_rate": 4.153026265702322e-06, + "loss": 1.4283, + "step": 753 + }, + { + "epoch": 0.7745249101181304, + "grad_norm": 2.3355393409729004, + "learning_rate": 4.1511229539398556e-06, + "loss": 1.4869, + "step": 754 + }, + { + "epoch": 0.7755521314843349, + "grad_norm": 2.629345417022705, + "learning_rate": 4.149219642177389e-06, + "loss": 1.4885, + "step": 755 + }, + { + "epoch": 0.7765793528505393, + "grad_norm": 2.458944797515869, + "learning_rate": 4.147316330414922e-06, + "loss": 1.415, + "step": 756 + }, + { + "epoch": 0.7776065742167437, + "grad_norm": 2.40051531791687, + "learning_rate": 4.145413018652456e-06, + "loss": 1.5129, + "step": 757 + }, + { + "epoch": 0.7786337955829481, + "grad_norm": 2.5249011516571045, + "learning_rate": 4.143509706889989e-06, + "loss": 1.5002, + "step": 758 + }, + { + "epoch": 0.7796610169491526, + "grad_norm": 2.413422107696533, + "learning_rate": 4.141606395127522e-06, + "loss": 1.4837, + "step": 759 + }, + { + "epoch": 0.7806882383153569, + "grad_norm": 2.3438029289245605, + "learning_rate": 4.139703083365056e-06, + "loss": 1.5213, + "step": 760 + }, + { + "epoch": 0.7806882383153569, + "eval_loss": 1.8037596940994263, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 760 + }, + { + "epoch": 0.7817154596815614, + "grad_norm": 2.326828718185425, + "learning_rate": 4.137799771602588e-06, + "loss": 1.5031, + "step": 761 + }, + { + "epoch": 0.7827426810477658, + "grad_norm": 2.4902493953704834, + "learning_rate": 4.135896459840122e-06, + "loss": 1.4602, + "step": 762 + }, + { + "epoch": 0.7837699024139703, + "grad_norm": 2.488004684448242, + "learning_rate": 4.133993148077656e-06, + "loss": 1.4764, + "step": 763 + }, + { + "epoch": 0.7847971237801746, + "grad_norm": 2.5515761375427246, + "learning_rate": 4.132089836315189e-06, + "loss": 1.4946, + "step": 764 + }, + { + "epoch": 0.785824345146379, + "grad_norm": 2.556278705596924, + "learning_rate": 4.1301865245527226e-06, + "loss": 1.4338, + "step": 765 + }, + { + "epoch": 0.7868515665125835, + "grad_norm": 2.3017969131469727, + "learning_rate": 4.128283212790256e-06, + "loss": 1.4053, + "step": 766 + }, + { + "epoch": 0.7878787878787878, + "grad_norm": 2.4371304512023926, + "learning_rate": 4.1263799010277884e-06, + "loss": 1.5238, + "step": 767 + }, + { + "epoch": 0.7889060092449923, + "grad_norm": 2.6256046295166016, + "learning_rate": 4.124476589265322e-06, + "loss": 1.4672, + "step": 768 + }, + { + "epoch": 0.7899332306111967, + "grad_norm": 2.3725669384002686, + "learning_rate": 4.122573277502855e-06, + "loss": 1.4677, + "step": 769 + }, + { + "epoch": 0.7909604519774012, + "grad_norm": 2.2928454875946045, + "learning_rate": 4.1206699657403885e-06, + "loss": 1.4927, + "step": 770 + }, + { + "epoch": 0.7909604519774012, + "eval_loss": 1.792726755142212, + "eval_runtime": 37.9273, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 770 + }, + { + "epoch": 0.7919876733436055, + "grad_norm": 2.315674066543579, + "learning_rate": 4.118766653977922e-06, + "loss": 1.4632, + "step": 771 + }, + { + "epoch": 0.79301489470981, + "grad_norm": 2.418849468231201, + "learning_rate": 4.116863342215455e-06, + "loss": 1.446, + "step": 772 + }, + { + "epoch": 0.7940421160760144, + "grad_norm": 2.3209879398345947, + "learning_rate": 4.114960030452989e-06, + "loss": 1.4755, + "step": 773 + }, + { + "epoch": 0.7950693374422187, + "grad_norm": 2.2322115898132324, + "learning_rate": 4.113056718690522e-06, + "loss": 1.4149, + "step": 774 + }, + { + "epoch": 0.7960965588084232, + "grad_norm": 2.6517558097839355, + "learning_rate": 4.111153406928055e-06, + "loss": 1.5323, + "step": 775 + }, + { + "epoch": 0.7971237801746276, + "grad_norm": 2.4602737426757812, + "learning_rate": 4.109250095165589e-06, + "loss": 1.5179, + "step": 776 + }, + { + "epoch": 0.7981510015408321, + "grad_norm": 2.4015536308288574, + "learning_rate": 4.107346783403122e-06, + "loss": 1.4711, + "step": 777 + }, + { + "epoch": 0.7991782229070364, + "grad_norm": 2.4367337226867676, + "learning_rate": 4.105443471640655e-06, + "loss": 1.4425, + "step": 778 + }, + { + "epoch": 0.8002054442732409, + "grad_norm": 2.5161020755767822, + "learning_rate": 4.103540159878188e-06, + "loss": 1.4832, + "step": 779 + }, + { + "epoch": 0.8012326656394453, + "grad_norm": 2.483154773712158, + "learning_rate": 4.101636848115721e-06, + "loss": 1.4508, + "step": 780 + }, + { + "epoch": 0.8012326656394453, + "eval_loss": 1.792397379875183, + "eval_runtime": 37.8937, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 780 + }, + { + "epoch": 0.8022598870056498, + "grad_norm": 2.4649806022644043, + "learning_rate": 4.099733536353255e-06, + "loss": 1.4919, + "step": 781 + }, + { + "epoch": 0.8032871083718541, + "grad_norm": 2.578634262084961, + "learning_rate": 4.097830224590788e-06, + "loss": 1.565, + "step": 782 + }, + { + "epoch": 0.8043143297380585, + "grad_norm": 2.444427967071533, + "learning_rate": 4.095926912828322e-06, + "loss": 1.5309, + "step": 783 + }, + { + "epoch": 0.805341551104263, + "grad_norm": 2.32051420211792, + "learning_rate": 4.094023601065855e-06, + "loss": 1.4777, + "step": 784 + }, + { + "epoch": 0.8063687724704673, + "grad_norm": 2.3207180500030518, + "learning_rate": 4.092120289303388e-06, + "loss": 1.4863, + "step": 785 + }, + { + "epoch": 0.8073959938366718, + "grad_norm": 2.1404120922088623, + "learning_rate": 4.0902169775409215e-06, + "loss": 1.437, + "step": 786 + }, + { + "epoch": 0.8084232152028762, + "grad_norm": 2.2980856895446777, + "learning_rate": 4.088313665778455e-06, + "loss": 1.4571, + "step": 787 + }, + { + "epoch": 0.8094504365690807, + "grad_norm": 2.2555243968963623, + "learning_rate": 4.086410354015988e-06, + "loss": 1.4451, + "step": 788 + }, + { + "epoch": 0.810477657935285, + "grad_norm": 2.3837077617645264, + "learning_rate": 4.0845070422535216e-06, + "loss": 1.4641, + "step": 789 + }, + { + "epoch": 0.8115048793014895, + "grad_norm": 2.5199060440063477, + "learning_rate": 4.082603730491055e-06, + "loss": 1.5048, + "step": 790 + }, + { + "epoch": 0.8115048793014895, + "eval_loss": 1.8019492626190186, + "eval_runtime": 37.8805, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 790 + }, + { + "epoch": 0.8125321006676939, + "grad_norm": 2.4010138511657715, + "learning_rate": 4.080700418728588e-06, + "loss": 1.4881, + "step": 791 + }, + { + "epoch": 0.8135593220338984, + "grad_norm": 2.342094898223877, + "learning_rate": 4.078797106966122e-06, + "loss": 1.3792, + "step": 792 + }, + { + "epoch": 0.8145865434001027, + "grad_norm": 2.276772975921631, + "learning_rate": 4.076893795203655e-06, + "loss": 1.4539, + "step": 793 + }, + { + "epoch": 0.8156137647663071, + "grad_norm": 2.569275140762329, + "learning_rate": 4.074990483441188e-06, + "loss": 1.5059, + "step": 794 + }, + { + "epoch": 0.8166409861325116, + "grad_norm": 2.345761299133301, + "learning_rate": 4.073087171678721e-06, + "loss": 1.494, + "step": 795 + }, + { + "epoch": 0.8176682074987159, + "grad_norm": 2.4425504207611084, + "learning_rate": 4.071183859916254e-06, + "loss": 1.471, + "step": 796 + }, + { + "epoch": 0.8186954288649204, + "grad_norm": 2.5758635997772217, + "learning_rate": 4.069280548153788e-06, + "loss": 1.4378, + "step": 797 + }, + { + "epoch": 0.8197226502311248, + "grad_norm": 2.4320778846740723, + "learning_rate": 4.067377236391321e-06, + "loss": 1.474, + "step": 798 + }, + { + "epoch": 0.8207498715973293, + "grad_norm": 2.466978073120117, + "learning_rate": 4.065473924628854e-06, + "loss": 1.4997, + "step": 799 + }, + { + "epoch": 0.8217770929635336, + "grad_norm": 2.462489128112793, + "learning_rate": 4.063570612866388e-06, + "loss": 1.5, + "step": 800 + }, + { + "epoch": 0.8217770929635336, + "eval_loss": 1.783312201499939, + "eval_runtime": 37.8832, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 800 + }, + { + "epoch": 0.8228043143297381, + "grad_norm": 2.367455005645752, + "learning_rate": 4.061667301103921e-06, + "loss": 1.4658, + "step": 801 + }, + { + "epoch": 0.8238315356959425, + "grad_norm": 2.292250394821167, + "learning_rate": 4.0597639893414544e-06, + "loss": 1.476, + "step": 802 + }, + { + "epoch": 0.8248587570621468, + "grad_norm": 2.3460025787353516, + "learning_rate": 4.057860677578988e-06, + "loss": 1.4932, + "step": 803 + }, + { + "epoch": 0.8258859784283513, + "grad_norm": 2.3708906173706055, + "learning_rate": 4.055957365816521e-06, + "loss": 1.4396, + "step": 804 + }, + { + "epoch": 0.8269131997945557, + "grad_norm": 2.353513479232788, + "learning_rate": 4.0540540540540545e-06, + "loss": 1.4285, + "step": 805 + }, + { + "epoch": 0.8279404211607602, + "grad_norm": 2.319577932357788, + "learning_rate": 4.052150742291588e-06, + "loss": 1.4326, + "step": 806 + }, + { + "epoch": 0.8289676425269645, + "grad_norm": 2.450364112854004, + "learning_rate": 4.050247430529121e-06, + "loss": 1.5109, + "step": 807 + }, + { + "epoch": 0.829994863893169, + "grad_norm": 2.393907070159912, + "learning_rate": 4.048344118766655e-06, + "loss": 1.4948, + "step": 808 + }, + { + "epoch": 0.8310220852593734, + "grad_norm": 2.336002826690674, + "learning_rate": 4.046440807004188e-06, + "loss": 1.4851, + "step": 809 + }, + { + "epoch": 0.8320493066255779, + "grad_norm": 2.281139850616455, + "learning_rate": 4.044537495241721e-06, + "loss": 1.4445, + "step": 810 + }, + { + "epoch": 0.8320493066255779, + "eval_loss": 1.784231185913086, + "eval_runtime": 37.6614, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 810 + }, + { + "epoch": 0.8330765279917822, + "grad_norm": 2.4848175048828125, + "learning_rate": 4.042634183479255e-06, + "loss": 1.4982, + "step": 811 + }, + { + "epoch": 0.8341037493579867, + "grad_norm": 2.263347864151001, + "learning_rate": 4.040730871716787e-06, + "loss": 1.4795, + "step": 812 + }, + { + "epoch": 0.8351309707241911, + "grad_norm": 2.394005060195923, + "learning_rate": 4.038827559954321e-06, + "loss": 1.5231, + "step": 813 + }, + { + "epoch": 0.8361581920903954, + "grad_norm": 2.5051357746124268, + "learning_rate": 4.036924248191854e-06, + "loss": 1.4886, + "step": 814 + }, + { + "epoch": 0.8371854134565999, + "grad_norm": 2.3580374717712402, + "learning_rate": 4.035020936429387e-06, + "loss": 1.4713, + "step": 815 + }, + { + "epoch": 0.8382126348228043, + "grad_norm": 2.3179264068603516, + "learning_rate": 4.033117624666921e-06, + "loss": 1.4369, + "step": 816 + }, + { + "epoch": 0.8392398561890088, + "grad_norm": 2.269911289215088, + "learning_rate": 4.031214312904454e-06, + "loss": 1.4326, + "step": 817 + }, + { + "epoch": 0.8402670775552131, + "grad_norm": 2.3177807331085205, + "learning_rate": 4.029311001141987e-06, + "loss": 1.4096, + "step": 818 + }, + { + "epoch": 0.8412942989214176, + "grad_norm": 2.394324541091919, + "learning_rate": 4.027407689379521e-06, + "loss": 1.4647, + "step": 819 + }, + { + "epoch": 0.842321520287622, + "grad_norm": 2.288536548614502, + "learning_rate": 4.025504377617054e-06, + "loss": 1.4472, + "step": 820 + }, + { + "epoch": 0.842321520287622, + "eval_loss": 1.7811495065689087, + "eval_runtime": 37.8132, + "eval_samples_per_second": 26.446, + "eval_steps_per_second": 1.111, + "step": 820 + }, + { + "epoch": 0.8433487416538265, + "grad_norm": 2.376574754714966, + "learning_rate": 4.0236010658545875e-06, + "loss": 1.4918, + "step": 821 + }, + { + "epoch": 0.8443759630200308, + "grad_norm": 2.5850327014923096, + "learning_rate": 4.02169775409212e-06, + "loss": 1.5342, + "step": 822 + }, + { + "epoch": 0.8454031843862352, + "grad_norm": 2.247206211090088, + "learning_rate": 4.019794442329653e-06, + "loss": 1.4177, + "step": 823 + }, + { + "epoch": 0.8464304057524397, + "grad_norm": 2.4411795139312744, + "learning_rate": 4.0178911305671876e-06, + "loss": 1.4992, + "step": 824 + }, + { + "epoch": 0.847457627118644, + "grad_norm": 2.3377397060394287, + "learning_rate": 4.015987818804721e-06, + "loss": 1.4359, + "step": 825 + }, + { + "epoch": 0.8484848484848485, + "grad_norm": 2.329857349395752, + "learning_rate": 4.014084507042254e-06, + "loss": 1.4096, + "step": 826 + }, + { + "epoch": 0.8495120698510529, + "grad_norm": 2.563685178756714, + "learning_rate": 4.012181195279788e-06, + "loss": 1.5174, + "step": 827 + }, + { + "epoch": 0.8505392912172574, + "grad_norm": 2.4521079063415527, + "learning_rate": 4.010277883517321e-06, + "loss": 1.5028, + "step": 828 + }, + { + "epoch": 0.8515665125834617, + "grad_norm": 2.4091477394104004, + "learning_rate": 4.0083745717548535e-06, + "loss": 1.4615, + "step": 829 + }, + { + "epoch": 0.8525937339496662, + "grad_norm": 2.4010677337646484, + "learning_rate": 4.006471259992387e-06, + "loss": 1.4834, + "step": 830 + }, + { + "epoch": 0.8525937339496662, + "eval_loss": 1.7797678709030151, + "eval_runtime": 37.9378, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 830 + }, + { + "epoch": 0.8536209553158706, + "grad_norm": 2.3712635040283203, + "learning_rate": 4.00456794822992e-06, + "loss": 1.5196, + "step": 831 + }, + { + "epoch": 0.8546481766820749, + "grad_norm": 2.4096601009368896, + "learning_rate": 4.002664636467454e-06, + "loss": 1.4769, + "step": 832 + }, + { + "epoch": 0.8556753980482794, + "grad_norm": 2.408466100692749, + "learning_rate": 4.000761324704987e-06, + "loss": 1.4268, + "step": 833 + }, + { + "epoch": 0.8567026194144838, + "grad_norm": 2.6857047080993652, + "learning_rate": 3.99885801294252e-06, + "loss": 1.492, + "step": 834 + }, + { + "epoch": 0.8577298407806883, + "grad_norm": 2.297550916671753, + "learning_rate": 3.996954701180054e-06, + "loss": 1.4184, + "step": 835 + }, + { + "epoch": 0.8587570621468926, + "grad_norm": 2.423224925994873, + "learning_rate": 3.995051389417587e-06, + "loss": 1.4337, + "step": 836 + }, + { + "epoch": 0.8597842835130971, + "grad_norm": 2.260493040084839, + "learning_rate": 3.9931480776551204e-06, + "loss": 1.5267, + "step": 837 + }, + { + "epoch": 0.8608115048793015, + "grad_norm": 2.3803012371063232, + "learning_rate": 3.991244765892654e-06, + "loss": 1.4426, + "step": 838 + }, + { + "epoch": 0.861838726245506, + "grad_norm": 2.1425721645355225, + "learning_rate": 3.989341454130186e-06, + "loss": 1.4148, + "step": 839 + }, + { + "epoch": 0.8628659476117103, + "grad_norm": 2.316624641418457, + "learning_rate": 3.98743814236772e-06, + "loss": 1.4544, + "step": 840 + }, + { + "epoch": 0.8628659476117103, + "eval_loss": 1.7824597358703613, + "eval_runtime": 37.9676, + "eval_samples_per_second": 26.338, + "eval_steps_per_second": 1.106, + "step": 840 + }, + { + "epoch": 0.8638931689779148, + "grad_norm": 2.3723156452178955, + "learning_rate": 3.985534830605253e-06, + "loss": 1.473, + "step": 841 + }, + { + "epoch": 0.8649203903441192, + "grad_norm": 2.5167760848999023, + "learning_rate": 3.983631518842786e-06, + "loss": 1.4919, + "step": 842 + }, + { + "epoch": 0.8659476117103235, + "grad_norm": 2.3018670082092285, + "learning_rate": 3.98172820708032e-06, + "loss": 1.447, + "step": 843 + }, + { + "epoch": 0.866974833076528, + "grad_norm": 2.420483112335205, + "learning_rate": 3.979824895317854e-06, + "loss": 1.4159, + "step": 844 + }, + { + "epoch": 0.8680020544427324, + "grad_norm": 2.39443039894104, + "learning_rate": 3.977921583555387e-06, + "loss": 1.4984, + "step": 845 + }, + { + "epoch": 0.8690292758089369, + "grad_norm": 2.2953920364379883, + "learning_rate": 3.97601827179292e-06, + "loss": 1.4297, + "step": 846 + }, + { + "epoch": 0.8700564971751412, + "grad_norm": 2.2919528484344482, + "learning_rate": 3.974114960030453e-06, + "loss": 1.4728, + "step": 847 + }, + { + "epoch": 0.8710837185413457, + "grad_norm": 2.327914237976074, + "learning_rate": 3.9722116482679866e-06, + "loss": 1.4584, + "step": 848 + }, + { + "epoch": 0.8721109399075501, + "grad_norm": 2.305720567703247, + "learning_rate": 3.97030833650552e-06, + "loss": 1.4688, + "step": 849 + }, + { + "epoch": 0.8731381612737545, + "grad_norm": 2.2969236373901367, + "learning_rate": 3.968405024743053e-06, + "loss": 1.3972, + "step": 850 + }, + { + "epoch": 0.8731381612737545, + "eval_loss": 1.775634765625, + "eval_runtime": 37.9916, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 850 + }, + { + "epoch": 0.8741653826399589, + "grad_norm": 2.314387083053589, + "learning_rate": 3.966501712980587e-06, + "loss": 1.4476, + "step": 851 + }, + { + "epoch": 0.8751926040061633, + "grad_norm": 2.3271584510803223, + "learning_rate": 3.96459840121812e-06, + "loss": 1.4615, + "step": 852 + }, + { + "epoch": 0.8762198253723678, + "grad_norm": 2.494795322418213, + "learning_rate": 3.962695089455653e-06, + "loss": 1.5062, + "step": 853 + }, + { + "epoch": 0.8772470467385721, + "grad_norm": 2.4132444858551025, + "learning_rate": 3.960791777693187e-06, + "loss": 1.4861, + "step": 854 + }, + { + "epoch": 0.8782742681047766, + "grad_norm": 2.3869900703430176, + "learning_rate": 3.95888846593072e-06, + "loss": 1.4923, + "step": 855 + }, + { + "epoch": 0.879301489470981, + "grad_norm": 2.5167367458343506, + "learning_rate": 3.956985154168253e-06, + "loss": 1.5137, + "step": 856 + }, + { + "epoch": 0.8803287108371854, + "grad_norm": 2.510361671447754, + "learning_rate": 3.955081842405786e-06, + "loss": 1.5414, + "step": 857 + }, + { + "epoch": 0.8813559322033898, + "grad_norm": 2.554427146911621, + "learning_rate": 3.953178530643319e-06, + "loss": 1.5569, + "step": 858 + }, + { + "epoch": 0.8823831535695943, + "grad_norm": 2.306509256362915, + "learning_rate": 3.951275218880853e-06, + "loss": 1.3982, + "step": 859 + }, + { + "epoch": 0.8834103749357987, + "grad_norm": 2.3805363178253174, + "learning_rate": 3.949371907118386e-06, + "loss": 1.4977, + "step": 860 + }, + { + "epoch": 0.8834103749357987, + "eval_loss": 1.7882694005966187, + "eval_runtime": 38.0072, + "eval_samples_per_second": 26.311, + "eval_steps_per_second": 1.105, + "step": 860 + }, + { + "epoch": 0.884437596302003, + "grad_norm": 2.5413448810577393, + "learning_rate": 3.9474685953559194e-06, + "loss": 1.472, + "step": 861 + }, + { + "epoch": 0.8854648176682075, + "grad_norm": 2.484834909439087, + "learning_rate": 3.945565283593453e-06, + "loss": 1.4597, + "step": 862 + }, + { + "epoch": 0.8864920390344119, + "grad_norm": 2.6060585975646973, + "learning_rate": 3.943661971830986e-06, + "loss": 1.509, + "step": 863 + }, + { + "epoch": 0.8875192604006163, + "grad_norm": 2.6536927223205566, + "learning_rate": 3.9417586600685195e-06, + "loss": 1.4836, + "step": 864 + }, + { + "epoch": 0.8885464817668207, + "grad_norm": 2.5198960304260254, + "learning_rate": 3.939855348306053e-06, + "loss": 1.486, + "step": 865 + }, + { + "epoch": 0.8895737031330252, + "grad_norm": 2.335204839706421, + "learning_rate": 3.937952036543586e-06, + "loss": 1.4701, + "step": 866 + }, + { + "epoch": 0.8906009244992296, + "grad_norm": 2.7058098316192627, + "learning_rate": 3.93604872478112e-06, + "loss": 1.4798, + "step": 867 + }, + { + "epoch": 0.891628145865434, + "grad_norm": 2.4798922538757324, + "learning_rate": 3.934145413018653e-06, + "loss": 1.4364, + "step": 868 + }, + { + "epoch": 0.8926553672316384, + "grad_norm": 2.3090760707855225, + "learning_rate": 3.932242101256186e-06, + "loss": 1.4405, + "step": 869 + }, + { + "epoch": 0.8936825885978429, + "grad_norm": 2.4251503944396973, + "learning_rate": 3.93033878949372e-06, + "loss": 1.4113, + "step": 870 + }, + { + "epoch": 0.8936825885978429, + "eval_loss": 1.794334888458252, + "eval_runtime": 38.0197, + "eval_samples_per_second": 26.302, + "eval_steps_per_second": 1.105, + "step": 870 + }, + { + "epoch": 0.8947098099640473, + "grad_norm": 2.594945192337036, + "learning_rate": 3.928435477731253e-06, + "loss": 1.502, + "step": 871 + }, + { + "epoch": 0.8957370313302516, + "grad_norm": 2.4224843978881836, + "learning_rate": 3.9265321659687864e-06, + "loss": 1.5005, + "step": 872 + }, + { + "epoch": 0.8967642526964561, + "grad_norm": 2.3968186378479004, + "learning_rate": 3.924628854206319e-06, + "loss": 1.481, + "step": 873 + }, + { + "epoch": 0.8977914740626605, + "grad_norm": 2.4923524856567383, + "learning_rate": 3.922725542443852e-06, + "loss": 1.4259, + "step": 874 + }, + { + "epoch": 0.8988186954288649, + "grad_norm": 2.5386104583740234, + "learning_rate": 3.920822230681386e-06, + "loss": 1.4497, + "step": 875 + }, + { + "epoch": 0.8998459167950693, + "grad_norm": 2.4014909267425537, + "learning_rate": 3.918918918918919e-06, + "loss": 1.4816, + "step": 876 + }, + { + "epoch": 0.9008731381612738, + "grad_norm": 2.4019317626953125, + "learning_rate": 3.917015607156452e-06, + "loss": 1.4669, + "step": 877 + }, + { + "epoch": 0.9019003595274782, + "grad_norm": 2.5408380031585693, + "learning_rate": 3.915112295393986e-06, + "loss": 1.5343, + "step": 878 + }, + { + "epoch": 0.9029275808936826, + "grad_norm": 2.364053964614868, + "learning_rate": 3.913208983631519e-06, + "loss": 1.5229, + "step": 879 + }, + { + "epoch": 0.903954802259887, + "grad_norm": 2.307412624359131, + "learning_rate": 3.9113056718690525e-06, + "loss": 1.4444, + "step": 880 + }, + { + "epoch": 0.903954802259887, + "eval_loss": 1.7910975217819214, + "eval_runtime": 38.0659, + "eval_samples_per_second": 26.27, + "eval_steps_per_second": 1.103, + "step": 880 + }, + { + "epoch": 0.9049820236260914, + "grad_norm": 2.496135950088501, + "learning_rate": 3.909402360106586e-06, + "loss": 1.4853, + "step": 881 + }, + { + "epoch": 0.9060092449922958, + "grad_norm": 2.4825589656829834, + "learning_rate": 3.907499048344119e-06, + "loss": 1.4886, + "step": 882 + }, + { + "epoch": 0.9070364663585002, + "grad_norm": 2.3458824157714844, + "learning_rate": 3.905595736581652e-06, + "loss": 1.4882, + "step": 883 + }, + { + "epoch": 0.9080636877247047, + "grad_norm": 2.2699403762817383, + "learning_rate": 3.903692424819186e-06, + "loss": 1.4417, + "step": 884 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 2.348593235015869, + "learning_rate": 3.901789113056719e-06, + "loss": 1.4991, + "step": 885 + }, + { + "epoch": 0.9101181304571135, + "grad_norm": 2.3292129039764404, + "learning_rate": 3.899885801294253e-06, + "loss": 1.4752, + "step": 886 + }, + { + "epoch": 0.9111453518233179, + "grad_norm": 2.487183094024658, + "learning_rate": 3.897982489531786e-06, + "loss": 1.5317, + "step": 887 + }, + { + "epoch": 0.9121725731895224, + "grad_norm": 2.3300845623016357, + "learning_rate": 3.896079177769319e-06, + "loss": 1.4964, + "step": 888 + }, + { + "epoch": 0.9131997945557268, + "grad_norm": 2.3292312622070312, + "learning_rate": 3.894175866006853e-06, + "loss": 1.5221, + "step": 889 + }, + { + "epoch": 0.9142270159219312, + "grad_norm": 2.3556439876556396, + "learning_rate": 3.892272554244385e-06, + "loss": 1.4471, + "step": 890 + }, + { + "epoch": 0.9142270159219312, + "eval_loss": 1.7807039022445679, + "eval_runtime": 38.0614, + "eval_samples_per_second": 26.273, + "eval_steps_per_second": 1.103, + "step": 890 + }, + { + "epoch": 0.9152542372881356, + "grad_norm": 2.3302760124206543, + "learning_rate": 3.890369242481919e-06, + "loss": 1.5179, + "step": 891 + }, + { + "epoch": 0.91628145865434, + "grad_norm": 2.327072858810425, + "learning_rate": 3.888465930719452e-06, + "loss": 1.4474, + "step": 892 + }, + { + "epoch": 0.9173086800205444, + "grad_norm": 2.2773101329803467, + "learning_rate": 3.886562618956985e-06, + "loss": 1.4783, + "step": 893 + }, + { + "epoch": 0.9183359013867488, + "grad_norm": 2.2717974185943604, + "learning_rate": 3.884659307194519e-06, + "loss": 1.4062, + "step": 894 + }, + { + "epoch": 0.9193631227529533, + "grad_norm": 2.3350367546081543, + "learning_rate": 3.882755995432052e-06, + "loss": 1.483, + "step": 895 + }, + { + "epoch": 0.9203903441191577, + "grad_norm": 2.302811622619629, + "learning_rate": 3.8808526836695854e-06, + "loss": 1.4407, + "step": 896 + }, + { + "epoch": 0.9214175654853621, + "grad_norm": 2.4031271934509277, + "learning_rate": 3.878949371907119e-06, + "loss": 1.4331, + "step": 897 + }, + { + "epoch": 0.9224447868515665, + "grad_norm": 2.4806716442108154, + "learning_rate": 3.877046060144652e-06, + "loss": 1.4341, + "step": 898 + }, + { + "epoch": 0.923472008217771, + "grad_norm": 2.4243357181549072, + "learning_rate": 3.8751427483821855e-06, + "loss": 1.4701, + "step": 899 + }, + { + "epoch": 0.9244992295839753, + "grad_norm": 2.4953269958496094, + "learning_rate": 3.873239436619718e-06, + "loss": 1.4287, + "step": 900 + }, + { + "epoch": 0.9244992295839753, + "eval_loss": 1.7872728109359741, + "eval_runtime": 38.0906, + "eval_samples_per_second": 26.253, + "eval_steps_per_second": 1.103, + "step": 900 + }, + { + "epoch": 0.9255264509501797, + "grad_norm": 2.307490587234497, + "learning_rate": 3.871336124857251e-06, + "loss": 1.4693, + "step": 901 + }, + { + "epoch": 0.9265536723163842, + "grad_norm": 2.4745755195617676, + "learning_rate": 3.869432813094785e-06, + "loss": 1.4779, + "step": 902 + }, + { + "epoch": 0.9275808936825886, + "grad_norm": 2.367715358734131, + "learning_rate": 3.867529501332318e-06, + "loss": 1.4771, + "step": 903 + }, + { + "epoch": 0.928608115048793, + "grad_norm": 2.388549566268921, + "learning_rate": 3.8656261895698515e-06, + "loss": 1.4648, + "step": 904 + }, + { + "epoch": 0.9296353364149974, + "grad_norm": 2.39432954788208, + "learning_rate": 3.863722877807386e-06, + "loss": 1.4932, + "step": 905 + }, + { + "epoch": 0.9306625577812019, + "grad_norm": 2.3921315670013428, + "learning_rate": 3.861819566044919e-06, + "loss": 1.5312, + "step": 906 + }, + { + "epoch": 0.9316897791474062, + "grad_norm": 2.400240659713745, + "learning_rate": 3.859916254282452e-06, + "loss": 1.4651, + "step": 907 + }, + { + "epoch": 0.9327170005136107, + "grad_norm": 2.394559860229492, + "learning_rate": 3.858012942519985e-06, + "loss": 1.4759, + "step": 908 + }, + { + "epoch": 0.9337442218798151, + "grad_norm": 2.3365273475646973, + "learning_rate": 3.856109630757518e-06, + "loss": 1.4502, + "step": 909 + }, + { + "epoch": 0.9347714432460195, + "grad_norm": 2.465691328048706, + "learning_rate": 3.854206318995052e-06, + "loss": 1.5036, + "step": 910 + }, + { + "epoch": 0.9347714432460195, + "eval_loss": 1.7682362794876099, + "eval_runtime": 37.6624, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 910 + }, + { + "epoch": 0.9357986646122239, + "grad_norm": 2.3376758098602295, + "learning_rate": 3.852303007232585e-06, + "loss": 1.4342, + "step": 911 + }, + { + "epoch": 0.9368258859784283, + "grad_norm": 2.345855236053467, + "learning_rate": 3.850399695470118e-06, + "loss": 1.4444, + "step": 912 + }, + { + "epoch": 0.9378531073446328, + "grad_norm": 2.4444501399993896, + "learning_rate": 3.848496383707652e-06, + "loss": 1.4315, + "step": 913 + }, + { + "epoch": 0.9388803287108372, + "grad_norm": 2.4896373748779297, + "learning_rate": 3.846593071945185e-06, + "loss": 1.4228, + "step": 914 + }, + { + "epoch": 0.9399075500770416, + "grad_norm": 2.3970463275909424, + "learning_rate": 3.8446897601827185e-06, + "loss": 1.5283, + "step": 915 + }, + { + "epoch": 0.940934771443246, + "grad_norm": 2.5194883346557617, + "learning_rate": 3.842786448420252e-06, + "loss": 1.5445, + "step": 916 + }, + { + "epoch": 0.9419619928094505, + "grad_norm": 2.446465253829956, + "learning_rate": 3.840883136657785e-06, + "loss": 1.5248, + "step": 917 + }, + { + "epoch": 0.9429892141756548, + "grad_norm": 2.348477840423584, + "learning_rate": 3.838979824895318e-06, + "loss": 1.402, + "step": 918 + }, + { + "epoch": 0.9440164355418593, + "grad_norm": 2.4511539936065674, + "learning_rate": 3.837076513132851e-06, + "loss": 1.4893, + "step": 919 + }, + { + "epoch": 0.9450436569080637, + "grad_norm": 2.3801777362823486, + "learning_rate": 3.8351732013703844e-06, + "loss": 1.4457, + "step": 920 + }, + { + "epoch": 0.9450436569080637, + "eval_loss": 1.7810813188552856, + "eval_runtime": 37.7647, + "eval_samples_per_second": 26.48, + "eval_steps_per_second": 1.112, + "step": 920 + }, + { + "epoch": 0.9460708782742681, + "grad_norm": 2.5260043144226074, + "learning_rate": 3.833269889607918e-06, + "loss": 1.4267, + "step": 921 + }, + { + "epoch": 0.9470980996404725, + "grad_norm": 2.270653009414673, + "learning_rate": 3.831366577845451e-06, + "loss": 1.4313, + "step": 922 + }, + { + "epoch": 0.9481253210066769, + "grad_norm": 2.5123250484466553, + "learning_rate": 3.8294632660829845e-06, + "loss": 1.4588, + "step": 923 + }, + { + "epoch": 0.9491525423728814, + "grad_norm": 2.4242913722991943, + "learning_rate": 3.827559954320518e-06, + "loss": 1.4659, + "step": 924 + }, + { + "epoch": 0.9501797637390857, + "grad_norm": 2.4337117671966553, + "learning_rate": 3.825656642558051e-06, + "loss": 1.4838, + "step": 925 + }, + { + "epoch": 0.9512069851052902, + "grad_norm": 2.3577871322631836, + "learning_rate": 3.823753330795585e-06, + "loss": 1.4674, + "step": 926 + }, + { + "epoch": 0.9522342064714946, + "grad_norm": 2.2574679851531982, + "learning_rate": 3.821850019033118e-06, + "loss": 1.4011, + "step": 927 + }, + { + "epoch": 0.9532614278376991, + "grad_norm": 2.38043475151062, + "learning_rate": 3.819946707270651e-06, + "loss": 1.459, + "step": 928 + }, + { + "epoch": 0.9542886492039034, + "grad_norm": 2.4249954223632812, + "learning_rate": 3.818043395508185e-06, + "loss": 1.4649, + "step": 929 + }, + { + "epoch": 0.9553158705701078, + "grad_norm": 2.4420788288116455, + "learning_rate": 3.816140083745718e-06, + "loss": 1.532, + "step": 930 + }, + { + "epoch": 0.9553158705701078, + "eval_loss": 1.7805145978927612, + "eval_runtime": 37.8606, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 930 + }, + { + "epoch": 0.9563430919363123, + "grad_norm": 2.391042470932007, + "learning_rate": 3.8142367719832514e-06, + "loss": 1.4365, + "step": 931 + }, + { + "epoch": 0.9573703133025167, + "grad_norm": 2.462303876876831, + "learning_rate": 3.8123334602207844e-06, + "loss": 1.4367, + "step": 932 + }, + { + "epoch": 0.9583975346687211, + "grad_norm": 2.330714702606201, + "learning_rate": 3.8104301484583177e-06, + "loss": 1.427, + "step": 933 + }, + { + "epoch": 0.9594247560349255, + "grad_norm": 2.3475441932678223, + "learning_rate": 3.808526836695851e-06, + "loss": 1.4836, + "step": 934 + }, + { + "epoch": 0.96045197740113, + "grad_norm": 2.4546470642089844, + "learning_rate": 3.8066235249333845e-06, + "loss": 1.4601, + "step": 935 + }, + { + "epoch": 0.9614791987673343, + "grad_norm": 2.4604640007019043, + "learning_rate": 3.804720213170918e-06, + "loss": 1.4616, + "step": 936 + }, + { + "epoch": 0.9625064201335388, + "grad_norm": 2.3087422847747803, + "learning_rate": 3.8028169014084508e-06, + "loss": 1.4816, + "step": 937 + }, + { + "epoch": 0.9635336414997432, + "grad_norm": 2.375377893447876, + "learning_rate": 3.800913589645984e-06, + "loss": 1.471, + "step": 938 + }, + { + "epoch": 0.9645608628659477, + "grad_norm": 2.358741283416748, + "learning_rate": 3.7990102778835175e-06, + "loss": 1.4754, + "step": 939 + }, + { + "epoch": 0.965588084232152, + "grad_norm": 2.3990139961242676, + "learning_rate": 3.797106966121051e-06, + "loss": 1.534, + "step": 940 + }, + { + "epoch": 0.965588084232152, + "eval_loss": 1.7745739221572876, + "eval_runtime": 37.8883, + "eval_samples_per_second": 26.393, + "eval_steps_per_second": 1.109, + "step": 940 + }, + { + "epoch": 0.9666153055983564, + "grad_norm": 2.417396068572998, + "learning_rate": 3.7952036543585842e-06, + "loss": 1.4994, + "step": 941 + }, + { + "epoch": 0.9676425269645609, + "grad_norm": 2.323125123977661, + "learning_rate": 3.793300342596117e-06, + "loss": 1.5045, + "step": 942 + }, + { + "epoch": 0.9686697483307652, + "grad_norm": 2.4424829483032227, + "learning_rate": 3.7913970308336505e-06, + "loss": 1.4936, + "step": 943 + }, + { + "epoch": 0.9696969696969697, + "grad_norm": 2.2769429683685303, + "learning_rate": 3.789493719071184e-06, + "loss": 1.4453, + "step": 944 + }, + { + "epoch": 0.9707241910631741, + "grad_norm": 2.183701276779175, + "learning_rate": 3.7875904073087177e-06, + "loss": 1.4005, + "step": 945 + }, + { + "epoch": 0.9717514124293786, + "grad_norm": 3.4538655281066895, + "learning_rate": 3.785687095546251e-06, + "loss": 1.4521, + "step": 946 + }, + { + "epoch": 0.9727786337955829, + "grad_norm": 2.3735523223876953, + "learning_rate": 3.7837837837837844e-06, + "loss": 1.4328, + "step": 947 + }, + { + "epoch": 0.9738058551617874, + "grad_norm": 2.6214165687561035, + "learning_rate": 3.7818804720213178e-06, + "loss": 1.4004, + "step": 948 + }, + { + "epoch": 0.9748330765279918, + "grad_norm": 2.365091562271118, + "learning_rate": 3.7799771602588507e-06, + "loss": 1.4006, + "step": 949 + }, + { + "epoch": 0.9758602978941961, + "grad_norm": 2.1926634311676025, + "learning_rate": 3.778073848496384e-06, + "loss": 1.4587, + "step": 950 + }, + { + "epoch": 0.9758602978941961, + "eval_loss": 1.767502784729004, + "eval_runtime": 37.9287, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 950 + }, + { + "epoch": 0.9768875192604006, + "grad_norm": 2.3047192096710205, + "learning_rate": 3.7761705367339174e-06, + "loss": 1.4826, + "step": 951 + }, + { + "epoch": 0.977914740626605, + "grad_norm": 2.3306028842926025, + "learning_rate": 3.7742672249714508e-06, + "loss": 1.4914, + "step": 952 + }, + { + "epoch": 0.9789419619928095, + "grad_norm": 2.3409693241119385, + "learning_rate": 3.772363913208984e-06, + "loss": 1.4556, + "step": 953 + }, + { + "epoch": 0.9799691833590138, + "grad_norm": 2.3337934017181396, + "learning_rate": 3.770460601446517e-06, + "loss": 1.4751, + "step": 954 + }, + { + "epoch": 0.9809964047252183, + "grad_norm": 2.3354830741882324, + "learning_rate": 3.7685572896840504e-06, + "loss": 1.4176, + "step": 955 + }, + { + "epoch": 0.9820236260914227, + "grad_norm": 2.4049782752990723, + "learning_rate": 3.766653977921584e-06, + "loss": 1.3715, + "step": 956 + }, + { + "epoch": 0.9830508474576272, + "grad_norm": 2.5127065181732178, + "learning_rate": 3.764750666159117e-06, + "loss": 1.468, + "step": 957 + }, + { + "epoch": 0.9840780688238315, + "grad_norm": 2.5576820373535156, + "learning_rate": 3.7628473543966505e-06, + "loss": 1.4237, + "step": 958 + }, + { + "epoch": 0.9851052901900359, + "grad_norm": 2.394098997116089, + "learning_rate": 3.7609440426341835e-06, + "loss": 1.5126, + "step": 959 + }, + { + "epoch": 0.9861325115562404, + "grad_norm": 2.3358702659606934, + "learning_rate": 3.759040730871717e-06, + "loss": 1.4109, + "step": 960 + }, + { + "epoch": 0.9861325115562404, + "eval_loss": 1.7665778398513794, + "eval_runtime": 37.9028, + "eval_samples_per_second": 26.383, + "eval_steps_per_second": 1.108, + "step": 960 + }, + { + "epoch": 0.9871597329224447, + "grad_norm": 2.4213271141052246, + "learning_rate": 3.75713741910925e-06, + "loss": 1.3995, + "step": 961 + }, + { + "epoch": 0.9881869542886492, + "grad_norm": 2.367835760116577, + "learning_rate": 3.7552341073467836e-06, + "loss": 1.4491, + "step": 962 + }, + { + "epoch": 0.9892141756548536, + "grad_norm": 2.433840751647949, + "learning_rate": 3.753330795584317e-06, + "loss": 1.5078, + "step": 963 + }, + { + "epoch": 0.9902413970210581, + "grad_norm": 2.3976447582244873, + "learning_rate": 3.75142748382185e-06, + "loss": 1.4107, + "step": 964 + }, + { + "epoch": 0.9912686183872624, + "grad_norm": 2.4369053840637207, + "learning_rate": 3.749524172059384e-06, + "loss": 1.4569, + "step": 965 + }, + { + "epoch": 0.9922958397534669, + "grad_norm": 2.3203182220458984, + "learning_rate": 3.747620860296917e-06, + "loss": 1.4976, + "step": 966 + }, + { + "epoch": 0.9933230611196713, + "grad_norm": 2.431885242462158, + "learning_rate": 3.7457175485344504e-06, + "loss": 1.4945, + "step": 967 + }, + { + "epoch": 0.9943502824858758, + "grad_norm": 2.2997021675109863, + "learning_rate": 3.7438142367719837e-06, + "loss": 1.48, + "step": 968 + }, + { + "epoch": 0.9953775038520801, + "grad_norm": 2.2502849102020264, + "learning_rate": 3.741910925009517e-06, + "loss": 1.4322, + "step": 969 + }, + { + "epoch": 0.9964047252182845, + "grad_norm": 2.311049222946167, + "learning_rate": 3.7400076132470505e-06, + "loss": 1.4498, + "step": 970 + }, + { + "epoch": 0.9964047252182845, + "eval_loss": 1.771975040435791, + "eval_runtime": 37.852, + "eval_samples_per_second": 26.419, + "eval_steps_per_second": 1.11, + "step": 970 + }, + { + "epoch": 0.997431946584489, + "grad_norm": 2.431821584701538, + "learning_rate": 3.7381043014845834e-06, + "loss": 1.4796, + "step": 971 + }, + { + "epoch": 0.9984591679506933, + "grad_norm": 2.319145441055298, + "learning_rate": 3.7362009897221168e-06, + "loss": 1.459, + "step": 972 + }, + { + "epoch": 0.9994863893168978, + "grad_norm": 2.4445269107818604, + "learning_rate": 3.73429767795965e-06, + "loss": 1.4764, + "step": 973 + }, + { + "epoch": 1.0005136106831023, + "grad_norm": 2.272233009338379, + "learning_rate": 3.7323943661971835e-06, + "loss": 1.3799, + "step": 974 + }, + { + "epoch": 1.0015408320493067, + "grad_norm": 2.5421202182769775, + "learning_rate": 3.730491054434717e-06, + "loss": 1.3257, + "step": 975 + }, + { + "epoch": 1.002568053415511, + "grad_norm": 2.3988611698150635, + "learning_rate": 3.72858774267225e-06, + "loss": 1.4244, + "step": 976 + }, + { + "epoch": 1.0035952747817154, + "grad_norm": 2.5080084800720215, + "learning_rate": 3.726684430909783e-06, + "loss": 1.4461, + "step": 977 + }, + { + "epoch": 1.00462249614792, + "grad_norm": 2.411081314086914, + "learning_rate": 3.7247811191473165e-06, + "loss": 1.45, + "step": 978 + }, + { + "epoch": 1.0056497175141244, + "grad_norm": 2.729546070098877, + "learning_rate": 3.72287780738485e-06, + "loss": 1.4296, + "step": 979 + }, + { + "epoch": 1.0066769388803287, + "grad_norm": 2.5932579040527344, + "learning_rate": 3.7209744956223832e-06, + "loss": 1.4312, + "step": 980 + }, + { + "epoch": 1.0066769388803287, + "eval_loss": 1.770633339881897, + "eval_runtime": 37.8977, + "eval_samples_per_second": 26.387, + "eval_steps_per_second": 1.108, + "step": 980 + }, + { + "epoch": 1.007704160246533, + "grad_norm": 2.4512336254119873, + "learning_rate": 3.719071183859916e-06, + "loss": 1.3637, + "step": 981 + }, + { + "epoch": 1.0087313816127375, + "grad_norm": 2.5416886806488037, + "learning_rate": 3.7171678720974495e-06, + "loss": 1.4591, + "step": 982 + }, + { + "epoch": 1.009758602978942, + "grad_norm": 2.464679002761841, + "learning_rate": 3.715264560334983e-06, + "loss": 1.4306, + "step": 983 + }, + { + "epoch": 1.0107858243451464, + "grad_norm": 2.3745839595794678, + "learning_rate": 3.7133612485725163e-06, + "loss": 1.4426, + "step": 984 + }, + { + "epoch": 1.0118130457113508, + "grad_norm": 2.3606340885162354, + "learning_rate": 3.71145793681005e-06, + "loss": 1.3618, + "step": 985 + }, + { + "epoch": 1.0128402670775551, + "grad_norm": 2.5836591720581055, + "learning_rate": 3.7095546250475834e-06, + "loss": 1.4776, + "step": 986 + }, + { + "epoch": 1.0138674884437597, + "grad_norm": 2.4108684062957764, + "learning_rate": 3.7076513132851168e-06, + "loss": 1.4047, + "step": 987 + }, + { + "epoch": 1.014894709809964, + "grad_norm": 2.4316983222961426, + "learning_rate": 3.7057480015226497e-06, + "loss": 1.3772, + "step": 988 + }, + { + "epoch": 1.0159219311761685, + "grad_norm": 2.5510315895080566, + "learning_rate": 3.703844689760183e-06, + "loss": 1.4536, + "step": 989 + }, + { + "epoch": 1.0169491525423728, + "grad_norm": 2.4947116374969482, + "learning_rate": 3.7019413779977164e-06, + "loss": 1.434, + "step": 990 + }, + { + "epoch": 1.0169491525423728, + "eval_loss": 1.7928223609924316, + "eval_runtime": 37.8838, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 990 + }, + { + "epoch": 1.0179763739085772, + "grad_norm": 2.5938122272491455, + "learning_rate": 3.70003806623525e-06, + "loss": 1.4083, + "step": 991 + }, + { + "epoch": 1.0190035952747818, + "grad_norm": 2.472865581512451, + "learning_rate": 3.698134754472783e-06, + "loss": 1.4696, + "step": 992 + }, + { + "epoch": 1.0200308166409862, + "grad_norm": 2.374344825744629, + "learning_rate": 3.696231442710316e-06, + "loss": 1.403, + "step": 993 + }, + { + "epoch": 1.0210580380071905, + "grad_norm": 2.6928529739379883, + "learning_rate": 3.6943281309478495e-06, + "loss": 1.4216, + "step": 994 + }, + { + "epoch": 1.022085259373395, + "grad_norm": 2.640890598297119, + "learning_rate": 3.692424819185383e-06, + "loss": 1.4362, + "step": 995 + }, + { + "epoch": 1.0231124807395995, + "grad_norm": 2.45357084274292, + "learning_rate": 3.690521507422916e-06, + "loss": 1.382, + "step": 996 + }, + { + "epoch": 1.0241397021058039, + "grad_norm": 2.712517023086548, + "learning_rate": 3.6886181956604496e-06, + "loss": 1.395, + "step": 997 + }, + { + "epoch": 1.0251669234720082, + "grad_norm": 2.7127771377563477, + "learning_rate": 3.686714883897983e-06, + "loss": 1.4589, + "step": 998 + }, + { + "epoch": 1.0261941448382126, + "grad_norm": 2.5379409790039062, + "learning_rate": 3.684811572135516e-06, + "loss": 1.4551, + "step": 999 + }, + { + "epoch": 1.027221366204417, + "grad_norm": 2.5015134811401367, + "learning_rate": 3.6829082603730492e-06, + "loss": 1.4112, + "step": 1000 + }, + { + "epoch": 1.027221366204417, + "eval_loss": 1.7806837558746338, + "eval_runtime": 37.9091, + "eval_samples_per_second": 26.379, + "eval_steps_per_second": 1.108, + "step": 1000 + }, + { + "epoch": 1.0282485875706215, + "grad_norm": 2.481306314468384, + "learning_rate": 3.6810049486105826e-06, + "loss": 1.4469, + "step": 1001 + }, + { + "epoch": 1.029275808936826, + "grad_norm": 2.481520891189575, + "learning_rate": 3.679101636848116e-06, + "loss": 1.4284, + "step": 1002 + }, + { + "epoch": 1.0303030303030303, + "grad_norm": 2.4291019439697266, + "learning_rate": 3.6771983250856493e-06, + "loss": 1.4067, + "step": 1003 + }, + { + "epoch": 1.0313302516692346, + "grad_norm": 2.5361173152923584, + "learning_rate": 3.6752950133231822e-06, + "loss": 1.4347, + "step": 1004 + }, + { + "epoch": 1.0323574730354392, + "grad_norm": 2.5588085651397705, + "learning_rate": 3.6733917015607156e-06, + "loss": 1.4455, + "step": 1005 + }, + { + "epoch": 1.0333846944016436, + "grad_norm": 2.555091619491577, + "learning_rate": 3.6714883897982494e-06, + "loss": 1.399, + "step": 1006 + }, + { + "epoch": 1.034411915767848, + "grad_norm": 2.5489652156829834, + "learning_rate": 3.6695850780357828e-06, + "loss": 1.3894, + "step": 1007 + }, + { + "epoch": 1.0354391371340523, + "grad_norm": 2.5258426666259766, + "learning_rate": 3.667681766273316e-06, + "loss": 1.3962, + "step": 1008 + }, + { + "epoch": 1.0364663585002567, + "grad_norm": 2.455916166305542, + "learning_rate": 3.6657784545108495e-06, + "loss": 1.4773, + "step": 1009 + }, + { + "epoch": 1.0374935798664613, + "grad_norm": 2.5466208457946777, + "learning_rate": 3.6638751427483824e-06, + "loss": 1.4337, + "step": 1010 + }, + { + "epoch": 1.0374935798664613, + "eval_loss": 1.7840261459350586, + "eval_runtime": 37.6685, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 1010 + }, + { + "epoch": 1.0385208012326657, + "grad_norm": 2.5446081161499023, + "learning_rate": 3.6619718309859158e-06, + "loss": 1.4777, + "step": 1011 + }, + { + "epoch": 1.03954802259887, + "grad_norm": 2.509493350982666, + "learning_rate": 3.660068519223449e-06, + "loss": 1.4295, + "step": 1012 + }, + { + "epoch": 1.0405752439650744, + "grad_norm": 2.417632818222046, + "learning_rate": 3.6581652074609825e-06, + "loss": 1.4044, + "step": 1013 + }, + { + "epoch": 1.041602465331279, + "grad_norm": 2.457902193069458, + "learning_rate": 3.656261895698516e-06, + "loss": 1.3948, + "step": 1014 + }, + { + "epoch": 1.0426296866974833, + "grad_norm": 2.4869253635406494, + "learning_rate": 3.6543585839360492e-06, + "loss": 1.397, + "step": 1015 + }, + { + "epoch": 1.0436569080636877, + "grad_norm": 2.495419979095459, + "learning_rate": 3.652455272173582e-06, + "loss": 1.4037, + "step": 1016 + }, + { + "epoch": 1.044684129429892, + "grad_norm": 2.5231642723083496, + "learning_rate": 3.6505519604111155e-06, + "loss": 1.4156, + "step": 1017 + }, + { + "epoch": 1.0457113507960965, + "grad_norm": 2.8302087783813477, + "learning_rate": 3.648648648648649e-06, + "loss": 1.4529, + "step": 1018 + }, + { + "epoch": 1.046738572162301, + "grad_norm": 2.467430353164673, + "learning_rate": 3.6467453368861823e-06, + "loss": 1.4078, + "step": 1019 + }, + { + "epoch": 1.0477657935285054, + "grad_norm": 2.469128131866455, + "learning_rate": 3.6448420251237156e-06, + "loss": 1.4344, + "step": 1020 + }, + { + "epoch": 1.0477657935285054, + "eval_loss": 1.7800015211105347, + "eval_runtime": 37.854, + "eval_samples_per_second": 26.417, + "eval_steps_per_second": 1.11, + "step": 1020 + }, + { + "epoch": 1.0487930148947098, + "grad_norm": 2.59389328956604, + "learning_rate": 3.6429387133612486e-06, + "loss": 1.451, + "step": 1021 + }, + { + "epoch": 1.0498202362609141, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.641035401598782e-06, + "loss": 1.4026, + "step": 1022 + }, + { + "epoch": 1.0508474576271187, + "grad_norm": 2.601926565170288, + "learning_rate": 3.6391320898363153e-06, + "loss": 1.4382, + "step": 1023 + }, + { + "epoch": 1.051874678993323, + "grad_norm": 2.5336716175079346, + "learning_rate": 3.6372287780738487e-06, + "loss": 1.4184, + "step": 1024 + }, + { + "epoch": 1.0529019003595275, + "grad_norm": 2.503507614135742, + "learning_rate": 3.635325466311382e-06, + "loss": 1.4426, + "step": 1025 + }, + { + "epoch": 1.0539291217257318, + "grad_norm": 2.483013153076172, + "learning_rate": 3.633422154548916e-06, + "loss": 1.4563, + "step": 1026 + }, + { + "epoch": 1.0549563430919364, + "grad_norm": 2.4803354740142822, + "learning_rate": 3.6315188427864487e-06, + "loss": 1.4284, + "step": 1027 + }, + { + "epoch": 1.0559835644581408, + "grad_norm": 2.3735713958740234, + "learning_rate": 3.629615531023982e-06, + "loss": 1.3772, + "step": 1028 + }, + { + "epoch": 1.0570107858243452, + "grad_norm": 2.471226215362549, + "learning_rate": 3.6277122192615155e-06, + "loss": 1.4399, + "step": 1029 + }, + { + "epoch": 1.0580380071905495, + "grad_norm": 2.3827438354492188, + "learning_rate": 3.625808907499049e-06, + "loss": 1.3951, + "step": 1030 + }, + { + "epoch": 1.0580380071905495, + "eval_loss": 1.7787058353424072, + "eval_runtime": 37.9484, + "eval_samples_per_second": 26.352, + "eval_steps_per_second": 1.107, + "step": 1030 + }, + { + "epoch": 1.059065228556754, + "grad_norm": 2.403416633605957, + "learning_rate": 3.623905595736582e-06, + "loss": 1.3946, + "step": 1031 + }, + { + "epoch": 1.0600924499229585, + "grad_norm": 2.455740213394165, + "learning_rate": 3.6220022839741156e-06, + "loss": 1.3859, + "step": 1032 + }, + { + "epoch": 1.0611196712891628, + "grad_norm": 2.6349902153015137, + "learning_rate": 3.6200989722116485e-06, + "loss": 1.4092, + "step": 1033 + }, + { + "epoch": 1.0621468926553672, + "grad_norm": 2.6292974948883057, + "learning_rate": 3.618195660449182e-06, + "loss": 1.419, + "step": 1034 + }, + { + "epoch": 1.0631741140215716, + "grad_norm": 2.4338958263397217, + "learning_rate": 3.6162923486867152e-06, + "loss": 1.3546, + "step": 1035 + }, + { + "epoch": 1.064201335387776, + "grad_norm": 2.376420259475708, + "learning_rate": 3.6143890369242486e-06, + "loss": 1.3899, + "step": 1036 + }, + { + "epoch": 1.0652285567539805, + "grad_norm": 2.428738832473755, + "learning_rate": 3.612485725161782e-06, + "loss": 1.3955, + "step": 1037 + }, + { + "epoch": 1.066255778120185, + "grad_norm": 2.4316651821136475, + "learning_rate": 3.610582413399315e-06, + "loss": 1.3875, + "step": 1038 + }, + { + "epoch": 1.0672829994863893, + "grad_norm": 2.3969266414642334, + "learning_rate": 3.6086791016368482e-06, + "loss": 1.404, + "step": 1039 + }, + { + "epoch": 1.0683102208525936, + "grad_norm": 2.662283420562744, + "learning_rate": 3.6067757898743816e-06, + "loss": 1.4639, + "step": 1040 + }, + { + "epoch": 1.0683102208525936, + "eval_loss": 1.7603929042816162, + "eval_runtime": 38.0098, + "eval_samples_per_second": 26.309, + "eval_steps_per_second": 1.105, + "step": 1040 + }, + { + "epoch": 1.0693374422187982, + "grad_norm": 2.601452112197876, + "learning_rate": 3.604872478111915e-06, + "loss": 1.47, + "step": 1041 + }, + { + "epoch": 1.0703646635850026, + "grad_norm": 2.4221863746643066, + "learning_rate": 3.6029691663494483e-06, + "loss": 1.4162, + "step": 1042 + }, + { + "epoch": 1.071391884951207, + "grad_norm": 2.4996650218963623, + "learning_rate": 3.6010658545869813e-06, + "loss": 1.443, + "step": 1043 + }, + { + "epoch": 1.0724191063174113, + "grad_norm": 2.547239303588867, + "learning_rate": 3.5991625428245146e-06, + "loss": 1.4464, + "step": 1044 + }, + { + "epoch": 1.073446327683616, + "grad_norm": 2.481187343597412, + "learning_rate": 3.597259231062048e-06, + "loss": 1.3749, + "step": 1045 + }, + { + "epoch": 1.0744735490498203, + "grad_norm": 2.5504934787750244, + "learning_rate": 3.5953559192995818e-06, + "loss": 1.4231, + "step": 1046 + }, + { + "epoch": 1.0755007704160247, + "grad_norm": 2.465390205383301, + "learning_rate": 3.593452607537115e-06, + "loss": 1.4805, + "step": 1047 + }, + { + "epoch": 1.076527991782229, + "grad_norm": 2.4399969577789307, + "learning_rate": 3.5915492957746485e-06, + "loss": 1.4091, + "step": 1048 + }, + { + "epoch": 1.0775552131484334, + "grad_norm": 2.4819061756134033, + "learning_rate": 3.589645984012182e-06, + "loss": 1.4367, + "step": 1049 + }, + { + "epoch": 1.078582434514638, + "grad_norm": 2.459517002105713, + "learning_rate": 3.587742672249715e-06, + "loss": 1.3882, + "step": 1050 + }, + { + "epoch": 1.078582434514638, + "eval_loss": 1.7557458877563477, + "eval_runtime": 38.0271, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1050 + }, + { + "epoch": 1.0796096558808423, + "grad_norm": 2.5434813499450684, + "learning_rate": 3.585839360487248e-06, + "loss": 1.3926, + "step": 1051 + }, + { + "epoch": 1.0806368772470467, + "grad_norm": 2.6173205375671387, + "learning_rate": 3.5839360487247815e-06, + "loss": 1.4044, + "step": 1052 + }, + { + "epoch": 1.081664098613251, + "grad_norm": 2.677584171295166, + "learning_rate": 3.582032736962315e-06, + "loss": 1.4695, + "step": 1053 + }, + { + "epoch": 1.0826913199794557, + "grad_norm": 2.4151623249053955, + "learning_rate": 3.5801294251998483e-06, + "loss": 1.3699, + "step": 1054 + }, + { + "epoch": 1.08371854134566, + "grad_norm": 2.4683942794799805, + "learning_rate": 3.578226113437381e-06, + "loss": 1.4132, + "step": 1055 + }, + { + "epoch": 1.0847457627118644, + "grad_norm": 2.6076581478118896, + "learning_rate": 3.5763228016749146e-06, + "loss": 1.4729, + "step": 1056 + }, + { + "epoch": 1.0857729840780688, + "grad_norm": 2.4766876697540283, + "learning_rate": 3.574419489912448e-06, + "loss": 1.4196, + "step": 1057 + }, + { + "epoch": 1.0868002054442731, + "grad_norm": 2.466458559036255, + "learning_rate": 3.5725161781499813e-06, + "loss": 1.4049, + "step": 1058 + }, + { + "epoch": 1.0878274268104777, + "grad_norm": 2.4220409393310547, + "learning_rate": 3.5706128663875147e-06, + "loss": 1.4246, + "step": 1059 + }, + { + "epoch": 1.088854648176682, + "grad_norm": 2.5571770668029785, + "learning_rate": 3.5687095546250476e-06, + "loss": 1.4397, + "step": 1060 + }, + { + "epoch": 1.088854648176682, + "eval_loss": 1.7701127529144287, + "eval_runtime": 38.0754, + "eval_samples_per_second": 26.264, + "eval_steps_per_second": 1.103, + "step": 1060 + }, + { + "epoch": 1.0898818695428865, + "grad_norm": 2.648768186569214, + "learning_rate": 3.566806242862581e-06, + "loss": 1.423, + "step": 1061 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 2.403651714324951, + "learning_rate": 3.5649029311001143e-06, + "loss": 1.3995, + "step": 1062 + }, + { + "epoch": 1.0919363122752954, + "grad_norm": 2.465538263320923, + "learning_rate": 3.5629996193376477e-06, + "loss": 1.4139, + "step": 1063 + }, + { + "epoch": 1.0929635336414998, + "grad_norm": 2.464662551879883, + "learning_rate": 3.561096307575181e-06, + "loss": 1.4499, + "step": 1064 + }, + { + "epoch": 1.0939907550077042, + "grad_norm": 2.4943597316741943, + "learning_rate": 3.559192995812714e-06, + "loss": 1.4555, + "step": 1065 + }, + { + "epoch": 1.0950179763739085, + "grad_norm": 2.6332199573516846, + "learning_rate": 3.557289684050248e-06, + "loss": 1.4484, + "step": 1066 + }, + { + "epoch": 1.0960451977401129, + "grad_norm": 2.4370667934417725, + "learning_rate": 3.555386372287781e-06, + "loss": 1.4588, + "step": 1067 + }, + { + "epoch": 1.0970724191063175, + "grad_norm": 2.568368673324585, + "learning_rate": 3.5534830605253145e-06, + "loss": 1.4269, + "step": 1068 + }, + { + "epoch": 1.0980996404725218, + "grad_norm": 2.4618077278137207, + "learning_rate": 3.551579748762848e-06, + "loss": 1.4165, + "step": 1069 + }, + { + "epoch": 1.0991268618387262, + "grad_norm": 2.551320791244507, + "learning_rate": 3.5496764370003812e-06, + "loss": 1.3991, + "step": 1070 + }, + { + "epoch": 1.0991268618387262, + "eval_loss": 1.7648736238479614, + "eval_runtime": 38.0273, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1070 + }, + { + "epoch": 1.1001540832049306, + "grad_norm": 2.350905418395996, + "learning_rate": 3.5477731252379146e-06, + "loss": 1.3556, + "step": 1071 + }, + { + "epoch": 1.1011813045711352, + "grad_norm": 2.4294588565826416, + "learning_rate": 3.5458698134754475e-06, + "loss": 1.3801, + "step": 1072 + }, + { + "epoch": 1.1022085259373395, + "grad_norm": 2.576774835586548, + "learning_rate": 3.543966501712981e-06, + "loss": 1.3876, + "step": 1073 + }, + { + "epoch": 1.103235747303544, + "grad_norm": 2.716935873031616, + "learning_rate": 3.5420631899505142e-06, + "loss": 1.4907, + "step": 1074 + }, + { + "epoch": 1.1042629686697483, + "grad_norm": 2.425305128097534, + "learning_rate": 3.5401598781880476e-06, + "loss": 1.3161, + "step": 1075 + }, + { + "epoch": 1.1052901900359529, + "grad_norm": 2.54738712310791, + "learning_rate": 3.538256566425581e-06, + "loss": 1.4462, + "step": 1076 + }, + { + "epoch": 1.1063174114021572, + "grad_norm": 2.5109341144561768, + "learning_rate": 3.536353254663114e-06, + "loss": 1.4022, + "step": 1077 + }, + { + "epoch": 1.1073446327683616, + "grad_norm": 2.468749523162842, + "learning_rate": 3.5344499429006473e-06, + "loss": 1.3857, + "step": 1078 + }, + { + "epoch": 1.108371854134566, + "grad_norm": 2.5285134315490723, + "learning_rate": 3.5325466311381806e-06, + "loss": 1.4309, + "step": 1079 + }, + { + "epoch": 1.1093990755007703, + "grad_norm": 2.6391055583953857, + "learning_rate": 3.530643319375714e-06, + "loss": 1.3644, + "step": 1080 + }, + { + "epoch": 1.1093990755007703, + "eval_loss": 1.7846598625183105, + "eval_runtime": 38.0604, + "eval_samples_per_second": 26.274, + "eval_steps_per_second": 1.104, + "step": 1080 + }, + { + "epoch": 1.110426296866975, + "grad_norm": 2.660356283187866, + "learning_rate": 3.5287400076132474e-06, + "loss": 1.3922, + "step": 1081 + }, + { + "epoch": 1.1114535182331793, + "grad_norm": 2.5648741722106934, + "learning_rate": 3.5268366958507803e-06, + "loss": 1.4282, + "step": 1082 + }, + { + "epoch": 1.1124807395993837, + "grad_norm": 2.6862876415252686, + "learning_rate": 3.5249333840883137e-06, + "loss": 1.4208, + "step": 1083 + }, + { + "epoch": 1.113507960965588, + "grad_norm": 2.6200191974639893, + "learning_rate": 3.523030072325847e-06, + "loss": 1.3883, + "step": 1084 + }, + { + "epoch": 1.1145351823317924, + "grad_norm": 2.5802111625671387, + "learning_rate": 3.5211267605633804e-06, + "loss": 1.4033, + "step": 1085 + }, + { + "epoch": 1.115562403697997, + "grad_norm": 2.4148080348968506, + "learning_rate": 3.519223448800914e-06, + "loss": 1.4085, + "step": 1086 + }, + { + "epoch": 1.1165896250642013, + "grad_norm": 2.7452127933502197, + "learning_rate": 3.5173201370384475e-06, + "loss": 1.4021, + "step": 1087 + }, + { + "epoch": 1.1176168464304057, + "grad_norm": 2.6437106132507324, + "learning_rate": 3.515416825275981e-06, + "loss": 1.4161, + "step": 1088 + }, + { + "epoch": 1.11864406779661, + "grad_norm": 2.4709842205047607, + "learning_rate": 3.513513513513514e-06, + "loss": 1.4322, + "step": 1089 + }, + { + "epoch": 1.1196712891628147, + "grad_norm": 2.5138099193573, + "learning_rate": 3.511610201751047e-06, + "loss": 1.3435, + "step": 1090 + }, + { + "epoch": 1.1196712891628147, + "eval_loss": 1.7687023878097534, + "eval_runtime": 38.0137, + "eval_samples_per_second": 26.306, + "eval_steps_per_second": 1.105, + "step": 1090 + }, + { + "epoch": 1.120698510529019, + "grad_norm": 2.4194881916046143, + "learning_rate": 3.5097068899885806e-06, + "loss": 1.4133, + "step": 1091 + }, + { + "epoch": 1.1217257318952234, + "grad_norm": 2.6564371585845947, + "learning_rate": 3.507803578226114e-06, + "loss": 1.4759, + "step": 1092 + }, + { + "epoch": 1.1227529532614278, + "grad_norm": 2.4480180740356445, + "learning_rate": 3.5059002664636473e-06, + "loss": 1.3577, + "step": 1093 + }, + { + "epoch": 1.1237801746276324, + "grad_norm": 2.5538978576660156, + "learning_rate": 3.5039969547011802e-06, + "loss": 1.3984, + "step": 1094 + }, + { + "epoch": 1.1248073959938367, + "grad_norm": 2.6467089653015137, + "learning_rate": 3.5020936429387136e-06, + "loss": 1.4049, + "step": 1095 + }, + { + "epoch": 1.125834617360041, + "grad_norm": 2.6502673625946045, + "learning_rate": 3.500190331176247e-06, + "loss": 1.462, + "step": 1096 + }, + { + "epoch": 1.1268618387262455, + "grad_norm": 2.5696403980255127, + "learning_rate": 3.4982870194137803e-06, + "loss": 1.3926, + "step": 1097 + }, + { + "epoch": 1.1278890600924498, + "grad_norm": 2.5951733589172363, + "learning_rate": 3.4963837076513137e-06, + "loss": 1.4031, + "step": 1098 + }, + { + "epoch": 1.1289162814586544, + "grad_norm": 2.4910218715667725, + "learning_rate": 3.4944803958888466e-06, + "loss": 1.3348, + "step": 1099 + }, + { + "epoch": 1.1299435028248588, + "grad_norm": 2.6325888633728027, + "learning_rate": 3.49257708412638e-06, + "loss": 1.4641, + "step": 1100 + }, + { + "epoch": 1.1299435028248588, + "eval_loss": 1.7720683813095093, + "eval_runtime": 37.9585, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 1100 + }, + { + "epoch": 1.1309707241910631, + "grad_norm": 2.6712775230407715, + "learning_rate": 3.4906737723639133e-06, + "loss": 1.4726, + "step": 1101 + }, + { + "epoch": 1.1319979455572675, + "grad_norm": 2.580404043197632, + "learning_rate": 3.4887704606014467e-06, + "loss": 1.4158, + "step": 1102 + }, + { + "epoch": 1.133025166923472, + "grad_norm": 2.467453718185425, + "learning_rate": 3.48686714883898e-06, + "loss": 1.4636, + "step": 1103 + }, + { + "epoch": 1.1340523882896765, + "grad_norm": 2.627136468887329, + "learning_rate": 3.484963837076513e-06, + "loss": 1.4082, + "step": 1104 + }, + { + "epoch": 1.1350796096558808, + "grad_norm": 2.6198604106903076, + "learning_rate": 3.4830605253140464e-06, + "loss": 1.3628, + "step": 1105 + }, + { + "epoch": 1.1361068310220852, + "grad_norm": 2.439512014389038, + "learning_rate": 3.4811572135515797e-06, + "loss": 1.3885, + "step": 1106 + }, + { + "epoch": 1.1371340523882898, + "grad_norm": 2.5543010234832764, + "learning_rate": 3.4792539017891135e-06, + "loss": 1.3674, + "step": 1107 + }, + { + "epoch": 1.1381612737544942, + "grad_norm": 2.7263364791870117, + "learning_rate": 3.477350590026647e-06, + "loss": 1.4387, + "step": 1108 + }, + { + "epoch": 1.1391884951206985, + "grad_norm": 2.5030503273010254, + "learning_rate": 3.4754472782641802e-06, + "loss": 1.4222, + "step": 1109 + }, + { + "epoch": 1.140215716486903, + "grad_norm": 2.5090320110321045, + "learning_rate": 3.4735439665017136e-06, + "loss": 1.3981, + "step": 1110 + }, + { + "epoch": 1.140215716486903, + "eval_loss": 1.769408941268921, + "eval_runtime": 37.608, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 1110 + }, + { + "epoch": 1.1412429378531073, + "grad_norm": 2.575934648513794, + "learning_rate": 3.4716406547392465e-06, + "loss": 1.4074, + "step": 1111 + }, + { + "epoch": 1.1422701592193119, + "grad_norm": 2.6381359100341797, + "learning_rate": 3.46973734297678e-06, + "loss": 1.4162, + "step": 1112 + }, + { + "epoch": 1.1432973805855162, + "grad_norm": 2.7605674266815186, + "learning_rate": 3.4678340312143133e-06, + "loss": 1.4616, + "step": 1113 + }, + { + "epoch": 1.1443246019517206, + "grad_norm": 2.5548479557037354, + "learning_rate": 3.4659307194518466e-06, + "loss": 1.4511, + "step": 1114 + }, + { + "epoch": 1.145351823317925, + "grad_norm": 2.471168279647827, + "learning_rate": 3.46402740768938e-06, + "loss": 1.3777, + "step": 1115 + }, + { + "epoch": 1.1463790446841293, + "grad_norm": 2.4416346549987793, + "learning_rate": 3.462124095926913e-06, + "loss": 1.3805, + "step": 1116 + }, + { + "epoch": 1.147406266050334, + "grad_norm": 2.4972171783447266, + "learning_rate": 3.4602207841644463e-06, + "loss": 1.4282, + "step": 1117 + }, + { + "epoch": 1.1484334874165383, + "grad_norm": 2.4379162788391113, + "learning_rate": 3.4583174724019797e-06, + "loss": 1.4744, + "step": 1118 + }, + { + "epoch": 1.1494607087827426, + "grad_norm": 2.508047580718994, + "learning_rate": 3.456414160639513e-06, + "loss": 1.4303, + "step": 1119 + }, + { + "epoch": 1.150487930148947, + "grad_norm": 2.4299631118774414, + "learning_rate": 3.4545108488770464e-06, + "loss": 1.4179, + "step": 1120 + }, + { + "epoch": 1.150487930148947, + "eval_loss": 1.7768158912658691, + "eval_runtime": 37.736, + "eval_samples_per_second": 26.5, + "eval_steps_per_second": 1.113, + "step": 1120 + }, + { + "epoch": 1.1515151515151516, + "grad_norm": 2.41538143157959, + "learning_rate": 3.4526075371145793e-06, + "loss": 1.4235, + "step": 1121 + }, + { + "epoch": 1.152542372881356, + "grad_norm": 2.5614655017852783, + "learning_rate": 3.4507042253521127e-06, + "loss": 1.4879, + "step": 1122 + }, + { + "epoch": 1.1535695942475603, + "grad_norm": 2.4826605319976807, + "learning_rate": 3.448800913589646e-06, + "loss": 1.3968, + "step": 1123 + }, + { + "epoch": 1.1545968156137647, + "grad_norm": 2.5341391563415527, + "learning_rate": 3.4468976018271794e-06, + "loss": 1.4903, + "step": 1124 + }, + { + "epoch": 1.1556240369799693, + "grad_norm": 2.493818998336792, + "learning_rate": 3.4449942900647128e-06, + "loss": 1.4466, + "step": 1125 + }, + { + "epoch": 1.1566512583461737, + "grad_norm": 2.6057965755462646, + "learning_rate": 3.443090978302246e-06, + "loss": 1.4425, + "step": 1126 + }, + { + "epoch": 1.157678479712378, + "grad_norm": 2.5481390953063965, + "learning_rate": 3.44118766653978e-06, + "loss": 1.4296, + "step": 1127 + }, + { + "epoch": 1.1587057010785824, + "grad_norm": 2.5753650665283203, + "learning_rate": 3.439284354777313e-06, + "loss": 1.4408, + "step": 1128 + }, + { + "epoch": 1.1597329224447868, + "grad_norm": 2.5423803329467773, + "learning_rate": 3.4373810430148462e-06, + "loss": 1.373, + "step": 1129 + }, + { + "epoch": 1.1607601438109914, + "grad_norm": 2.593487024307251, + "learning_rate": 3.4354777312523796e-06, + "loss": 1.3992, + "step": 1130 + }, + { + "epoch": 1.1607601438109914, + "eval_loss": 1.7784411907196045, + "eval_runtime": 37.8687, + "eval_samples_per_second": 26.407, + "eval_steps_per_second": 1.109, + "step": 1130 + }, + { + "epoch": 1.1617873651771957, + "grad_norm": 2.4693503379821777, + "learning_rate": 3.433574419489913e-06, + "loss": 1.4339, + "step": 1131 + }, + { + "epoch": 1.1628145865434, + "grad_norm": 2.5477423667907715, + "learning_rate": 3.4316711077274463e-06, + "loss": 1.4035, + "step": 1132 + }, + { + "epoch": 1.1638418079096045, + "grad_norm": 2.5650672912597656, + "learning_rate": 3.4297677959649792e-06, + "loss": 1.4517, + "step": 1133 + }, + { + "epoch": 1.1648690292758088, + "grad_norm": 2.3217973709106445, + "learning_rate": 3.4278644842025126e-06, + "loss": 1.3983, + "step": 1134 + }, + { + "epoch": 1.1658962506420134, + "grad_norm": 2.674342155456543, + "learning_rate": 3.425961172440046e-06, + "loss": 1.4799, + "step": 1135 + }, + { + "epoch": 1.1669234720082178, + "grad_norm": 2.4861397743225098, + "learning_rate": 3.4240578606775793e-06, + "loss": 1.414, + "step": 1136 + }, + { + "epoch": 1.1679506933744221, + "grad_norm": 2.4306106567382812, + "learning_rate": 3.4221545489151127e-06, + "loss": 1.3918, + "step": 1137 + }, + { + "epoch": 1.1689779147406265, + "grad_norm": 2.6873955726623535, + "learning_rate": 3.4202512371526456e-06, + "loss": 1.4979, + "step": 1138 + }, + { + "epoch": 1.170005136106831, + "grad_norm": 2.559042453765869, + "learning_rate": 3.418347925390179e-06, + "loss": 1.4424, + "step": 1139 + }, + { + "epoch": 1.1710323574730355, + "grad_norm": 2.397489547729492, + "learning_rate": 3.4164446136277124e-06, + "loss": 1.418, + "step": 1140 + }, + { + "epoch": 1.1710323574730355, + "eval_loss": 1.7702736854553223, + "eval_runtime": 37.8963, + "eval_samples_per_second": 26.388, + "eval_steps_per_second": 1.108, + "step": 1140 + }, + { + "epoch": 1.1720595788392398, + "grad_norm": 2.4295787811279297, + "learning_rate": 3.4145413018652457e-06, + "loss": 1.4436, + "step": 1141 + }, + { + "epoch": 1.1730868002054442, + "grad_norm": 2.503202199935913, + "learning_rate": 3.412637990102779e-06, + "loss": 1.4598, + "step": 1142 + }, + { + "epoch": 1.1741140215716488, + "grad_norm": 2.540195941925049, + "learning_rate": 3.4107346783403125e-06, + "loss": 1.3704, + "step": 1143 + }, + { + "epoch": 1.1751412429378532, + "grad_norm": 2.4832522869110107, + "learning_rate": 3.4088313665778454e-06, + "loss": 1.4081, + "step": 1144 + }, + { + "epoch": 1.1761684643040575, + "grad_norm": 2.5639121532440186, + "learning_rate": 3.4069280548153788e-06, + "loss": 1.3866, + "step": 1145 + }, + { + "epoch": 1.177195685670262, + "grad_norm": 2.526610851287842, + "learning_rate": 3.405024743052912e-06, + "loss": 1.4181, + "step": 1146 + }, + { + "epoch": 1.1782229070364663, + "grad_norm": 2.6945254802703857, + "learning_rate": 3.403121431290446e-06, + "loss": 1.3871, + "step": 1147 + }, + { + "epoch": 1.1792501284026708, + "grad_norm": 2.6561484336853027, + "learning_rate": 3.4012181195279793e-06, + "loss": 1.3987, + "step": 1148 + }, + { + "epoch": 1.1802773497688752, + "grad_norm": 2.6104650497436523, + "learning_rate": 3.3993148077655126e-06, + "loss": 1.4057, + "step": 1149 + }, + { + "epoch": 1.1813045711350796, + "grad_norm": 2.571707248687744, + "learning_rate": 3.3974114960030456e-06, + "loss": 1.3606, + "step": 1150 + }, + { + "epoch": 1.1813045711350796, + "eval_loss": 1.7780840396881104, + "eval_runtime": 37.8616, + "eval_samples_per_second": 26.412, + "eval_steps_per_second": 1.109, + "step": 1150 + }, + { + "epoch": 1.182331792501284, + "grad_norm": 2.5804402828216553, + "learning_rate": 3.395508184240579e-06, + "loss": 1.3876, + "step": 1151 + }, + { + "epoch": 1.1833590138674885, + "grad_norm": 2.832446336746216, + "learning_rate": 3.3936048724781123e-06, + "loss": 1.4674, + "step": 1152 + }, + { + "epoch": 1.184386235233693, + "grad_norm": 2.6478688716888428, + "learning_rate": 3.3917015607156457e-06, + "loss": 1.3643, + "step": 1153 + }, + { + "epoch": 1.1854134565998973, + "grad_norm": 2.579416513442993, + "learning_rate": 3.389798248953179e-06, + "loss": 1.3874, + "step": 1154 + }, + { + "epoch": 1.1864406779661016, + "grad_norm": 2.487861156463623, + "learning_rate": 3.387894937190712e-06, + "loss": 1.4323, + "step": 1155 + }, + { + "epoch": 1.1874678993323062, + "grad_norm": 2.5020716190338135, + "learning_rate": 3.3859916254282453e-06, + "loss": 1.423, + "step": 1156 + }, + { + "epoch": 1.1884951206985106, + "grad_norm": 2.6588852405548096, + "learning_rate": 3.3840883136657787e-06, + "loss": 1.4808, + "step": 1157 + }, + { + "epoch": 1.189522342064715, + "grad_norm": 2.5702178478240967, + "learning_rate": 3.382185001903312e-06, + "loss": 1.4007, + "step": 1158 + }, + { + "epoch": 1.1905495634309193, + "grad_norm": 2.5082061290740967, + "learning_rate": 3.3802816901408454e-06, + "loss": 1.4278, + "step": 1159 + }, + { + "epoch": 1.1915767847971237, + "grad_norm": 2.4165053367614746, + "learning_rate": 3.3783783783783788e-06, + "loss": 1.4088, + "step": 1160 + }, + { + "epoch": 1.1915767847971237, + "eval_loss": 1.7722209692001343, + "eval_runtime": 37.8843, + "eval_samples_per_second": 26.396, + "eval_steps_per_second": 1.109, + "step": 1160 + }, + { + "epoch": 1.1926040061633283, + "grad_norm": 2.4510512351989746, + "learning_rate": 3.3764750666159117e-06, + "loss": 1.4089, + "step": 1161 + }, + { + "epoch": 1.1936312275295327, + "grad_norm": 2.5006556510925293, + "learning_rate": 3.374571754853445e-06, + "loss": 1.3988, + "step": 1162 + }, + { + "epoch": 1.194658448895737, + "grad_norm": 2.416264057159424, + "learning_rate": 3.3726684430909784e-06, + "loss": 1.3576, + "step": 1163 + }, + { + "epoch": 1.1956856702619414, + "grad_norm": 2.465613842010498, + "learning_rate": 3.370765131328512e-06, + "loss": 1.3796, + "step": 1164 + }, + { + "epoch": 1.1967128916281458, + "grad_norm": 2.510542154312134, + "learning_rate": 3.368861819566045e-06, + "loss": 1.391, + "step": 1165 + }, + { + "epoch": 1.1977401129943503, + "grad_norm": 2.729495048522949, + "learning_rate": 3.366958507803578e-06, + "loss": 1.431, + "step": 1166 + }, + { + "epoch": 1.1987673343605547, + "grad_norm": 2.4473495483398438, + "learning_rate": 3.365055196041112e-06, + "loss": 1.376, + "step": 1167 + }, + { + "epoch": 1.199794555726759, + "grad_norm": 2.3869211673736572, + "learning_rate": 3.3631518842786452e-06, + "loss": 1.4498, + "step": 1168 + }, + { + "epoch": 1.2008217770929634, + "grad_norm": 2.4202256202697754, + "learning_rate": 3.3612485725161786e-06, + "loss": 1.3754, + "step": 1169 + }, + { + "epoch": 1.201848998459168, + "grad_norm": 2.452707529067993, + "learning_rate": 3.359345260753712e-06, + "loss": 1.4647, + "step": 1170 + }, + { + "epoch": 1.201848998459168, + "eval_loss": 1.769252896308899, + "eval_runtime": 37.9385, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1170 + }, + { + "epoch": 1.2028762198253724, + "grad_norm": 2.56616473197937, + "learning_rate": 3.3574419489912453e-06, + "loss": 1.443, + "step": 1171 + }, + { + "epoch": 1.2039034411915768, + "grad_norm": 2.6483540534973145, + "learning_rate": 3.3555386372287783e-06, + "loss": 1.4755, + "step": 1172 + }, + { + "epoch": 1.2049306625577811, + "grad_norm": 2.5579090118408203, + "learning_rate": 3.3536353254663116e-06, + "loss": 1.3732, + "step": 1173 + }, + { + "epoch": 1.2059578839239857, + "grad_norm": 2.5310683250427246, + "learning_rate": 3.351732013703845e-06, + "loss": 1.4077, + "step": 1174 + }, + { + "epoch": 1.20698510529019, + "grad_norm": 2.616957187652588, + "learning_rate": 3.3498287019413784e-06, + "loss": 1.4237, + "step": 1175 + }, + { + "epoch": 1.2080123266563945, + "grad_norm": 2.6008999347686768, + "learning_rate": 3.3479253901789117e-06, + "loss": 1.4381, + "step": 1176 + }, + { + "epoch": 1.2090395480225988, + "grad_norm": 2.5235824584960938, + "learning_rate": 3.346022078416445e-06, + "loss": 1.3471, + "step": 1177 + }, + { + "epoch": 1.2100667693888032, + "grad_norm": 2.3749704360961914, + "learning_rate": 3.344118766653978e-06, + "loss": 1.417, + "step": 1178 + }, + { + "epoch": 1.2110939907550078, + "grad_norm": 2.493380069732666, + "learning_rate": 3.3422154548915114e-06, + "loss": 1.3924, + "step": 1179 + }, + { + "epoch": 1.2121212121212122, + "grad_norm": 2.4233086109161377, + "learning_rate": 3.3403121431290448e-06, + "loss": 1.4257, + "step": 1180 + }, + { + "epoch": 1.2121212121212122, + "eval_loss": 1.7660146951675415, + "eval_runtime": 37.9268, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 1180 + }, + { + "epoch": 1.2131484334874165, + "grad_norm": 2.535982608795166, + "learning_rate": 3.338408831366578e-06, + "loss": 1.3587, + "step": 1181 + }, + { + "epoch": 1.214175654853621, + "grad_norm": 2.5744845867156982, + "learning_rate": 3.3365055196041115e-06, + "loss": 1.436, + "step": 1182 + }, + { + "epoch": 1.2152028762198253, + "grad_norm": 2.5544042587280273, + "learning_rate": 3.3346022078416444e-06, + "loss": 1.464, + "step": 1183 + }, + { + "epoch": 1.2162300975860298, + "grad_norm": 2.4980483055114746, + "learning_rate": 3.3326988960791778e-06, + "loss": 1.3915, + "step": 1184 + }, + { + "epoch": 1.2172573189522342, + "grad_norm": 2.403468608856201, + "learning_rate": 3.330795584316711e-06, + "loss": 1.3239, + "step": 1185 + }, + { + "epoch": 1.2182845403184386, + "grad_norm": 2.6861846446990967, + "learning_rate": 3.3288922725542445e-06, + "loss": 1.4013, + "step": 1186 + }, + { + "epoch": 1.219311761684643, + "grad_norm": 2.525033712387085, + "learning_rate": 3.326988960791778e-06, + "loss": 1.4135, + "step": 1187 + }, + { + "epoch": 1.2203389830508475, + "grad_norm": 2.6705894470214844, + "learning_rate": 3.3250856490293117e-06, + "loss": 1.4587, + "step": 1188 + }, + { + "epoch": 1.221366204417052, + "grad_norm": 2.6108217239379883, + "learning_rate": 3.323182337266845e-06, + "loss": 1.4482, + "step": 1189 + }, + { + "epoch": 1.2223934257832563, + "grad_norm": 2.5227413177490234, + "learning_rate": 3.321279025504378e-06, + "loss": 1.432, + "step": 1190 + }, + { + "epoch": 1.2223934257832563, + "eval_loss": 1.756821870803833, + "eval_runtime": 37.9929, + "eval_samples_per_second": 26.321, + "eval_steps_per_second": 1.105, + "step": 1190 + }, + { + "epoch": 1.2234206471494606, + "grad_norm": 2.469766139984131, + "learning_rate": 3.3193757137419113e-06, + "loss": 1.4215, + "step": 1191 + }, + { + "epoch": 1.2244478685156652, + "grad_norm": 2.5257861614227295, + "learning_rate": 3.3174724019794447e-06, + "loss": 1.4624, + "step": 1192 + }, + { + "epoch": 1.2254750898818696, + "grad_norm": 2.5281760692596436, + "learning_rate": 3.315569090216978e-06, + "loss": 1.3799, + "step": 1193 + }, + { + "epoch": 1.226502311248074, + "grad_norm": 2.701521396636963, + "learning_rate": 3.3136657784545114e-06, + "loss": 1.4631, + "step": 1194 + }, + { + "epoch": 1.2275295326142783, + "grad_norm": 2.4239444732666016, + "learning_rate": 3.3117624666920443e-06, + "loss": 1.3746, + "step": 1195 + }, + { + "epoch": 1.2285567539804827, + "grad_norm": 2.430377721786499, + "learning_rate": 3.3098591549295777e-06, + "loss": 1.3584, + "step": 1196 + }, + { + "epoch": 1.2295839753466873, + "grad_norm": 2.7177658081054688, + "learning_rate": 3.307955843167111e-06, + "loss": 1.4301, + "step": 1197 + }, + { + "epoch": 1.2306111967128917, + "grad_norm": 2.512943983078003, + "learning_rate": 3.3060525314046444e-06, + "loss": 1.4111, + "step": 1198 + }, + { + "epoch": 1.231638418079096, + "grad_norm": 2.5269784927368164, + "learning_rate": 3.304149219642178e-06, + "loss": 1.4096, + "step": 1199 + }, + { + "epoch": 1.2326656394453004, + "grad_norm": 2.476409673690796, + "learning_rate": 3.3022459078797107e-06, + "loss": 1.4322, + "step": 1200 + }, + { + "epoch": 1.2326656394453004, + "eval_loss": 1.7645230293273926, + "eval_runtime": 37.9898, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1200 + }, + { + "epoch": 1.2336928608115048, + "grad_norm": 2.5339248180389404, + "learning_rate": 3.300342596117244e-06, + "loss": 1.4093, + "step": 1201 + }, + { + "epoch": 1.2347200821777093, + "grad_norm": 2.3505656719207764, + "learning_rate": 3.2984392843547775e-06, + "loss": 1.3623, + "step": 1202 + }, + { + "epoch": 1.2357473035439137, + "grad_norm": 2.400259256362915, + "learning_rate": 3.296535972592311e-06, + "loss": 1.3512, + "step": 1203 + }, + { + "epoch": 1.236774524910118, + "grad_norm": 2.561558723449707, + "learning_rate": 3.294632660829844e-06, + "loss": 1.3946, + "step": 1204 + }, + { + "epoch": 1.2378017462763227, + "grad_norm": 2.623100519180298, + "learning_rate": 3.292729349067377e-06, + "loss": 1.3972, + "step": 1205 + }, + { + "epoch": 1.238828967642527, + "grad_norm": 2.4889469146728516, + "learning_rate": 3.2908260373049105e-06, + "loss": 1.381, + "step": 1206 + }, + { + "epoch": 1.2398561890087314, + "grad_norm": 2.457334518432617, + "learning_rate": 3.288922725542444e-06, + "loss": 1.4345, + "step": 1207 + }, + { + "epoch": 1.2408834103749358, + "grad_norm": 2.440988540649414, + "learning_rate": 3.2870194137799776e-06, + "loss": 1.3901, + "step": 1208 + }, + { + "epoch": 1.2419106317411401, + "grad_norm": 2.4992308616638184, + "learning_rate": 3.285116102017511e-06, + "loss": 1.3694, + "step": 1209 + }, + { + "epoch": 1.2429378531073447, + "grad_norm": 2.611128091812134, + "learning_rate": 3.2832127902550444e-06, + "loss": 1.4344, + "step": 1210 + }, + { + "epoch": 1.2429378531073447, + "eval_loss": 1.763407826423645, + "eval_runtime": 37.7142, + "eval_samples_per_second": 26.515, + "eval_steps_per_second": 1.114, + "step": 1210 + }, + { + "epoch": 1.243965074473549, + "grad_norm": 2.6573522090911865, + "learning_rate": 3.2813094784925777e-06, + "loss": 1.4204, + "step": 1211 + }, + { + "epoch": 1.2449922958397535, + "grad_norm": 2.4785778522491455, + "learning_rate": 3.2794061667301107e-06, + "loss": 1.3617, + "step": 1212 + }, + { + "epoch": 1.2460195172059578, + "grad_norm": 2.4831974506378174, + "learning_rate": 3.277502854967644e-06, + "loss": 1.4004, + "step": 1213 + }, + { + "epoch": 1.2470467385721622, + "grad_norm": 2.6142327785491943, + "learning_rate": 3.2755995432051774e-06, + "loss": 1.3881, + "step": 1214 + }, + { + "epoch": 1.2480739599383668, + "grad_norm": 2.447962999343872, + "learning_rate": 3.2736962314427107e-06, + "loss": 1.3984, + "step": 1215 + }, + { + "epoch": 1.2491011813045712, + "grad_norm": 2.3895885944366455, + "learning_rate": 3.271792919680244e-06, + "loss": 1.4099, + "step": 1216 + }, + { + "epoch": 1.2501284026707755, + "grad_norm": 2.599364995956421, + "learning_rate": 3.269889607917777e-06, + "loss": 1.4575, + "step": 1217 + }, + { + "epoch": 1.25115562403698, + "grad_norm": 2.6294758319854736, + "learning_rate": 3.2679862961553104e-06, + "loss": 1.3752, + "step": 1218 + }, + { + "epoch": 1.2521828454031843, + "grad_norm": 2.50923228263855, + "learning_rate": 3.2660829843928438e-06, + "loss": 1.4017, + "step": 1219 + }, + { + "epoch": 1.2532100667693888, + "grad_norm": 2.626434564590454, + "learning_rate": 3.264179672630377e-06, + "loss": 1.4505, + "step": 1220 + }, + { + "epoch": 1.2532100667693888, + "eval_loss": 1.7672324180603027, + "eval_runtime": 37.8605, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 1220 + }, + { + "epoch": 1.2542372881355932, + "grad_norm": 2.541663885116577, + "learning_rate": 3.2622763608679105e-06, + "loss": 1.4129, + "step": 1221 + }, + { + "epoch": 1.2552645095017976, + "grad_norm": 2.559877395629883, + "learning_rate": 3.2603730491054434e-06, + "loss": 1.3793, + "step": 1222 + }, + { + "epoch": 1.2562917308680022, + "grad_norm": 2.4085795879364014, + "learning_rate": 3.258469737342977e-06, + "loss": 1.3895, + "step": 1223 + }, + { + "epoch": 1.2573189522342065, + "grad_norm": 2.2942678928375244, + "learning_rate": 3.25656642558051e-06, + "loss": 1.3713, + "step": 1224 + }, + { + "epoch": 1.258346173600411, + "grad_norm": 2.561086893081665, + "learning_rate": 3.2546631138180435e-06, + "loss": 1.3835, + "step": 1225 + }, + { + "epoch": 1.2593733949666153, + "grad_norm": 2.4947667121887207, + "learning_rate": 3.252759802055577e-06, + "loss": 1.3793, + "step": 1226 + }, + { + "epoch": 1.2604006163328196, + "grad_norm": 2.6293253898620605, + "learning_rate": 3.25085649029311e-06, + "loss": 1.3935, + "step": 1227 + }, + { + "epoch": 1.2614278376990242, + "grad_norm": 2.5461983680725098, + "learning_rate": 3.248953178530644e-06, + "loss": 1.4201, + "step": 1228 + }, + { + "epoch": 1.2624550590652286, + "grad_norm": 2.381537675857544, + "learning_rate": 3.247049866768177e-06, + "loss": 1.3932, + "step": 1229 + }, + { + "epoch": 1.263482280431433, + "grad_norm": 2.381558418273926, + "learning_rate": 3.2451465550057103e-06, + "loss": 1.3425, + "step": 1230 + }, + { + "epoch": 1.263482280431433, + "eval_loss": 1.7713288068771362, + "eval_runtime": 37.9629, + "eval_samples_per_second": 26.342, + "eval_steps_per_second": 1.106, + "step": 1230 + }, + { + "epoch": 1.2645095017976373, + "grad_norm": 2.5877468585968018, + "learning_rate": 3.2432432432432437e-06, + "loss": 1.3917, + "step": 1231 + }, + { + "epoch": 1.2655367231638417, + "grad_norm": 2.6057848930358887, + "learning_rate": 3.241339931480777e-06, + "loss": 1.4217, + "step": 1232 + }, + { + "epoch": 1.2665639445300463, + "grad_norm": 2.5157413482666016, + "learning_rate": 3.2394366197183104e-06, + "loss": 1.427, + "step": 1233 + }, + { + "epoch": 1.2675911658962506, + "grad_norm": 2.4153692722320557, + "learning_rate": 3.2375333079558434e-06, + "loss": 1.3822, + "step": 1234 + }, + { + "epoch": 1.268618387262455, + "grad_norm": 2.5459582805633545, + "learning_rate": 3.2356299961933767e-06, + "loss": 1.4152, + "step": 1235 + }, + { + "epoch": 1.2696456086286596, + "grad_norm": 2.606468915939331, + "learning_rate": 3.23372668443091e-06, + "loss": 1.4278, + "step": 1236 + }, + { + "epoch": 1.270672829994864, + "grad_norm": 2.5061633586883545, + "learning_rate": 3.2318233726684435e-06, + "loss": 1.4512, + "step": 1237 + }, + { + "epoch": 1.2717000513610683, + "grad_norm": 2.3693087100982666, + "learning_rate": 3.229920060905977e-06, + "loss": 1.3791, + "step": 1238 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 2.479583978652954, + "learning_rate": 3.2280167491435098e-06, + "loss": 1.3814, + "step": 1239 + }, + { + "epoch": 1.273754494093477, + "grad_norm": 2.3258883953094482, + "learning_rate": 3.226113437381043e-06, + "loss": 1.3904, + "step": 1240 + }, + { + "epoch": 1.273754494093477, + "eval_loss": 1.7602728605270386, + "eval_runtime": 37.9604, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 1240 + }, + { + "epoch": 1.2747817154596817, + "grad_norm": 2.4170308113098145, + "learning_rate": 3.2242101256185765e-06, + "loss": 1.4045, + "step": 1241 + }, + { + "epoch": 1.275808936825886, + "grad_norm": 2.578066825866699, + "learning_rate": 3.22230681385611e-06, + "loss": 1.4556, + "step": 1242 + }, + { + "epoch": 1.2768361581920904, + "grad_norm": 2.558187246322632, + "learning_rate": 3.220403502093643e-06, + "loss": 1.4036, + "step": 1243 + }, + { + "epoch": 1.2778633795582948, + "grad_norm": 2.598916530609131, + "learning_rate": 3.218500190331176e-06, + "loss": 1.4462, + "step": 1244 + }, + { + "epoch": 1.2788906009244991, + "grad_norm": 2.4528257846832275, + "learning_rate": 3.2165968785687095e-06, + "loss": 1.3855, + "step": 1245 + }, + { + "epoch": 1.2799178222907037, + "grad_norm": 2.5937459468841553, + "learning_rate": 3.214693566806243e-06, + "loss": 1.4191, + "step": 1246 + }, + { + "epoch": 1.280945043656908, + "grad_norm": 2.5324676036834717, + "learning_rate": 3.2127902550437762e-06, + "loss": 1.4162, + "step": 1247 + }, + { + "epoch": 1.2819722650231125, + "grad_norm": 2.4849791526794434, + "learning_rate": 3.21088694328131e-06, + "loss": 1.4037, + "step": 1248 + }, + { + "epoch": 1.2829994863893168, + "grad_norm": 2.389582633972168, + "learning_rate": 3.2089836315188434e-06, + "loss": 1.3966, + "step": 1249 + }, + { + "epoch": 1.2840267077555212, + "grad_norm": 2.387909173965454, + "learning_rate": 3.2070803197563767e-06, + "loss": 1.396, + "step": 1250 + }, + { + "epoch": 1.2840267077555212, + "eval_loss": 1.7642556428909302, + "eval_runtime": 37.9446, + "eval_samples_per_second": 26.354, + "eval_steps_per_second": 1.107, + "step": 1250 + }, + { + "epoch": 1.2850539291217258, + "grad_norm": 2.5014612674713135, + "learning_rate": 3.2051770079939097e-06, + "loss": 1.4128, + "step": 1251 + }, + { + "epoch": 1.2860811504879301, + "grad_norm": 2.497950553894043, + "learning_rate": 3.203273696231443e-06, + "loss": 1.4412, + "step": 1252 + }, + { + "epoch": 1.2871083718541345, + "grad_norm": 2.4371297359466553, + "learning_rate": 3.2013703844689764e-06, + "loss": 1.3754, + "step": 1253 + }, + { + "epoch": 1.288135593220339, + "grad_norm": 2.4336745738983154, + "learning_rate": 3.1994670727065098e-06, + "loss": 1.4258, + "step": 1254 + }, + { + "epoch": 1.2891628145865435, + "grad_norm": 2.5165536403656006, + "learning_rate": 3.197563760944043e-06, + "loss": 1.4414, + "step": 1255 + }, + { + "epoch": 1.2901900359527478, + "grad_norm": 2.5060458183288574, + "learning_rate": 3.195660449181576e-06, + "loss": 1.4336, + "step": 1256 + }, + { + "epoch": 1.2912172573189522, + "grad_norm": 2.584284543991089, + "learning_rate": 3.1937571374191094e-06, + "loss": 1.3768, + "step": 1257 + }, + { + "epoch": 1.2922444786851566, + "grad_norm": 2.615629196166992, + "learning_rate": 3.191853825656643e-06, + "loss": 1.401, + "step": 1258 + }, + { + "epoch": 1.2932717000513612, + "grad_norm": 2.5989596843719482, + "learning_rate": 3.189950513894176e-06, + "loss": 1.3915, + "step": 1259 + }, + { + "epoch": 1.2942989214175655, + "grad_norm": 2.5507843494415283, + "learning_rate": 3.1880472021317095e-06, + "loss": 1.3979, + "step": 1260 + }, + { + "epoch": 1.2942989214175655, + "eval_loss": 1.7715612649917603, + "eval_runtime": 37.9859, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1260 + }, + { + "epoch": 1.29532614278377, + "grad_norm": 2.512140989303589, + "learning_rate": 3.1861438903692425e-06, + "loss": 1.3523, + "step": 1261 + }, + { + "epoch": 1.2963533641499743, + "grad_norm": 2.4311885833740234, + "learning_rate": 3.184240578606776e-06, + "loss": 1.3828, + "step": 1262 + }, + { + "epoch": 1.2973805855161786, + "grad_norm": 2.439265489578247, + "learning_rate": 3.182337266844309e-06, + "loss": 1.402, + "step": 1263 + }, + { + "epoch": 1.2984078068823832, + "grad_norm": 2.5773651599884033, + "learning_rate": 3.1804339550818426e-06, + "loss": 1.4362, + "step": 1264 + }, + { + "epoch": 1.2994350282485876, + "grad_norm": 2.456021547317505, + "learning_rate": 3.178530643319376e-06, + "loss": 1.3918, + "step": 1265 + }, + { + "epoch": 1.300462249614792, + "grad_norm": 2.44404673576355, + "learning_rate": 3.176627331556909e-06, + "loss": 1.3662, + "step": 1266 + }, + { + "epoch": 1.3014894709809965, + "grad_norm": 2.778027057647705, + "learning_rate": 3.1747240197944422e-06, + "loss": 1.4112, + "step": 1267 + }, + { + "epoch": 1.3025166923472007, + "grad_norm": 2.476496934890747, + "learning_rate": 3.172820708031976e-06, + "loss": 1.3654, + "step": 1268 + }, + { + "epoch": 1.3035439137134053, + "grad_norm": 2.6415998935699463, + "learning_rate": 3.1709173962695094e-06, + "loss": 1.4441, + "step": 1269 + }, + { + "epoch": 1.3045711350796096, + "grad_norm": 2.5046234130859375, + "learning_rate": 3.1690140845070427e-06, + "loss": 1.4316, + "step": 1270 + }, + { + "epoch": 1.3045711350796096, + "eval_loss": 1.7736135721206665, + "eval_runtime": 37.9392, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1270 + }, + { + "epoch": 1.305598356445814, + "grad_norm": 2.6234474182128906, + "learning_rate": 3.167110772744576e-06, + "loss": 1.4, + "step": 1271 + }, + { + "epoch": 1.3066255778120186, + "grad_norm": 2.488297700881958, + "learning_rate": 3.1652074609821095e-06, + "loss": 1.3908, + "step": 1272 + }, + { + "epoch": 1.307652799178223, + "grad_norm": 2.3759512901306152, + "learning_rate": 3.1633041492196424e-06, + "loss": 1.3908, + "step": 1273 + }, + { + "epoch": 1.3086800205444273, + "grad_norm": 2.487281084060669, + "learning_rate": 3.1614008374571758e-06, + "loss": 1.4217, + "step": 1274 + }, + { + "epoch": 1.3097072419106317, + "grad_norm": 2.5366780757904053, + "learning_rate": 3.159497525694709e-06, + "loss": 1.4229, + "step": 1275 + }, + { + "epoch": 1.310734463276836, + "grad_norm": 2.416141986846924, + "learning_rate": 3.1575942139322425e-06, + "loss": 1.3865, + "step": 1276 + }, + { + "epoch": 1.3117616846430407, + "grad_norm": 2.537964344024658, + "learning_rate": 3.155690902169776e-06, + "loss": 1.4141, + "step": 1277 + }, + { + "epoch": 1.312788906009245, + "grad_norm": 2.481692314147949, + "learning_rate": 3.1537875904073088e-06, + "loss": 1.383, + "step": 1278 + }, + { + "epoch": 1.3138161273754494, + "grad_norm": 2.4898593425750732, + "learning_rate": 3.151884278644842e-06, + "loss": 1.4078, + "step": 1279 + }, + { + "epoch": 1.3148433487416538, + "grad_norm": 2.4692323207855225, + "learning_rate": 3.1499809668823755e-06, + "loss": 1.3818, + "step": 1280 + }, + { + "epoch": 1.3148433487416538, + "eval_loss": 1.7660051584243774, + "eval_runtime": 37.8922, + "eval_samples_per_second": 26.391, + "eval_steps_per_second": 1.108, + "step": 1280 + }, + { + "epoch": 1.3158705701078581, + "grad_norm": 2.475862503051758, + "learning_rate": 3.148077655119909e-06, + "loss": 1.4384, + "step": 1281 + }, + { + "epoch": 1.3168977914740627, + "grad_norm": 2.2839648723602295, + "learning_rate": 3.1461743433574422e-06, + "loss": 1.3722, + "step": 1282 + }, + { + "epoch": 1.317925012840267, + "grad_norm": 2.3965585231781006, + "learning_rate": 3.144271031594975e-06, + "loss": 1.4262, + "step": 1283 + }, + { + "epoch": 1.3189522342064715, + "grad_norm": 2.4107236862182617, + "learning_rate": 3.1423677198325085e-06, + "loss": 1.4216, + "step": 1284 + }, + { + "epoch": 1.319979455572676, + "grad_norm": 2.438899040222168, + "learning_rate": 3.140464408070042e-06, + "loss": 1.383, + "step": 1285 + }, + { + "epoch": 1.3210066769388804, + "grad_norm": 2.4146060943603516, + "learning_rate": 3.1385610963075753e-06, + "loss": 1.3827, + "step": 1286 + }, + { + "epoch": 1.3220338983050848, + "grad_norm": 2.4193496704101562, + "learning_rate": 3.1366577845451086e-06, + "loss": 1.3707, + "step": 1287 + }, + { + "epoch": 1.3230611196712891, + "grad_norm": 2.5284323692321777, + "learning_rate": 3.134754472782642e-06, + "loss": 1.3623, + "step": 1288 + }, + { + "epoch": 1.3240883410374935, + "grad_norm": 2.421509265899658, + "learning_rate": 3.1328511610201758e-06, + "loss": 1.35, + "step": 1289 + }, + { + "epoch": 1.325115562403698, + "grad_norm": 2.3964431285858154, + "learning_rate": 3.1309478492577087e-06, + "loss": 1.3486, + "step": 1290 + }, + { + "epoch": 1.325115562403698, + "eval_loss": 1.7699168920516968, + "eval_runtime": 37.9003, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 1290 + }, + { + "epoch": 1.3261427837699025, + "grad_norm": 2.4371817111968994, + "learning_rate": 3.129044537495242e-06, + "loss": 1.3958, + "step": 1291 + }, + { + "epoch": 1.3271700051361068, + "grad_norm": 2.45233154296875, + "learning_rate": 3.1271412257327754e-06, + "loss": 1.4175, + "step": 1292 + }, + { + "epoch": 1.3281972265023112, + "grad_norm": 2.314384937286377, + "learning_rate": 3.125237913970309e-06, + "loss": 1.3701, + "step": 1293 + }, + { + "epoch": 1.3292244478685156, + "grad_norm": 2.5603694915771484, + "learning_rate": 3.123334602207842e-06, + "loss": 1.3489, + "step": 1294 + }, + { + "epoch": 1.3302516692347202, + "grad_norm": 2.5919623374938965, + "learning_rate": 3.121431290445375e-06, + "loss": 1.4472, + "step": 1295 + }, + { + "epoch": 1.3312788906009245, + "grad_norm": 2.6912758350372314, + "learning_rate": 3.1195279786829085e-06, + "loss": 1.4917, + "step": 1296 + }, + { + "epoch": 1.332306111967129, + "grad_norm": 2.439089298248291, + "learning_rate": 3.117624666920442e-06, + "loss": 1.3776, + "step": 1297 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 2.5207953453063965, + "learning_rate": 3.115721355157975e-06, + "loss": 1.3784, + "step": 1298 + }, + { + "epoch": 1.3343605546995376, + "grad_norm": 2.507791519165039, + "learning_rate": 3.1138180433955085e-06, + "loss": 1.3346, + "step": 1299 + }, + { + "epoch": 1.3353877760657422, + "grad_norm": 2.450531482696533, + "learning_rate": 3.1119147316330415e-06, + "loss": 1.4151, + "step": 1300 + }, + { + "epoch": 1.3353877760657422, + "eval_loss": 1.7761608362197876, + "eval_runtime": 37.9041, + "eval_samples_per_second": 26.382, + "eval_steps_per_second": 1.108, + "step": 1300 + }, + { + "epoch": 1.3364149974319466, + "grad_norm": 2.488255262374878, + "learning_rate": 3.110011419870575e-06, + "loss": 1.4475, + "step": 1301 + }, + { + "epoch": 1.337442218798151, + "grad_norm": 2.5059807300567627, + "learning_rate": 3.1081081081081082e-06, + "loss": 1.3812, + "step": 1302 + }, + { + "epoch": 1.3384694401643555, + "grad_norm": 2.5579488277435303, + "learning_rate": 3.1062047963456416e-06, + "loss": 1.4134, + "step": 1303 + }, + { + "epoch": 1.33949666153056, + "grad_norm": 2.6170148849487305, + "learning_rate": 3.104301484583175e-06, + "loss": 1.459, + "step": 1304 + }, + { + "epoch": 1.3405238828967643, + "grad_norm": 2.5572125911712646, + "learning_rate": 3.1023981728207083e-06, + "loss": 1.4354, + "step": 1305 + }, + { + "epoch": 1.3415511042629686, + "grad_norm": 2.396883487701416, + "learning_rate": 3.1004948610582412e-06, + "loss": 1.4016, + "step": 1306 + }, + { + "epoch": 1.342578325629173, + "grad_norm": 2.5869390964508057, + "learning_rate": 3.0985915492957746e-06, + "loss": 1.3945, + "step": 1307 + }, + { + "epoch": 1.3436055469953776, + "grad_norm": 2.4271085262298584, + "learning_rate": 3.096688237533308e-06, + "loss": 1.3506, + "step": 1308 + }, + { + "epoch": 1.344632768361582, + "grad_norm": 2.398407220840454, + "learning_rate": 3.0947849257708418e-06, + "loss": 1.4475, + "step": 1309 + }, + { + "epoch": 1.3456599897277863, + "grad_norm": 2.441934823989868, + "learning_rate": 3.092881614008375e-06, + "loss": 1.3972, + "step": 1310 + }, + { + "epoch": 1.3456599897277863, + "eval_loss": 1.774910569190979, + "eval_runtime": 37.6105, + "eval_samples_per_second": 26.588, + "eval_steps_per_second": 1.117, + "step": 1310 + }, + { + "epoch": 1.3466872110939907, + "grad_norm": 2.5015370845794678, + "learning_rate": 3.0909783022459085e-06, + "loss": 1.4418, + "step": 1311 + }, + { + "epoch": 1.347714432460195, + "grad_norm": 2.4616997241973877, + "learning_rate": 3.0890749904834414e-06, + "loss": 1.3833, + "step": 1312 + }, + { + "epoch": 1.3487416538263997, + "grad_norm": 2.5993902683258057, + "learning_rate": 3.0871716787209748e-06, + "loss": 1.4385, + "step": 1313 + }, + { + "epoch": 1.349768875192604, + "grad_norm": 2.518195867538452, + "learning_rate": 3.085268366958508e-06, + "loss": 1.4127, + "step": 1314 + }, + { + "epoch": 1.3507960965588084, + "grad_norm": 2.545189380645752, + "learning_rate": 3.0833650551960415e-06, + "loss": 1.42, + "step": 1315 + }, + { + "epoch": 1.3518233179250128, + "grad_norm": 2.461818218231201, + "learning_rate": 3.081461743433575e-06, + "loss": 1.3796, + "step": 1316 + }, + { + "epoch": 1.3528505392912171, + "grad_norm": 2.4081408977508545, + "learning_rate": 3.079558431671108e-06, + "loss": 1.3939, + "step": 1317 + }, + { + "epoch": 1.3538777606574217, + "grad_norm": 2.2885539531707764, + "learning_rate": 3.077655119908641e-06, + "loss": 1.3645, + "step": 1318 + }, + { + "epoch": 1.354904982023626, + "grad_norm": 2.453791618347168, + "learning_rate": 3.0757518081461745e-06, + "loss": 1.3906, + "step": 1319 + }, + { + "epoch": 1.3559322033898304, + "grad_norm": 2.4538352489471436, + "learning_rate": 3.073848496383708e-06, + "loss": 1.4127, + "step": 1320 + }, + { + "epoch": 1.3559322033898304, + "eval_loss": 1.7821755409240723, + "eval_runtime": 37.7585, + "eval_samples_per_second": 26.484, + "eval_steps_per_second": 1.112, + "step": 1320 + }, + { + "epoch": 1.356959424756035, + "grad_norm": 2.438425064086914, + "learning_rate": 3.0719451846212413e-06, + "loss": 1.3763, + "step": 1321 + }, + { + "epoch": 1.3579866461222394, + "grad_norm": 2.4398279190063477, + "learning_rate": 3.0700418728587746e-06, + "loss": 1.3963, + "step": 1322 + }, + { + "epoch": 1.3590138674884438, + "grad_norm": 2.6763861179351807, + "learning_rate": 3.0681385610963076e-06, + "loss": 1.3837, + "step": 1323 + }, + { + "epoch": 1.3600410888546481, + "grad_norm": 2.685056209564209, + "learning_rate": 3.066235249333841e-06, + "loss": 1.4158, + "step": 1324 + }, + { + "epoch": 1.3610683102208525, + "grad_norm": 2.6663355827331543, + "learning_rate": 3.0643319375713743e-06, + "loss": 1.4533, + "step": 1325 + }, + { + "epoch": 1.362095531587057, + "grad_norm": 2.4613428115844727, + "learning_rate": 3.0624286258089076e-06, + "loss": 1.3927, + "step": 1326 + }, + { + "epoch": 1.3631227529532615, + "grad_norm": 2.348334789276123, + "learning_rate": 3.060525314046441e-06, + "loss": 1.3782, + "step": 1327 + }, + { + "epoch": 1.3641499743194658, + "grad_norm": 2.5011332035064697, + "learning_rate": 3.058622002283974e-06, + "loss": 1.4175, + "step": 1328 + }, + { + "epoch": 1.3651771956856702, + "grad_norm": 2.4692256450653076, + "learning_rate": 3.0567186905215077e-06, + "loss": 1.4319, + "step": 1329 + }, + { + "epoch": 1.3662044170518746, + "grad_norm": 2.616990089416504, + "learning_rate": 3.054815378759041e-06, + "loss": 1.4264, + "step": 1330 + }, + { + "epoch": 1.3662044170518746, + "eval_loss": 1.781583309173584, + "eval_runtime": 37.9291, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 1330 + }, + { + "epoch": 1.3672316384180792, + "grad_norm": 2.5416364669799805, + "learning_rate": 3.0529120669965745e-06, + "loss": 1.4249, + "step": 1331 + }, + { + "epoch": 1.3682588597842835, + "grad_norm": 2.5346689224243164, + "learning_rate": 3.051008755234108e-06, + "loss": 1.3647, + "step": 1332 + }, + { + "epoch": 1.3692860811504879, + "grad_norm": 2.585714340209961, + "learning_rate": 3.049105443471641e-06, + "loss": 1.3714, + "step": 1333 + }, + { + "epoch": 1.3703133025166925, + "grad_norm": 2.5958940982818604, + "learning_rate": 3.0472021317091745e-06, + "loss": 1.5041, + "step": 1334 + }, + { + "epoch": 1.3713405238828968, + "grad_norm": 2.4795925617218018, + "learning_rate": 3.0452988199467075e-06, + "loss": 1.4317, + "step": 1335 + }, + { + "epoch": 1.3723677452491012, + "grad_norm": 2.4608285427093506, + "learning_rate": 3.043395508184241e-06, + "loss": 1.4259, + "step": 1336 + }, + { + "epoch": 1.3733949666153056, + "grad_norm": 2.5970380306243896, + "learning_rate": 3.041492196421774e-06, + "loss": 1.3859, + "step": 1337 + }, + { + "epoch": 1.37442218798151, + "grad_norm": 2.517047166824341, + "learning_rate": 3.0395888846593076e-06, + "loss": 1.4086, + "step": 1338 + }, + { + "epoch": 1.3754494093477145, + "grad_norm": 2.4541358947753906, + "learning_rate": 3.037685572896841e-06, + "loss": 1.3959, + "step": 1339 + }, + { + "epoch": 1.376476630713919, + "grad_norm": 2.6143250465393066, + "learning_rate": 3.035782261134374e-06, + "loss": 1.4088, + "step": 1340 + }, + { + "epoch": 1.376476630713919, + "eval_loss": 1.7629026174545288, + "eval_runtime": 38.0156, + "eval_samples_per_second": 26.305, + "eval_steps_per_second": 1.105, + "step": 1340 + }, + { + "epoch": 1.3775038520801233, + "grad_norm": 2.763033151626587, + "learning_rate": 3.0338789493719072e-06, + "loss": 1.4104, + "step": 1341 + }, + { + "epoch": 1.3785310734463276, + "grad_norm": 2.523888349533081, + "learning_rate": 3.0319756376094406e-06, + "loss": 1.4073, + "step": 1342 + }, + { + "epoch": 1.379558294812532, + "grad_norm": 2.5393309593200684, + "learning_rate": 3.030072325846974e-06, + "loss": 1.4145, + "step": 1343 + }, + { + "epoch": 1.3805855161787366, + "grad_norm": 2.455733060836792, + "learning_rate": 3.0281690140845073e-06, + "loss": 1.3782, + "step": 1344 + }, + { + "epoch": 1.381612737544941, + "grad_norm": 2.600311517715454, + "learning_rate": 3.0262657023220403e-06, + "loss": 1.4158, + "step": 1345 + }, + { + "epoch": 1.3826399589111453, + "grad_norm": 2.453249931335449, + "learning_rate": 3.0243623905595736e-06, + "loss": 1.414, + "step": 1346 + }, + { + "epoch": 1.3836671802773497, + "grad_norm": 2.603515625, + "learning_rate": 3.022459078797107e-06, + "loss": 1.4102, + "step": 1347 + }, + { + "epoch": 1.384694401643554, + "grad_norm": 2.6515872478485107, + "learning_rate": 3.0205557670346404e-06, + "loss": 1.4063, + "step": 1348 + }, + { + "epoch": 1.3857216230097587, + "grad_norm": 2.412846565246582, + "learning_rate": 3.018652455272174e-06, + "loss": 1.3739, + "step": 1349 + }, + { + "epoch": 1.386748844375963, + "grad_norm": 2.748563051223755, + "learning_rate": 3.0167491435097075e-06, + "loss": 1.4909, + "step": 1350 + }, + { + "epoch": 1.386748844375963, + "eval_loss": 1.7653924226760864, + "eval_runtime": 37.9855, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1350 + }, + { + "epoch": 1.3877760657421674, + "grad_norm": 2.7861363887786865, + "learning_rate": 3.014845831747241e-06, + "loss": 1.4028, + "step": 1351 + }, + { + "epoch": 1.388803287108372, + "grad_norm": 2.5898189544677734, + "learning_rate": 3.012942519984774e-06, + "loss": 1.3793, + "step": 1352 + }, + { + "epoch": 1.3898305084745763, + "grad_norm": 2.4240355491638184, + "learning_rate": 3.011039208222307e-06, + "loss": 1.3889, + "step": 1353 + }, + { + "epoch": 1.3908577298407807, + "grad_norm": 2.4492504596710205, + "learning_rate": 3.0091358964598405e-06, + "loss": 1.3464, + "step": 1354 + }, + { + "epoch": 1.391884951206985, + "grad_norm": 2.4425718784332275, + "learning_rate": 3.007232584697374e-06, + "loss": 1.4073, + "step": 1355 + }, + { + "epoch": 1.3929121725731894, + "grad_norm": 2.5057594776153564, + "learning_rate": 3.0053292729349073e-06, + "loss": 1.4053, + "step": 1356 + }, + { + "epoch": 1.393939393939394, + "grad_norm": 2.5386273860931396, + "learning_rate": 3.00342596117244e-06, + "loss": 1.3991, + "step": 1357 + }, + { + "epoch": 1.3949666153055984, + "grad_norm": 2.6289730072021484, + "learning_rate": 3.0015226494099736e-06, + "loss": 1.3971, + "step": 1358 + }, + { + "epoch": 1.3959938366718028, + "grad_norm": 2.5091774463653564, + "learning_rate": 2.999619337647507e-06, + "loss": 1.3815, + "step": 1359 + }, + { + "epoch": 1.3970210580380071, + "grad_norm": 2.567659616470337, + "learning_rate": 2.9977160258850403e-06, + "loss": 1.3786, + "step": 1360 + }, + { + "epoch": 1.3970210580380071, + "eval_loss": 1.7706127166748047, + "eval_runtime": 37.9897, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1360 + }, + { + "epoch": 1.3980482794042115, + "grad_norm": 2.5508553981781006, + "learning_rate": 2.9958127141225736e-06, + "loss": 1.3629, + "step": 1361 + }, + { + "epoch": 1.399075500770416, + "grad_norm": 2.624800205230713, + "learning_rate": 2.9939094023601066e-06, + "loss": 1.4355, + "step": 1362 + }, + { + "epoch": 1.4001027221366205, + "grad_norm": 2.432114839553833, + "learning_rate": 2.99200609059764e-06, + "loss": 1.3995, + "step": 1363 + }, + { + "epoch": 1.4011299435028248, + "grad_norm": 2.4633607864379883, + "learning_rate": 2.9901027788351733e-06, + "loss": 1.3499, + "step": 1364 + }, + { + "epoch": 1.4021571648690292, + "grad_norm": 2.557415246963501, + "learning_rate": 2.9881994670727067e-06, + "loss": 1.3614, + "step": 1365 + }, + { + "epoch": 1.4031843862352336, + "grad_norm": 2.411295175552368, + "learning_rate": 2.98629615531024e-06, + "loss": 1.3857, + "step": 1366 + }, + { + "epoch": 1.4042116076014381, + "grad_norm": 2.429797410964966, + "learning_rate": 2.984392843547773e-06, + "loss": 1.3701, + "step": 1367 + }, + { + "epoch": 1.4052388289676425, + "grad_norm": 2.6080522537231445, + "learning_rate": 2.9824895317853063e-06, + "loss": 1.3996, + "step": 1368 + }, + { + "epoch": 1.4062660503338469, + "grad_norm": 2.5128798484802246, + "learning_rate": 2.98058622002284e-06, + "loss": 1.3973, + "step": 1369 + }, + { + "epoch": 1.4072932717000515, + "grad_norm": 2.5906126499176025, + "learning_rate": 2.9786829082603735e-06, + "loss": 1.4412, + "step": 1370 + }, + { + "epoch": 1.4072932717000515, + "eval_loss": 1.7700779438018799, + "eval_runtime": 38.017, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 1370 + }, + { + "epoch": 1.4083204930662558, + "grad_norm": 2.4890248775482178, + "learning_rate": 2.976779596497907e-06, + "loss": 1.4251, + "step": 1371 + }, + { + "epoch": 1.4093477144324602, + "grad_norm": 2.3803439140319824, + "learning_rate": 2.97487628473544e-06, + "loss": 1.4363, + "step": 1372 + }, + { + "epoch": 1.4103749357986646, + "grad_norm": 2.461958646774292, + "learning_rate": 2.9729729729729736e-06, + "loss": 1.3948, + "step": 1373 + }, + { + "epoch": 1.411402157164869, + "grad_norm": 2.419793128967285, + "learning_rate": 2.9710696612105065e-06, + "loss": 1.4367, + "step": 1374 + }, + { + "epoch": 1.4124293785310735, + "grad_norm": 2.5748846530914307, + "learning_rate": 2.96916634944804e-06, + "loss": 1.3873, + "step": 1375 + }, + { + "epoch": 1.413456599897278, + "grad_norm": 2.564558506011963, + "learning_rate": 2.9672630376855732e-06, + "loss": 1.4395, + "step": 1376 + }, + { + "epoch": 1.4144838212634823, + "grad_norm": 2.596085548400879, + "learning_rate": 2.9653597259231066e-06, + "loss": 1.4091, + "step": 1377 + }, + { + "epoch": 1.4155110426296866, + "grad_norm": 2.49029278755188, + "learning_rate": 2.96345641416064e-06, + "loss": 1.3639, + "step": 1378 + }, + { + "epoch": 1.416538263995891, + "grad_norm": 2.506887912750244, + "learning_rate": 2.961553102398173e-06, + "loss": 1.3065, + "step": 1379 + }, + { + "epoch": 1.4175654853620956, + "grad_norm": 2.517609119415283, + "learning_rate": 2.9596497906357063e-06, + "loss": 1.4365, + "step": 1380 + }, + { + "epoch": 1.4175654853620956, + "eval_loss": 1.7673696279525757, + "eval_runtime": 37.9941, + "eval_samples_per_second": 26.32, + "eval_steps_per_second": 1.105, + "step": 1380 + }, + { + "epoch": 1.4185927067283, + "grad_norm": 2.6870028972625732, + "learning_rate": 2.9577464788732396e-06, + "loss": 1.4386, + "step": 1381 + }, + { + "epoch": 1.4196199280945043, + "grad_norm": 2.3195550441741943, + "learning_rate": 2.955843167110773e-06, + "loss": 1.3771, + "step": 1382 + }, + { + "epoch": 1.420647149460709, + "grad_norm": 2.4306302070617676, + "learning_rate": 2.9539398553483064e-06, + "loss": 1.3963, + "step": 1383 + }, + { + "epoch": 1.4216743708269133, + "grad_norm": 2.4227781295776367, + "learning_rate": 2.9520365435858393e-06, + "loss": 1.394, + "step": 1384 + }, + { + "epoch": 1.4227015921931176, + "grad_norm": 2.521425485610962, + "learning_rate": 2.9501332318233727e-06, + "loss": 1.4664, + "step": 1385 + }, + { + "epoch": 1.423728813559322, + "grad_norm": 2.533363103866577, + "learning_rate": 2.948229920060906e-06, + "loss": 1.4072, + "step": 1386 + }, + { + "epoch": 1.4247560349255264, + "grad_norm": 2.378007173538208, + "learning_rate": 2.9463266082984394e-06, + "loss": 1.38, + "step": 1387 + }, + { + "epoch": 1.425783256291731, + "grad_norm": 2.476451873779297, + "learning_rate": 2.9444232965359727e-06, + "loss": 1.3864, + "step": 1388 + }, + { + "epoch": 1.4268104776579353, + "grad_norm": 2.428365707397461, + "learning_rate": 2.9425199847735057e-06, + "loss": 1.4081, + "step": 1389 + }, + { + "epoch": 1.4278376990241397, + "grad_norm": 2.4382991790771484, + "learning_rate": 2.94061667301104e-06, + "loss": 1.3806, + "step": 1390 + }, + { + "epoch": 1.4278376990241397, + "eval_loss": 1.7905043363571167, + "eval_runtime": 38.0264, + "eval_samples_per_second": 26.298, + "eval_steps_per_second": 1.104, + "step": 1390 + }, + { + "epoch": 1.428864920390344, + "grad_norm": 2.414433240890503, + "learning_rate": 2.938713361248573e-06, + "loss": 1.3113, + "step": 1391 + }, + { + "epoch": 1.4298921417565484, + "grad_norm": 2.450570821762085, + "learning_rate": 2.936810049486106e-06, + "loss": 1.4023, + "step": 1392 + }, + { + "epoch": 1.430919363122753, + "grad_norm": 2.4372148513793945, + "learning_rate": 2.9349067377236396e-06, + "loss": 1.4057, + "step": 1393 + }, + { + "epoch": 1.4319465844889574, + "grad_norm": 2.4278311729431152, + "learning_rate": 2.933003425961173e-06, + "loss": 1.3574, + "step": 1394 + }, + { + "epoch": 1.4329738058551618, + "grad_norm": 2.448517322540283, + "learning_rate": 2.9311001141987063e-06, + "loss": 1.3722, + "step": 1395 + }, + { + "epoch": 1.4340010272213661, + "grad_norm": 2.354310989379883, + "learning_rate": 2.9291968024362392e-06, + "loss": 1.3682, + "step": 1396 + }, + { + "epoch": 1.4350282485875705, + "grad_norm": 2.3318262100219727, + "learning_rate": 2.9272934906737726e-06, + "loss": 1.3604, + "step": 1397 + }, + { + "epoch": 1.436055469953775, + "grad_norm": 2.377561092376709, + "learning_rate": 2.925390178911306e-06, + "loss": 1.372, + "step": 1398 + }, + { + "epoch": 1.4370826913199795, + "grad_norm": 2.5543324947357178, + "learning_rate": 2.9234868671488393e-06, + "loss": 1.3846, + "step": 1399 + }, + { + "epoch": 1.4381099126861838, + "grad_norm": 2.6063973903656006, + "learning_rate": 2.9215835553863727e-06, + "loss": 1.3989, + "step": 1400 + }, + { + "epoch": 1.4381099126861838, + "eval_loss": 1.7730437517166138, + "eval_runtime": 38.0407, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 1400 + }, + { + "epoch": 1.4391371340523884, + "grad_norm": 2.4241244792938232, + "learning_rate": 2.9196802436239056e-06, + "loss": 1.3945, + "step": 1401 + }, + { + "epoch": 1.4401643554185928, + "grad_norm": 2.306528329849243, + "learning_rate": 2.917776931861439e-06, + "loss": 1.3686, + "step": 1402 + }, + { + "epoch": 1.4411915767847971, + "grad_norm": 2.4980850219726562, + "learning_rate": 2.9158736200989723e-06, + "loss": 1.3714, + "step": 1403 + }, + { + "epoch": 1.4422187981510015, + "grad_norm": 2.385195732116699, + "learning_rate": 2.9139703083365057e-06, + "loss": 1.3706, + "step": 1404 + }, + { + "epoch": 1.4432460195172059, + "grad_norm": 2.4087095260620117, + "learning_rate": 2.912066996574039e-06, + "loss": 1.3967, + "step": 1405 + }, + { + "epoch": 1.4442732408834105, + "grad_norm": 2.470900535583496, + "learning_rate": 2.910163684811572e-06, + "loss": 1.3351, + "step": 1406 + }, + { + "epoch": 1.4453004622496148, + "grad_norm": 2.8330743312835693, + "learning_rate": 2.9082603730491054e-06, + "loss": 1.4413, + "step": 1407 + }, + { + "epoch": 1.4463276836158192, + "grad_norm": 2.5553359985351562, + "learning_rate": 2.9063570612866387e-06, + "loss": 1.4207, + "step": 1408 + }, + { + "epoch": 1.4473549049820236, + "grad_norm": 2.509373903274536, + "learning_rate": 2.904453749524172e-06, + "loss": 1.3387, + "step": 1409 + }, + { + "epoch": 1.448382126348228, + "grad_norm": 2.5485401153564453, + "learning_rate": 2.902550437761706e-06, + "loss": 1.3765, + "step": 1410 + }, + { + "epoch": 1.448382126348228, + "eval_loss": 1.7703789472579956, + "eval_runtime": 37.7078, + "eval_samples_per_second": 26.52, + "eval_steps_per_second": 1.114, + "step": 1410 + }, + { + "epoch": 1.4494093477144325, + "grad_norm": 2.502002716064453, + "learning_rate": 2.9006471259992392e-06, + "loss": 1.3842, + "step": 1411 + }, + { + "epoch": 1.450436569080637, + "grad_norm": 2.4621520042419434, + "learning_rate": 2.8987438142367726e-06, + "loss": 1.3809, + "step": 1412 + }, + { + "epoch": 1.4514637904468413, + "grad_norm": 2.4255263805389404, + "learning_rate": 2.8968405024743055e-06, + "loss": 1.4411, + "step": 1413 + }, + { + "epoch": 1.4524910118130456, + "grad_norm": 2.2450802326202393, + "learning_rate": 2.894937190711839e-06, + "loss": 1.3389, + "step": 1414 + }, + { + "epoch": 1.45351823317925, + "grad_norm": 2.502903938293457, + "learning_rate": 2.8930338789493723e-06, + "loss": 1.4066, + "step": 1415 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 2.3685200214385986, + "learning_rate": 2.8911305671869056e-06, + "loss": 1.3995, + "step": 1416 + }, + { + "epoch": 1.455572675911659, + "grad_norm": 2.355095624923706, + "learning_rate": 2.889227255424439e-06, + "loss": 1.3352, + "step": 1417 + }, + { + "epoch": 1.4565998972778633, + "grad_norm": 2.363035202026367, + "learning_rate": 2.887323943661972e-06, + "loss": 1.3792, + "step": 1418 + }, + { + "epoch": 1.457627118644068, + "grad_norm": 2.451078176498413, + "learning_rate": 2.8854206318995053e-06, + "loss": 1.358, + "step": 1419 + }, + { + "epoch": 1.4586543400102723, + "grad_norm": 2.554501533508301, + "learning_rate": 2.8835173201370386e-06, + "loss": 1.3811, + "step": 1420 + }, + { + "epoch": 1.4586543400102723, + "eval_loss": 1.773111343383789, + "eval_runtime": 37.7693, + "eval_samples_per_second": 26.477, + "eval_steps_per_second": 1.112, + "step": 1420 + }, + { + "epoch": 1.4596815613764766, + "grad_norm": 2.556537389755249, + "learning_rate": 2.881614008374572e-06, + "loss": 1.4397, + "step": 1421 + }, + { + "epoch": 1.460708782742681, + "grad_norm": 2.5945091247558594, + "learning_rate": 2.8797106966121054e-06, + "loss": 1.3976, + "step": 1422 + }, + { + "epoch": 1.4617360041088854, + "grad_norm": 2.5158495903015137, + "learning_rate": 2.8778073848496383e-06, + "loss": 1.381, + "step": 1423 + }, + { + "epoch": 1.46276322547509, + "grad_norm": 2.359591007232666, + "learning_rate": 2.8759040730871717e-06, + "loss": 1.4156, + "step": 1424 + }, + { + "epoch": 1.4637904468412943, + "grad_norm": 2.3388748168945312, + "learning_rate": 2.874000761324705e-06, + "loss": 1.3808, + "step": 1425 + }, + { + "epoch": 1.4648176682074987, + "grad_norm": 2.438465118408203, + "learning_rate": 2.8720974495622384e-06, + "loss": 1.4158, + "step": 1426 + }, + { + "epoch": 1.465844889573703, + "grad_norm": 2.5624589920043945, + "learning_rate": 2.8701941377997718e-06, + "loss": 1.3917, + "step": 1427 + }, + { + "epoch": 1.4668721109399074, + "grad_norm": 2.467921495437622, + "learning_rate": 2.8682908260373047e-06, + "loss": 1.4316, + "step": 1428 + }, + { + "epoch": 1.467899332306112, + "grad_norm": 2.6034083366394043, + "learning_rate": 2.866387514274838e-06, + "loss": 1.4058, + "step": 1429 + }, + { + "epoch": 1.4689265536723164, + "grad_norm": 2.560844898223877, + "learning_rate": 2.864484202512372e-06, + "loss": 1.4114, + "step": 1430 + }, + { + "epoch": 1.4689265536723164, + "eval_loss": 1.7799324989318848, + "eval_runtime": 37.8814, + "eval_samples_per_second": 26.398, + "eval_steps_per_second": 1.109, + "step": 1430 + }, + { + "epoch": 1.4699537750385208, + "grad_norm": 2.469799280166626, + "learning_rate": 2.8625808907499052e-06, + "loss": 1.3767, + "step": 1431 + }, + { + "epoch": 1.4709809964047253, + "grad_norm": 2.408665895462036, + "learning_rate": 2.8606775789874386e-06, + "loss": 1.4064, + "step": 1432 + }, + { + "epoch": 1.4720082177709297, + "grad_norm": 2.497950553894043, + "learning_rate": 2.858774267224972e-06, + "loss": 1.3747, + "step": 1433 + }, + { + "epoch": 1.473035439137134, + "grad_norm": 2.6274573802948, + "learning_rate": 2.8568709554625053e-06, + "loss": 1.3994, + "step": 1434 + }, + { + "epoch": 1.4740626605033385, + "grad_norm": 2.545614719390869, + "learning_rate": 2.8549676437000382e-06, + "loss": 1.4354, + "step": 1435 + }, + { + "epoch": 1.4750898818695428, + "grad_norm": 2.796013355255127, + "learning_rate": 2.8530643319375716e-06, + "loss": 1.4081, + "step": 1436 + }, + { + "epoch": 1.4761171032357474, + "grad_norm": 2.68571400642395, + "learning_rate": 2.851161020175105e-06, + "loss": 1.3833, + "step": 1437 + }, + { + "epoch": 1.4771443246019518, + "grad_norm": 2.455244541168213, + "learning_rate": 2.8492577084126383e-06, + "loss": 1.3587, + "step": 1438 + }, + { + "epoch": 1.4781715459681561, + "grad_norm": 2.552747964859009, + "learning_rate": 2.8473543966501717e-06, + "loss": 1.4004, + "step": 1439 + }, + { + "epoch": 1.4791987673343605, + "grad_norm": 2.9428722858428955, + "learning_rate": 2.8454510848877046e-06, + "loss": 1.437, + "step": 1440 + }, + { + "epoch": 1.4791987673343605, + "eval_loss": 1.7853318452835083, + "eval_runtime": 37.8848, + "eval_samples_per_second": 26.396, + "eval_steps_per_second": 1.109, + "step": 1440 + }, + { + "epoch": 1.4802259887005649, + "grad_norm": 2.686199188232422, + "learning_rate": 2.843547773125238e-06, + "loss": 1.3901, + "step": 1441 + }, + { + "epoch": 1.4812532100667695, + "grad_norm": 2.47476863861084, + "learning_rate": 2.8416444613627714e-06, + "loss": 1.3887, + "step": 1442 + }, + { + "epoch": 1.4822804314329738, + "grad_norm": 2.425445795059204, + "learning_rate": 2.8397411496003047e-06, + "loss": 1.3502, + "step": 1443 + }, + { + "epoch": 1.4833076527991782, + "grad_norm": 2.529341459274292, + "learning_rate": 2.837837837837838e-06, + "loss": 1.3973, + "step": 1444 + }, + { + "epoch": 1.4843348741653826, + "grad_norm": 2.3861210346221924, + "learning_rate": 2.835934526075371e-06, + "loss": 1.4072, + "step": 1445 + }, + { + "epoch": 1.485362095531587, + "grad_norm": 2.359403371810913, + "learning_rate": 2.8340312143129044e-06, + "loss": 1.3553, + "step": 1446 + }, + { + "epoch": 1.4863893168977915, + "grad_norm": 2.417792558670044, + "learning_rate": 2.8321279025504377e-06, + "loss": 1.3539, + "step": 1447 + }, + { + "epoch": 1.487416538263996, + "grad_norm": 2.7177648544311523, + "learning_rate": 2.830224590787971e-06, + "loss": 1.4575, + "step": 1448 + }, + { + "epoch": 1.4884437596302003, + "grad_norm": 2.514181613922119, + "learning_rate": 2.8283212790255045e-06, + "loss": 1.388, + "step": 1449 + }, + { + "epoch": 1.4894709809964048, + "grad_norm": 2.4293761253356934, + "learning_rate": 2.8264179672630383e-06, + "loss": 1.4143, + "step": 1450 + }, + { + "epoch": 1.4894709809964048, + "eval_loss": 1.781496286392212, + "eval_runtime": 37.8942, + "eval_samples_per_second": 26.389, + "eval_steps_per_second": 1.108, + "step": 1450 + }, + { + "epoch": 1.4904982023626092, + "grad_norm": 2.495591878890991, + "learning_rate": 2.8245146555005716e-06, + "loss": 1.3809, + "step": 1451 + }, + { + "epoch": 1.4915254237288136, + "grad_norm": 2.520158290863037, + "learning_rate": 2.8226113437381046e-06, + "loss": 1.3297, + "step": 1452 + }, + { + "epoch": 1.492552645095018, + "grad_norm": 2.403061866760254, + "learning_rate": 2.820708031975638e-06, + "loss": 1.3985, + "step": 1453 + }, + { + "epoch": 1.4935798664612223, + "grad_norm": 2.606884479522705, + "learning_rate": 2.8188047202131713e-06, + "loss": 1.4179, + "step": 1454 + }, + { + "epoch": 1.494607087827427, + "grad_norm": 2.4543404579162598, + "learning_rate": 2.8169014084507046e-06, + "loss": 1.4325, + "step": 1455 + }, + { + "epoch": 1.4956343091936313, + "grad_norm": 2.3947877883911133, + "learning_rate": 2.814998096688238e-06, + "loss": 1.3876, + "step": 1456 + }, + { + "epoch": 1.4966615305598356, + "grad_norm": 2.4632408618927, + "learning_rate": 2.813094784925771e-06, + "loss": 1.3392, + "step": 1457 + }, + { + "epoch": 1.49768875192604, + "grad_norm": 2.507312059402466, + "learning_rate": 2.8111914731633043e-06, + "loss": 1.4351, + "step": 1458 + }, + { + "epoch": 1.4987159732922444, + "grad_norm": 2.6338255405426025, + "learning_rate": 2.8092881614008377e-06, + "loss": 1.389, + "step": 1459 + }, + { + "epoch": 1.499743194658449, + "grad_norm": 2.3906662464141846, + "learning_rate": 2.807384849638371e-06, + "loss": 1.3655, + "step": 1460 + }, + { + "epoch": 1.499743194658449, + "eval_loss": 1.7716007232666016, + "eval_runtime": 37.8931, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 1460 + }, + { + "epoch": 1.5007704160246533, + "grad_norm": 2.4841926097869873, + "learning_rate": 2.8054815378759044e-06, + "loss": 1.4046, + "step": 1461 + }, + { + "epoch": 1.5017976373908577, + "grad_norm": 2.4934608936309814, + "learning_rate": 2.8035782261134378e-06, + "loss": 1.3781, + "step": 1462 + }, + { + "epoch": 1.5028248587570623, + "grad_norm": 2.538504123687744, + "learning_rate": 2.8016749143509707e-06, + "loss": 1.383, + "step": 1463 + }, + { + "epoch": 1.5038520801232664, + "grad_norm": 2.461595058441162, + "learning_rate": 2.799771602588504e-06, + "loss": 1.3892, + "step": 1464 + }, + { + "epoch": 1.504879301489471, + "grad_norm": 2.3686397075653076, + "learning_rate": 2.7978682908260374e-06, + "loss": 1.4037, + "step": 1465 + }, + { + "epoch": 1.5059065228556754, + "grad_norm": 2.47755765914917, + "learning_rate": 2.795964979063571e-06, + "loss": 1.4173, + "step": 1466 + }, + { + "epoch": 1.5069337442218798, + "grad_norm": 2.504168748855591, + "learning_rate": 2.794061667301104e-06, + "loss": 1.3955, + "step": 1467 + }, + { + "epoch": 1.5079609655880843, + "grad_norm": 2.4311411380767822, + "learning_rate": 2.792158355538637e-06, + "loss": 1.4079, + "step": 1468 + }, + { + "epoch": 1.5089881869542885, + "grad_norm": 2.523200273513794, + "learning_rate": 2.7902550437761705e-06, + "loss": 1.4019, + "step": 1469 + }, + { + "epoch": 1.510015408320493, + "grad_norm": 2.3743972778320312, + "learning_rate": 2.7883517320137042e-06, + "loss": 1.3677, + "step": 1470 + }, + { + "epoch": 1.510015408320493, + "eval_loss": 1.7622252702713013, + "eval_runtime": 37.8933, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 1470 + }, + { + "epoch": 1.5110426296866974, + "grad_norm": 2.4073667526245117, + "learning_rate": 2.7864484202512376e-06, + "loss": 1.3623, + "step": 1471 + }, + { + "epoch": 1.5120698510529018, + "grad_norm": 2.5012941360473633, + "learning_rate": 2.784545108488771e-06, + "loss": 1.382, + "step": 1472 + }, + { + "epoch": 1.5130970724191064, + "grad_norm": 2.425480365753174, + "learning_rate": 2.7826417967263043e-06, + "loss": 1.3896, + "step": 1473 + }, + { + "epoch": 1.5141242937853108, + "grad_norm": 2.4600207805633545, + "learning_rate": 2.7807384849638373e-06, + "loss": 1.4364, + "step": 1474 + }, + { + "epoch": 1.5151515151515151, + "grad_norm": 2.5489892959594727, + "learning_rate": 2.7788351732013706e-06, + "loss": 1.3832, + "step": 1475 + }, + { + "epoch": 1.5161787365177197, + "grad_norm": 2.4228315353393555, + "learning_rate": 2.776931861438904e-06, + "loss": 1.3755, + "step": 1476 + }, + { + "epoch": 1.5172059578839239, + "grad_norm": 2.519340991973877, + "learning_rate": 2.7750285496764374e-06, + "loss": 1.3915, + "step": 1477 + }, + { + "epoch": 1.5182331792501285, + "grad_norm": 2.3544020652770996, + "learning_rate": 2.7731252379139707e-06, + "loss": 1.394, + "step": 1478 + }, + { + "epoch": 1.5192604006163328, + "grad_norm": 2.4037892818450928, + "learning_rate": 2.771221926151504e-06, + "loss": 1.3396, + "step": 1479 + }, + { + "epoch": 1.5202876219825372, + "grad_norm": 2.409222364425659, + "learning_rate": 2.769318614389037e-06, + "loss": 1.4052, + "step": 1480 + }, + { + "epoch": 1.5202876219825372, + "eval_loss": 1.7622238397598267, + "eval_runtime": 37.9064, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 1480 + }, + { + "epoch": 1.5213148433487418, + "grad_norm": 2.3649749755859375, + "learning_rate": 2.7674153026265704e-06, + "loss": 1.401, + "step": 1481 + }, + { + "epoch": 1.522342064714946, + "grad_norm": 2.4484262466430664, + "learning_rate": 2.7655119908641037e-06, + "loss": 1.4026, + "step": 1482 + }, + { + "epoch": 1.5233692860811505, + "grad_norm": 2.492107629776001, + "learning_rate": 2.763608679101637e-06, + "loss": 1.3985, + "step": 1483 + }, + { + "epoch": 1.5243965074473549, + "grad_norm": 2.5192503929138184, + "learning_rate": 2.7617053673391705e-06, + "loss": 1.4282, + "step": 1484 + }, + { + "epoch": 1.5254237288135593, + "grad_norm": 2.471250534057617, + "learning_rate": 2.7598020555767034e-06, + "loss": 1.4217, + "step": 1485 + }, + { + "epoch": 1.5264509501797638, + "grad_norm": 2.2985496520996094, + "learning_rate": 2.7578987438142368e-06, + "loss": 1.3599, + "step": 1486 + }, + { + "epoch": 1.5274781715459682, + "grad_norm": 2.492047071456909, + "learning_rate": 2.75599543205177e-06, + "loss": 1.4036, + "step": 1487 + }, + { + "epoch": 1.5285053929121726, + "grad_norm": 2.5131945610046387, + "learning_rate": 2.7540921202893035e-06, + "loss": 1.4206, + "step": 1488 + }, + { + "epoch": 1.529532614278377, + "grad_norm": 2.619874954223633, + "learning_rate": 2.752188808526837e-06, + "loss": 1.4055, + "step": 1489 + }, + { + "epoch": 1.5305598356445813, + "grad_norm": 2.4526877403259277, + "learning_rate": 2.75028549676437e-06, + "loss": 1.3955, + "step": 1490 + }, + { + "epoch": 1.5305598356445813, + "eval_loss": 1.758982539176941, + "eval_runtime": 37.8756, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1490 + }, + { + "epoch": 1.531587057010786, + "grad_norm": 2.4925529956817627, + "learning_rate": 2.7483821850019036e-06, + "loss": 1.3899, + "step": 1491 + }, + { + "epoch": 1.5326142783769903, + "grad_norm": 2.3789961338043213, + "learning_rate": 2.746478873239437e-06, + "loss": 1.377, + "step": 1492 + }, + { + "epoch": 1.5336414997431946, + "grad_norm": 2.4063172340393066, + "learning_rate": 2.7445755614769703e-06, + "loss": 1.384, + "step": 1493 + }, + { + "epoch": 1.5346687211093992, + "grad_norm": 2.3189985752105713, + "learning_rate": 2.7426722497145037e-06, + "loss": 1.3635, + "step": 1494 + }, + { + "epoch": 1.5356959424756034, + "grad_norm": 2.4792535305023193, + "learning_rate": 2.740768937952037e-06, + "loss": 1.3676, + "step": 1495 + }, + { + "epoch": 1.536723163841808, + "grad_norm": 2.342355489730835, + "learning_rate": 2.7388656261895704e-06, + "loss": 1.3812, + "step": 1496 + }, + { + "epoch": 1.5377503852080123, + "grad_norm": 2.556312084197998, + "learning_rate": 2.7369623144271033e-06, + "loss": 1.4529, + "step": 1497 + }, + { + "epoch": 1.5387776065742167, + "grad_norm": 2.555600643157959, + "learning_rate": 2.7350590026646367e-06, + "loss": 1.3819, + "step": 1498 + }, + { + "epoch": 1.5398048279404213, + "grad_norm": 2.4179797172546387, + "learning_rate": 2.73315569090217e-06, + "loss": 1.3383, + "step": 1499 + }, + { + "epoch": 1.5408320493066254, + "grad_norm": 2.5075631141662598, + "learning_rate": 2.7312523791397034e-06, + "loss": 1.4217, + "step": 1500 + }, + { + "epoch": 1.5408320493066254, + "eval_loss": 1.769559621810913, + "eval_runtime": 37.8701, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 1500 + }, + { + "epoch": 1.54185927067283, + "grad_norm": 2.5675160884857178, + "learning_rate": 2.7293490673772368e-06, + "loss": 1.4437, + "step": 1501 + }, + { + "epoch": 1.5428864920390344, + "grad_norm": 2.4768013954162598, + "learning_rate": 2.7274457556147697e-06, + "loss": 1.4022, + "step": 1502 + }, + { + "epoch": 1.5439137134052388, + "grad_norm": 2.4523043632507324, + "learning_rate": 2.725542443852303e-06, + "loss": 1.385, + "step": 1503 + }, + { + "epoch": 1.5449409347714433, + "grad_norm": 2.504263401031494, + "learning_rate": 2.7236391320898365e-06, + "loss": 1.398, + "step": 1504 + }, + { + "epoch": 1.5459681561376477, + "grad_norm": 2.6203269958496094, + "learning_rate": 2.72173582032737e-06, + "loss": 1.3651, + "step": 1505 + }, + { + "epoch": 1.546995377503852, + "grad_norm": 2.3805699348449707, + "learning_rate": 2.719832508564903e-06, + "loss": 1.3861, + "step": 1506 + }, + { + "epoch": 1.5480225988700564, + "grad_norm": 2.5000486373901367, + "learning_rate": 2.717929196802436e-06, + "loss": 1.3931, + "step": 1507 + }, + { + "epoch": 1.5490498202362608, + "grad_norm": 2.5567827224731445, + "learning_rate": 2.7160258850399695e-06, + "loss": 1.3608, + "step": 1508 + }, + { + "epoch": 1.5500770416024654, + "grad_norm": 2.531614303588867, + "learning_rate": 2.714122573277503e-06, + "loss": 1.3863, + "step": 1509 + }, + { + "epoch": 1.5511042629686698, + "grad_norm": 2.419722318649292, + "learning_rate": 2.712219261515036e-06, + "loss": 1.3963, + "step": 1510 + }, + { + "epoch": 1.5511042629686698, + "eval_loss": 1.7669111490249634, + "eval_runtime": 37.6622, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 1510 + }, + { + "epoch": 1.5521314843348741, + "grad_norm": 2.4597935676574707, + "learning_rate": 2.71031594975257e-06, + "loss": 1.4044, + "step": 1511 + }, + { + "epoch": 1.5531587057010787, + "grad_norm": 2.386899471282959, + "learning_rate": 2.7084126379901034e-06, + "loss": 1.4092, + "step": 1512 + }, + { + "epoch": 1.5541859270672829, + "grad_norm": 2.4398958683013916, + "learning_rate": 2.7065093262276367e-06, + "loss": 1.4185, + "step": 1513 + }, + { + "epoch": 1.5552131484334875, + "grad_norm": 2.319274663925171, + "learning_rate": 2.7046060144651697e-06, + "loss": 1.4316, + "step": 1514 + }, + { + "epoch": 1.5562403697996918, + "grad_norm": 2.4519565105438232, + "learning_rate": 2.702702702702703e-06, + "loss": 1.4387, + "step": 1515 + }, + { + "epoch": 1.5572675911658962, + "grad_norm": 2.2890849113464355, + "learning_rate": 2.7007993909402364e-06, + "loss": 1.3167, + "step": 1516 + }, + { + "epoch": 1.5582948125321008, + "grad_norm": 2.6062512397766113, + "learning_rate": 2.6988960791777697e-06, + "loss": 1.4332, + "step": 1517 + }, + { + "epoch": 1.559322033898305, + "grad_norm": 2.508957862854004, + "learning_rate": 2.696992767415303e-06, + "loss": 1.3995, + "step": 1518 + }, + { + "epoch": 1.5603492552645095, + "grad_norm": 2.4134879112243652, + "learning_rate": 2.695089455652836e-06, + "loss": 1.3793, + "step": 1519 + }, + { + "epoch": 1.5613764766307139, + "grad_norm": 2.498729705810547, + "learning_rate": 2.6931861438903694e-06, + "loss": 1.3564, + "step": 1520 + }, + { + "epoch": 1.5613764766307139, + "eval_loss": 1.76124107837677, + "eval_runtime": 37.831, + "eval_samples_per_second": 26.433, + "eval_steps_per_second": 1.11, + "step": 1520 + }, + { + "epoch": 1.5624036979969183, + "grad_norm": 2.455510139465332, + "learning_rate": 2.6912828321279028e-06, + "loss": 1.4109, + "step": 1521 + }, + { + "epoch": 1.5634309193631228, + "grad_norm": 2.422098159790039, + "learning_rate": 2.689379520365436e-06, + "loss": 1.3811, + "step": 1522 + }, + { + "epoch": 1.5644581407293272, + "grad_norm": 2.380890369415283, + "learning_rate": 2.6874762086029695e-06, + "loss": 1.4215, + "step": 1523 + }, + { + "epoch": 1.5654853620955316, + "grad_norm": 2.39363169670105, + "learning_rate": 2.6855728968405024e-06, + "loss": 1.3708, + "step": 1524 + }, + { + "epoch": 1.5665125834617362, + "grad_norm": 2.3624465465545654, + "learning_rate": 2.683669585078036e-06, + "loss": 1.4084, + "step": 1525 + }, + { + "epoch": 1.5675398048279403, + "grad_norm": 2.3589534759521484, + "learning_rate": 2.681766273315569e-06, + "loss": 1.3465, + "step": 1526 + }, + { + "epoch": 1.568567026194145, + "grad_norm": 2.4358372688293457, + "learning_rate": 2.6798629615531025e-06, + "loss": 1.3915, + "step": 1527 + }, + { + "epoch": 1.5695942475603493, + "grad_norm": 2.4162042140960693, + "learning_rate": 2.677959649790636e-06, + "loss": 1.3842, + "step": 1528 + }, + { + "epoch": 1.5706214689265536, + "grad_norm": 2.568127155303955, + "learning_rate": 2.676056338028169e-06, + "loss": 1.4197, + "step": 1529 + }, + { + "epoch": 1.5716486902927582, + "grad_norm": 2.5098628997802734, + "learning_rate": 2.674153026265702e-06, + "loss": 1.4132, + "step": 1530 + }, + { + "epoch": 1.5716486902927582, + "eval_loss": 1.7691000699996948, + "eval_runtime": 37.9709, + "eval_samples_per_second": 26.336, + "eval_steps_per_second": 1.106, + "step": 1530 + }, + { + "epoch": 1.5726759116589624, + "grad_norm": 2.5882396697998047, + "learning_rate": 2.672249714503236e-06, + "loss": 1.4223, + "step": 1531 + }, + { + "epoch": 1.573703133025167, + "grad_norm": 2.5202438831329346, + "learning_rate": 2.6703464027407693e-06, + "loss": 1.3804, + "step": 1532 + }, + { + "epoch": 1.5747303543913713, + "grad_norm": 2.4582319259643555, + "learning_rate": 2.6684430909783027e-06, + "loss": 1.3987, + "step": 1533 + }, + { + "epoch": 1.5757575757575757, + "grad_norm": 2.3516597747802734, + "learning_rate": 2.666539779215836e-06, + "loss": 1.3878, + "step": 1534 + }, + { + "epoch": 1.5767847971237803, + "grad_norm": 2.410268783569336, + "learning_rate": 2.6646364674533694e-06, + "loss": 1.3737, + "step": 1535 + }, + { + "epoch": 1.5778120184899846, + "grad_norm": 2.2420742511749268, + "learning_rate": 2.6627331556909024e-06, + "loss": 1.2908, + "step": 1536 + }, + { + "epoch": 1.578839239856189, + "grad_norm": 2.498166084289551, + "learning_rate": 2.6608298439284357e-06, + "loss": 1.383, + "step": 1537 + }, + { + "epoch": 1.5798664612223934, + "grad_norm": 2.389152765274048, + "learning_rate": 2.658926532165969e-06, + "loss": 1.3829, + "step": 1538 + }, + { + "epoch": 1.5808936825885977, + "grad_norm": 2.4197394847869873, + "learning_rate": 2.6570232204035024e-06, + "loss": 1.3438, + "step": 1539 + }, + { + "epoch": 1.5819209039548023, + "grad_norm": 2.321922540664673, + "learning_rate": 2.655119908641036e-06, + "loss": 1.3741, + "step": 1540 + }, + { + "epoch": 1.5819209039548023, + "eval_loss": 1.7643449306488037, + "eval_runtime": 38.0047, + "eval_samples_per_second": 26.313, + "eval_steps_per_second": 1.105, + "step": 1540 + }, + { + "epoch": 1.5829481253210067, + "grad_norm": 2.4056994915008545, + "learning_rate": 2.6532165968785687e-06, + "loss": 1.3857, + "step": 1541 + }, + { + "epoch": 1.583975346687211, + "grad_norm": 2.2752463817596436, + "learning_rate": 2.651313285116102e-06, + "loss": 1.3517, + "step": 1542 + }, + { + "epoch": 1.5850025680534157, + "grad_norm": 2.411092519760132, + "learning_rate": 2.6494099733536355e-06, + "loss": 1.3548, + "step": 1543 + }, + { + "epoch": 1.5860297894196198, + "grad_norm": 2.7468392848968506, + "learning_rate": 2.647506661591169e-06, + "loss": 1.5273, + "step": 1544 + }, + { + "epoch": 1.5870570107858244, + "grad_norm": 2.5081777572631836, + "learning_rate": 2.645603349828702e-06, + "loss": 1.4064, + "step": 1545 + }, + { + "epoch": 1.5880842321520288, + "grad_norm": 2.3444366455078125, + "learning_rate": 2.643700038066235e-06, + "loss": 1.3444, + "step": 1546 + }, + { + "epoch": 1.5891114535182331, + "grad_norm": 2.437304973602295, + "learning_rate": 2.6417967263037685e-06, + "loss": 1.4078, + "step": 1547 + }, + { + "epoch": 1.5901386748844377, + "grad_norm": 2.456725597381592, + "learning_rate": 2.639893414541302e-06, + "loss": 1.4212, + "step": 1548 + }, + { + "epoch": 1.5911658962506419, + "grad_norm": 2.4959378242492676, + "learning_rate": 2.6379901027788352e-06, + "loss": 1.4439, + "step": 1549 + }, + { + "epoch": 1.5921931176168465, + "grad_norm": 2.3856401443481445, + "learning_rate": 2.6360867910163686e-06, + "loss": 1.3748, + "step": 1550 + }, + { + "epoch": 1.5921931176168465, + "eval_loss": 1.7585556507110596, + "eval_runtime": 38.038, + "eval_samples_per_second": 26.289, + "eval_steps_per_second": 1.104, + "step": 1550 + }, + { + "epoch": 1.5932203389830508, + "grad_norm": 2.3285882472991943, + "learning_rate": 2.6341834792539024e-06, + "loss": 1.3912, + "step": 1551 + }, + { + "epoch": 1.5942475603492552, + "grad_norm": 2.365633249282837, + "learning_rate": 2.6322801674914357e-06, + "loss": 1.3922, + "step": 1552 + }, + { + "epoch": 1.5952747817154598, + "grad_norm": 2.460118293762207, + "learning_rate": 2.6303768557289687e-06, + "loss": 1.3899, + "step": 1553 + }, + { + "epoch": 1.5963020030816641, + "grad_norm": 2.244420051574707, + "learning_rate": 2.628473543966502e-06, + "loss": 1.347, + "step": 1554 + }, + { + "epoch": 1.5973292244478685, + "grad_norm": 2.404733180999756, + "learning_rate": 2.6265702322040354e-06, + "loss": 1.4163, + "step": 1555 + }, + { + "epoch": 1.5983564458140729, + "grad_norm": 2.3252789974212646, + "learning_rate": 2.6246669204415688e-06, + "loss": 1.3578, + "step": 1556 + }, + { + "epoch": 1.5993836671802772, + "grad_norm": 2.3664159774780273, + "learning_rate": 2.622763608679102e-06, + "loss": 1.3585, + "step": 1557 + }, + { + "epoch": 1.6004108885464818, + "grad_norm": 2.3188118934631348, + "learning_rate": 2.620860296916635e-06, + "loss": 1.3794, + "step": 1558 + }, + { + "epoch": 1.6014381099126862, + "grad_norm": 2.3163065910339355, + "learning_rate": 2.6189569851541684e-06, + "loss": 1.3157, + "step": 1559 + }, + { + "epoch": 1.6024653312788906, + "grad_norm": 2.4271090030670166, + "learning_rate": 2.617053673391702e-06, + "loss": 1.413, + "step": 1560 + }, + { + "epoch": 1.6024653312788906, + "eval_loss": 1.7588003873825073, + "eval_runtime": 38.0406, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 1560 + }, + { + "epoch": 1.6034925526450952, + "grad_norm": 2.5792505741119385, + "learning_rate": 2.615150361629235e-06, + "loss": 1.4431, + "step": 1561 + }, + { + "epoch": 1.6045197740112993, + "grad_norm": 2.4612159729003906, + "learning_rate": 2.6132470498667685e-06, + "loss": 1.3967, + "step": 1562 + }, + { + "epoch": 1.605546995377504, + "grad_norm": 2.4364020824432373, + "learning_rate": 2.6113437381043015e-06, + "loss": 1.3779, + "step": 1563 + }, + { + "epoch": 1.6065742167437083, + "grad_norm": 2.464442253112793, + "learning_rate": 2.609440426341835e-06, + "loss": 1.3685, + "step": 1564 + }, + { + "epoch": 1.6076014381099126, + "grad_norm": 2.4871442317962646, + "learning_rate": 2.607537114579368e-06, + "loss": 1.3979, + "step": 1565 + }, + { + "epoch": 1.6086286594761172, + "grad_norm": 2.375364303588867, + "learning_rate": 2.6056338028169015e-06, + "loss": 1.3836, + "step": 1566 + }, + { + "epoch": 1.6096558808423214, + "grad_norm": 2.364389657974243, + "learning_rate": 2.603730491054435e-06, + "loss": 1.3655, + "step": 1567 + }, + { + "epoch": 1.610683102208526, + "grad_norm": 2.421839714050293, + "learning_rate": 2.601827179291968e-06, + "loss": 1.398, + "step": 1568 + }, + { + "epoch": 1.6117103235747303, + "grad_norm": 2.3270041942596436, + "learning_rate": 2.599923867529501e-06, + "loss": 1.4061, + "step": 1569 + }, + { + "epoch": 1.6127375449409347, + "grad_norm": 2.4831383228302, + "learning_rate": 2.5980205557670346e-06, + "loss": 1.3808, + "step": 1570 + }, + { + "epoch": 1.6127375449409347, + "eval_loss": 1.773654818534851, + "eval_runtime": 38.046, + "eval_samples_per_second": 26.284, + "eval_steps_per_second": 1.104, + "step": 1570 + }, + { + "epoch": 1.6137647663071393, + "grad_norm": 2.499079942703247, + "learning_rate": 2.5961172440045684e-06, + "loss": 1.4167, + "step": 1571 + }, + { + "epoch": 1.6147919876733436, + "grad_norm": 2.3598642349243164, + "learning_rate": 2.5942139322421017e-06, + "loss": 1.4319, + "step": 1572 + }, + { + "epoch": 1.615819209039548, + "grad_norm": 2.3554258346557617, + "learning_rate": 2.592310620479635e-06, + "loss": 1.3723, + "step": 1573 + }, + { + "epoch": 1.6168464304057526, + "grad_norm": 2.2909231185913086, + "learning_rate": 2.5904073087171684e-06, + "loss": 1.339, + "step": 1574 + }, + { + "epoch": 1.6178736517719567, + "grad_norm": 2.3029019832611084, + "learning_rate": 2.5885039969547014e-06, + "loss": 1.4026, + "step": 1575 + }, + { + "epoch": 1.6189008731381613, + "grad_norm": 2.2370176315307617, + "learning_rate": 2.5866006851922347e-06, + "loss": 1.3531, + "step": 1576 + }, + { + "epoch": 1.6199280945043657, + "grad_norm": 2.2932159900665283, + "learning_rate": 2.584697373429768e-06, + "loss": 1.3585, + "step": 1577 + }, + { + "epoch": 1.62095531587057, + "grad_norm": 2.391493797302246, + "learning_rate": 2.5827940616673015e-06, + "loss": 1.3817, + "step": 1578 + }, + { + "epoch": 1.6219825372367747, + "grad_norm": 2.476572036743164, + "learning_rate": 2.580890749904835e-06, + "loss": 1.4241, + "step": 1579 + }, + { + "epoch": 1.6230097586029788, + "grad_norm": 2.3396525382995605, + "learning_rate": 2.5789874381423678e-06, + "loss": 1.3734, + "step": 1580 + }, + { + "epoch": 1.6230097586029788, + "eval_loss": 1.7641068696975708, + "eval_runtime": 38.0413, + "eval_samples_per_second": 26.287, + "eval_steps_per_second": 1.104, + "step": 1580 + }, + { + "epoch": 1.6240369799691834, + "grad_norm": 2.381436824798584, + "learning_rate": 2.577084126379901e-06, + "loss": 1.3764, + "step": 1581 + }, + { + "epoch": 1.6250642013353878, + "grad_norm": 2.391737937927246, + "learning_rate": 2.5751808146174345e-06, + "loss": 1.3419, + "step": 1582 + }, + { + "epoch": 1.6260914227015921, + "grad_norm": 2.453653573989868, + "learning_rate": 2.573277502854968e-06, + "loss": 1.398, + "step": 1583 + }, + { + "epoch": 1.6271186440677967, + "grad_norm": 2.3906450271606445, + "learning_rate": 2.5713741910925012e-06, + "loss": 1.3605, + "step": 1584 + }, + { + "epoch": 1.6281458654340009, + "grad_norm": 2.347041606903076, + "learning_rate": 2.569470879330034e-06, + "loss": 1.369, + "step": 1585 + }, + { + "epoch": 1.6291730868002054, + "grad_norm": 2.571047306060791, + "learning_rate": 2.5675675675675675e-06, + "loss": 1.421, + "step": 1586 + }, + { + "epoch": 1.6302003081664098, + "grad_norm": 2.3640646934509277, + "learning_rate": 2.565664255805101e-06, + "loss": 1.3736, + "step": 1587 + }, + { + "epoch": 1.6312275295326142, + "grad_norm": 2.419844388961792, + "learning_rate": 2.5637609440426343e-06, + "loss": 1.4008, + "step": 1588 + }, + { + "epoch": 1.6322547508988188, + "grad_norm": 2.3414361476898193, + "learning_rate": 2.5618576322801676e-06, + "loss": 1.3473, + "step": 1589 + }, + { + "epoch": 1.6332819722650231, + "grad_norm": 2.462259292602539, + "learning_rate": 2.5599543205177006e-06, + "loss": 1.4418, + "step": 1590 + }, + { + "epoch": 1.6332819722650231, + "eval_loss": 1.760501503944397, + "eval_runtime": 38.0428, + "eval_samples_per_second": 26.286, + "eval_steps_per_second": 1.104, + "step": 1590 + }, + { + "epoch": 1.6343091936312275, + "grad_norm": 2.49070405960083, + "learning_rate": 2.558051008755234e-06, + "loss": 1.3859, + "step": 1591 + }, + { + "epoch": 1.635336414997432, + "grad_norm": 2.500180244445801, + "learning_rate": 2.5561476969927677e-06, + "loss": 1.4379, + "step": 1592 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 2.4660637378692627, + "learning_rate": 2.554244385230301e-06, + "loss": 1.3656, + "step": 1593 + }, + { + "epoch": 1.6373908577298408, + "grad_norm": 2.5678634643554688, + "learning_rate": 2.5523410734678344e-06, + "loss": 1.3939, + "step": 1594 + }, + { + "epoch": 1.6384180790960452, + "grad_norm": 2.358930826187134, + "learning_rate": 2.550437761705368e-06, + "loss": 1.3281, + "step": 1595 + }, + { + "epoch": 1.6394453004622496, + "grad_norm": 2.462233543395996, + "learning_rate": 2.548534449942901e-06, + "loss": 1.3643, + "step": 1596 + }, + { + "epoch": 1.6404725218284542, + "grad_norm": 2.5294501781463623, + "learning_rate": 2.546631138180434e-06, + "loss": 1.4009, + "step": 1597 + }, + { + "epoch": 1.6414997431946583, + "grad_norm": 2.5325710773468018, + "learning_rate": 2.5447278264179675e-06, + "loss": 1.3797, + "step": 1598 + }, + { + "epoch": 1.642526964560863, + "grad_norm": 2.3293585777282715, + "learning_rate": 2.542824514655501e-06, + "loss": 1.3492, + "step": 1599 + }, + { + "epoch": 1.6435541859270673, + "grad_norm": 2.5357868671417236, + "learning_rate": 2.540921202893034e-06, + "loss": 1.4295, + "step": 1600 + }, + { + "epoch": 1.6435541859270673, + "eval_loss": 1.7499933242797852, + "eval_runtime": 38.0219, + "eval_samples_per_second": 26.301, + "eval_steps_per_second": 1.105, + "step": 1600 + }, + { + "epoch": 1.6445814072932716, + "grad_norm": 2.509333848953247, + "learning_rate": 2.5390178911305675e-06, + "loss": 1.4174, + "step": 1601 + }, + { + "epoch": 1.6456086286594762, + "grad_norm": 2.4453277587890625, + "learning_rate": 2.5371145793681005e-06, + "loss": 1.3898, + "step": 1602 + }, + { + "epoch": 1.6466358500256806, + "grad_norm": 2.506847381591797, + "learning_rate": 2.535211267605634e-06, + "loss": 1.3881, + "step": 1603 + }, + { + "epoch": 1.647663071391885, + "grad_norm": 2.2735178470611572, + "learning_rate": 2.533307955843167e-06, + "loss": 1.3268, + "step": 1604 + }, + { + "epoch": 1.6486902927580893, + "grad_norm": 2.479081630706787, + "learning_rate": 2.5314046440807006e-06, + "loss": 1.3897, + "step": 1605 + }, + { + "epoch": 1.6497175141242937, + "grad_norm": 2.3882839679718018, + "learning_rate": 2.529501332318234e-06, + "loss": 1.3734, + "step": 1606 + }, + { + "epoch": 1.6507447354904983, + "grad_norm": 2.521341323852539, + "learning_rate": 2.5275980205557673e-06, + "loss": 1.4298, + "step": 1607 + }, + { + "epoch": 1.6517719568567026, + "grad_norm": 2.55405855178833, + "learning_rate": 2.5256947087933002e-06, + "loss": 1.4497, + "step": 1608 + }, + { + "epoch": 1.652799178222907, + "grad_norm": 2.534478187561035, + "learning_rate": 2.5237913970308336e-06, + "loss": 1.438, + "step": 1609 + }, + { + "epoch": 1.6538263995891116, + "grad_norm": 2.453374147415161, + "learning_rate": 2.521888085268367e-06, + "loss": 1.3834, + "step": 1610 + }, + { + "epoch": 1.6538263995891116, + "eval_loss": 1.7550541162490845, + "eval_runtime": 37.6671, + "eval_samples_per_second": 26.548, + "eval_steps_per_second": 1.115, + "step": 1610 + }, + { + "epoch": 1.6548536209553157, + "grad_norm": 2.4754152297973633, + "learning_rate": 2.5199847735059003e-06, + "loss": 1.434, + "step": 1611 + }, + { + "epoch": 1.6558808423215203, + "grad_norm": 2.3785831928253174, + "learning_rate": 2.518081461743434e-06, + "loss": 1.3706, + "step": 1612 + }, + { + "epoch": 1.6569080636877247, + "grad_norm": 2.3936359882354736, + "learning_rate": 2.5161781499809675e-06, + "loss": 1.3955, + "step": 1613 + }, + { + "epoch": 1.657935285053929, + "grad_norm": 2.338012456893921, + "learning_rate": 2.5142748382185004e-06, + "loss": 1.3532, + "step": 1614 + }, + { + "epoch": 1.6589625064201337, + "grad_norm": 2.5027408599853516, + "learning_rate": 2.5123715264560338e-06, + "loss": 1.3522, + "step": 1615 + }, + { + "epoch": 1.6599897277863378, + "grad_norm": 2.4059648513793945, + "learning_rate": 2.510468214693567e-06, + "loss": 1.3767, + "step": 1616 + }, + { + "epoch": 1.6610169491525424, + "grad_norm": 2.324235200881958, + "learning_rate": 2.5085649029311005e-06, + "loss": 1.3248, + "step": 1617 + }, + { + "epoch": 1.6620441705187468, + "grad_norm": 2.336317300796509, + "learning_rate": 2.506661591168634e-06, + "loss": 1.288, + "step": 1618 + }, + { + "epoch": 1.6630713918849511, + "grad_norm": 2.3386120796203613, + "learning_rate": 2.504758279406167e-06, + "loss": 1.3715, + "step": 1619 + }, + { + "epoch": 1.6640986132511557, + "grad_norm": 2.5584142208099365, + "learning_rate": 2.5028549676437e-06, + "loss": 1.4157, + "step": 1620 + }, + { + "epoch": 1.6640986132511557, + "eval_loss": 1.7575956583023071, + "eval_runtime": 37.7897, + "eval_samples_per_second": 26.462, + "eval_steps_per_second": 1.111, + "step": 1620 + }, + { + "epoch": 1.66512583461736, + "grad_norm": 2.475881338119507, + "learning_rate": 2.5009516558812335e-06, + "loss": 1.357, + "step": 1621 + }, + { + "epoch": 1.6661530559835644, + "grad_norm": 2.4611690044403076, + "learning_rate": 2.499048344118767e-06, + "loss": 1.4077, + "step": 1622 + }, + { + "epoch": 1.667180277349769, + "grad_norm": 2.4230902194976807, + "learning_rate": 2.4971450323563002e-06, + "loss": 1.443, + "step": 1623 + }, + { + "epoch": 1.6682074987159732, + "grad_norm": 2.359095335006714, + "learning_rate": 2.4952417205938336e-06, + "loss": 1.3595, + "step": 1624 + }, + { + "epoch": 1.6692347200821778, + "grad_norm": 2.4128329753875732, + "learning_rate": 2.4933384088313666e-06, + "loss": 1.3636, + "step": 1625 + }, + { + "epoch": 1.6702619414483821, + "grad_norm": 2.3391597270965576, + "learning_rate": 2.4914350970689e-06, + "loss": 1.3536, + "step": 1626 + }, + { + "epoch": 1.6712891628145865, + "grad_norm": 2.478581190109253, + "learning_rate": 2.4895317853064337e-06, + "loss": 1.3642, + "step": 1627 + }, + { + "epoch": 1.672316384180791, + "grad_norm": 2.7438251972198486, + "learning_rate": 2.4876284735439666e-06, + "loss": 1.4192, + "step": 1628 + }, + { + "epoch": 1.6733436055469952, + "grad_norm": 2.508089542388916, + "learning_rate": 2.4857251617815e-06, + "loss": 1.3879, + "step": 1629 + }, + { + "epoch": 1.6743708269131998, + "grad_norm": 2.5251338481903076, + "learning_rate": 2.4838218500190334e-06, + "loss": 1.3936, + "step": 1630 + }, + { + "epoch": 1.6743708269131998, + "eval_loss": 1.7578516006469727, + "eval_runtime": 37.8753, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1630 + }, + { + "epoch": 1.6753980482794042, + "grad_norm": 2.3655474185943604, + "learning_rate": 2.4819185382565667e-06, + "loss": 1.3839, + "step": 1631 + }, + { + "epoch": 1.6764252696456086, + "grad_norm": 2.4703869819641113, + "learning_rate": 2.4800152264941e-06, + "loss": 1.3767, + "step": 1632 + }, + { + "epoch": 1.6774524910118132, + "grad_norm": 2.4739503860473633, + "learning_rate": 2.478111914731633e-06, + "loss": 1.3465, + "step": 1633 + }, + { + "epoch": 1.6784797123780173, + "grad_norm": 2.356006383895874, + "learning_rate": 2.4762086029691664e-06, + "loss": 1.3828, + "step": 1634 + }, + { + "epoch": 1.6795069337442219, + "grad_norm": 2.7019715309143066, + "learning_rate": 2.4743052912066998e-06, + "loss": 1.4875, + "step": 1635 + }, + { + "epoch": 1.6805341551104263, + "grad_norm": 2.3504486083984375, + "learning_rate": 2.472401979444233e-06, + "loss": 1.3494, + "step": 1636 + }, + { + "epoch": 1.6815613764766306, + "grad_norm": 2.6056807041168213, + "learning_rate": 2.4704986676817665e-06, + "loss": 1.4399, + "step": 1637 + }, + { + "epoch": 1.6825885978428352, + "grad_norm": 2.481422185897827, + "learning_rate": 2.4685953559193e-06, + "loss": 1.3803, + "step": 1638 + }, + { + "epoch": 1.6836158192090396, + "grad_norm": 2.483227491378784, + "learning_rate": 2.466692044156833e-06, + "loss": 1.4255, + "step": 1639 + }, + { + "epoch": 1.684643040575244, + "grad_norm": 2.4171221256256104, + "learning_rate": 2.4647887323943666e-06, + "loss": 1.3541, + "step": 1640 + }, + { + "epoch": 1.684643040575244, + "eval_loss": 1.7658556699752808, + "eval_runtime": 37.8673, + "eval_samples_per_second": 26.408, + "eval_steps_per_second": 1.109, + "step": 1640 + }, + { + "epoch": 1.6856702619414485, + "grad_norm": 2.2551865577697754, + "learning_rate": 2.4628854206319e-06, + "loss": 1.334, + "step": 1641 + }, + { + "epoch": 1.6866974833076527, + "grad_norm": 2.45120906829834, + "learning_rate": 2.460982108869433e-06, + "loss": 1.3416, + "step": 1642 + }, + { + "epoch": 1.6877247046738573, + "grad_norm": 2.4356141090393066, + "learning_rate": 2.4590787971069662e-06, + "loss": 1.4222, + "step": 1643 + }, + { + "epoch": 1.6887519260400616, + "grad_norm": 2.3313231468200684, + "learning_rate": 2.4571754853444996e-06, + "loss": 1.3654, + "step": 1644 + }, + { + "epoch": 1.689779147406266, + "grad_norm": 2.3518543243408203, + "learning_rate": 2.455272173582033e-06, + "loss": 1.3521, + "step": 1645 + }, + { + "epoch": 1.6908063687724706, + "grad_norm": 2.477158546447754, + "learning_rate": 2.4533688618195663e-06, + "loss": 1.3665, + "step": 1646 + }, + { + "epoch": 1.6918335901386747, + "grad_norm": 2.2717649936676025, + "learning_rate": 2.4514655500570997e-06, + "loss": 1.2992, + "step": 1647 + }, + { + "epoch": 1.6928608115048793, + "grad_norm": 2.2843265533447266, + "learning_rate": 2.449562238294633e-06, + "loss": 1.3458, + "step": 1648 + }, + { + "epoch": 1.6938880328710837, + "grad_norm": 2.4243931770324707, + "learning_rate": 2.4476589265321664e-06, + "loss": 1.3143, + "step": 1649 + }, + { + "epoch": 1.694915254237288, + "grad_norm": 2.4439947605133057, + "learning_rate": 2.4457556147696993e-06, + "loss": 1.41, + "step": 1650 + }, + { + "epoch": 1.694915254237288, + "eval_loss": 1.762351155281067, + "eval_runtime": 37.8973, + "eval_samples_per_second": 26.387, + "eval_steps_per_second": 1.108, + "step": 1650 + }, + { + "epoch": 1.6959424756034926, + "grad_norm": 2.3726727962493896, + "learning_rate": 2.4438523030072327e-06, + "loss": 1.388, + "step": 1651 + }, + { + "epoch": 1.696969696969697, + "grad_norm": 2.4449899196624756, + "learning_rate": 2.441948991244766e-06, + "loss": 1.4023, + "step": 1652 + }, + { + "epoch": 1.6979969183359014, + "grad_norm": 2.4462502002716064, + "learning_rate": 2.4400456794822994e-06, + "loss": 1.3761, + "step": 1653 + }, + { + "epoch": 1.6990241397021058, + "grad_norm": 2.413308620452881, + "learning_rate": 2.438142367719833e-06, + "loss": 1.4017, + "step": 1654 + }, + { + "epoch": 1.7000513610683101, + "grad_norm": 2.398103952407837, + "learning_rate": 2.4362390559573657e-06, + "loss": 1.3709, + "step": 1655 + }, + { + "epoch": 1.7010785824345147, + "grad_norm": 2.4637506008148193, + "learning_rate": 2.434335744194899e-06, + "loss": 1.3657, + "step": 1656 + }, + { + "epoch": 1.702105803800719, + "grad_norm": 2.419947385787964, + "learning_rate": 2.432432432432433e-06, + "loss": 1.4071, + "step": 1657 + }, + { + "epoch": 1.7031330251669234, + "grad_norm": 2.4477648735046387, + "learning_rate": 2.4305291206699662e-06, + "loss": 1.3288, + "step": 1658 + }, + { + "epoch": 1.704160246533128, + "grad_norm": 2.3359293937683105, + "learning_rate": 2.428625808907499e-06, + "loss": 1.396, + "step": 1659 + }, + { + "epoch": 1.7051874678993322, + "grad_norm": 2.396193742752075, + "learning_rate": 2.4267224971450325e-06, + "loss": 1.4246, + "step": 1660 + }, + { + "epoch": 1.7051874678993322, + "eval_loss": 1.7526435852050781, + "eval_runtime": 37.9178, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 1660 + }, + { + "epoch": 1.7062146892655368, + "grad_norm": 2.511453628540039, + "learning_rate": 2.424819185382566e-06, + "loss": 1.3877, + "step": 1661 + }, + { + "epoch": 1.7072419106317411, + "grad_norm": 2.4472320079803467, + "learning_rate": 2.4229158736200993e-06, + "loss": 1.3346, + "step": 1662 + }, + { + "epoch": 1.7082691319979455, + "grad_norm": 2.2219882011413574, + "learning_rate": 2.4210125618576326e-06, + "loss": 1.3352, + "step": 1663 + }, + { + "epoch": 1.70929635336415, + "grad_norm": 2.440121650695801, + "learning_rate": 2.4191092500951656e-06, + "loss": 1.3644, + "step": 1664 + }, + { + "epoch": 1.7103235747303542, + "grad_norm": 2.5501232147216797, + "learning_rate": 2.417205938332699e-06, + "loss": 1.4073, + "step": 1665 + }, + { + "epoch": 1.7113507960965588, + "grad_norm": 2.4315030574798584, + "learning_rate": 2.4153026265702323e-06, + "loss": 1.3137, + "step": 1666 + }, + { + "epoch": 1.7123780174627632, + "grad_norm": 2.3719727993011475, + "learning_rate": 2.4133993148077657e-06, + "loss": 1.3484, + "step": 1667 + }, + { + "epoch": 1.7134052388289676, + "grad_norm": 2.3800482749938965, + "learning_rate": 2.411496003045299e-06, + "loss": 1.4106, + "step": 1668 + }, + { + "epoch": 1.7144324601951721, + "grad_norm": 2.4554190635681152, + "learning_rate": 2.4095926912828324e-06, + "loss": 1.3603, + "step": 1669 + }, + { + "epoch": 1.7154596815613765, + "grad_norm": 2.2984702587127686, + "learning_rate": 2.4076893795203658e-06, + "loss": 1.3739, + "step": 1670 + }, + { + "epoch": 1.7154596815613765, + "eval_loss": 1.752260446548462, + "eval_runtime": 37.9208, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 1670 + }, + { + "epoch": 1.7164869029275809, + "grad_norm": 2.3967223167419434, + "learning_rate": 2.405786067757899e-06, + "loss": 1.3598, + "step": 1671 + }, + { + "epoch": 1.7175141242937855, + "grad_norm": 2.585935592651367, + "learning_rate": 2.403882755995432e-06, + "loss": 1.4489, + "step": 1672 + }, + { + "epoch": 1.7185413456599896, + "grad_norm": 2.4227662086486816, + "learning_rate": 2.4019794442329654e-06, + "loss": 1.3745, + "step": 1673 + }, + { + "epoch": 1.7195685670261942, + "grad_norm": 2.485344648361206, + "learning_rate": 2.4000761324704988e-06, + "loss": 1.3405, + "step": 1674 + }, + { + "epoch": 1.7205957883923986, + "grad_norm": 2.4055545330047607, + "learning_rate": 2.398172820708032e-06, + "loss": 1.3914, + "step": 1675 + }, + { + "epoch": 1.721623009758603, + "grad_norm": 2.3449783325195312, + "learning_rate": 2.3962695089455655e-06, + "loss": 1.3637, + "step": 1676 + }, + { + "epoch": 1.7226502311248075, + "grad_norm": 2.338590621948242, + "learning_rate": 2.3943661971830984e-06, + "loss": 1.3571, + "step": 1677 + }, + { + "epoch": 1.7236774524910117, + "grad_norm": 2.3274776935577393, + "learning_rate": 2.3924628854206322e-06, + "loss": 1.3632, + "step": 1678 + }, + { + "epoch": 1.7247046738572163, + "grad_norm": 2.3467724323272705, + "learning_rate": 2.3905595736581656e-06, + "loss": 1.3568, + "step": 1679 + }, + { + "epoch": 1.7257318952234206, + "grad_norm": 2.552039861679077, + "learning_rate": 2.388656261895699e-06, + "loss": 1.4099, + "step": 1680 + }, + { + "epoch": 1.7257318952234206, + "eval_loss": 1.745519995689392, + "eval_runtime": 37.9527, + "eval_samples_per_second": 26.349, + "eval_steps_per_second": 1.107, + "step": 1680 + }, + { + "epoch": 1.726759116589625, + "grad_norm": 2.5332300662994385, + "learning_rate": 2.386752950133232e-06, + "loss": 1.4017, + "step": 1681 + }, + { + "epoch": 1.7277863379558296, + "grad_norm": 2.4033122062683105, + "learning_rate": 2.3848496383707653e-06, + "loss": 1.3866, + "step": 1682 + }, + { + "epoch": 1.7288135593220337, + "grad_norm": 2.460071086883545, + "learning_rate": 2.3829463266082986e-06, + "loss": 1.3824, + "step": 1683 + }, + { + "epoch": 1.7298407806882383, + "grad_norm": 2.5089826583862305, + "learning_rate": 2.381043014845832e-06, + "loss": 1.398, + "step": 1684 + }, + { + "epoch": 1.7308680020544427, + "grad_norm": 2.419689893722534, + "learning_rate": 2.3791397030833653e-06, + "loss": 1.3852, + "step": 1685 + }, + { + "epoch": 1.731895223420647, + "grad_norm": 2.314080238342285, + "learning_rate": 2.3772363913208983e-06, + "loss": 1.3666, + "step": 1686 + }, + { + "epoch": 1.7329224447868516, + "grad_norm": 2.324474334716797, + "learning_rate": 2.3753330795584316e-06, + "loss": 1.3495, + "step": 1687 + }, + { + "epoch": 1.733949666153056, + "grad_norm": 2.5079402923583984, + "learning_rate": 2.3734297677959654e-06, + "loss": 1.3731, + "step": 1688 + }, + { + "epoch": 1.7349768875192604, + "grad_norm": 2.492154598236084, + "learning_rate": 2.3715264560334984e-06, + "loss": 1.413, + "step": 1689 + }, + { + "epoch": 1.736004108885465, + "grad_norm": 2.3400940895080566, + "learning_rate": 2.3696231442710317e-06, + "loss": 1.3952, + "step": 1690 + }, + { + "epoch": 1.736004108885465, + "eval_loss": 1.7412214279174805, + "eval_runtime": 37.9159, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 1690 + }, + { + "epoch": 1.7370313302516691, + "grad_norm": 2.3744688034057617, + "learning_rate": 2.367719832508565e-06, + "loss": 1.396, + "step": 1691 + }, + { + "epoch": 1.7380585516178737, + "grad_norm": 2.4762585163116455, + "learning_rate": 2.3658165207460985e-06, + "loss": 1.4113, + "step": 1692 + }, + { + "epoch": 1.739085772984078, + "grad_norm": 2.361391305923462, + "learning_rate": 2.363913208983632e-06, + "loss": 1.3962, + "step": 1693 + }, + { + "epoch": 1.7401129943502824, + "grad_norm": 2.4838762283325195, + "learning_rate": 2.362009897221165e-06, + "loss": 1.4098, + "step": 1694 + }, + { + "epoch": 1.741140215716487, + "grad_norm": 2.4860999584198, + "learning_rate": 2.360106585458698e-06, + "loss": 1.4006, + "step": 1695 + }, + { + "epoch": 1.7421674370826912, + "grad_norm": 2.3834593296051025, + "learning_rate": 2.3582032736962315e-06, + "loss": 1.3445, + "step": 1696 + }, + { + "epoch": 1.7431946584488958, + "grad_norm": 2.5383975505828857, + "learning_rate": 2.356299961933765e-06, + "loss": 1.3849, + "step": 1697 + }, + { + "epoch": 1.7442218798151001, + "grad_norm": 2.414038896560669, + "learning_rate": 2.354396650171298e-06, + "loss": 1.356, + "step": 1698 + }, + { + "epoch": 1.7452491011813045, + "grad_norm": 2.3932273387908936, + "learning_rate": 2.3524933384088316e-06, + "loss": 1.3422, + "step": 1699 + }, + { + "epoch": 1.746276322547509, + "grad_norm": 2.5618491172790527, + "learning_rate": 2.350590026646365e-06, + "loss": 1.4324, + "step": 1700 + }, + { + "epoch": 1.746276322547509, + "eval_loss": 1.7560412883758545, + "eval_runtime": 37.9455, + "eval_samples_per_second": 26.354, + "eval_steps_per_second": 1.107, + "step": 1700 + }, + { + "epoch": 1.7473035439137135, + "grad_norm": 2.450718879699707, + "learning_rate": 2.3486867148838983e-06, + "loss": 1.3631, + "step": 1701 + }, + { + "epoch": 1.7483307652799178, + "grad_norm": 2.4136962890625, + "learning_rate": 2.3467834031214317e-06, + "loss": 1.3528, + "step": 1702 + }, + { + "epoch": 1.7493579866461222, + "grad_norm": 2.4310078620910645, + "learning_rate": 2.3448800913589646e-06, + "loss": 1.369, + "step": 1703 + }, + { + "epoch": 1.7503852080123266, + "grad_norm": 2.5222883224487305, + "learning_rate": 2.342976779596498e-06, + "loss": 1.3897, + "step": 1704 + }, + { + "epoch": 1.7514124293785311, + "grad_norm": 2.3903377056121826, + "learning_rate": 2.3410734678340313e-06, + "loss": 1.3526, + "step": 1705 + }, + { + "epoch": 1.7524396507447355, + "grad_norm": 2.3869972229003906, + "learning_rate": 2.3391701560715647e-06, + "loss": 1.4091, + "step": 1706 + }, + { + "epoch": 1.7534668721109399, + "grad_norm": 2.3368723392486572, + "learning_rate": 2.337266844309098e-06, + "loss": 1.3671, + "step": 1707 + }, + { + "epoch": 1.7544940934771445, + "grad_norm": 2.4195168018341064, + "learning_rate": 2.3353635325466314e-06, + "loss": 1.3701, + "step": 1708 + }, + { + "epoch": 1.7555213148433486, + "grad_norm": 2.4312965869903564, + "learning_rate": 2.3334602207841648e-06, + "loss": 1.3795, + "step": 1709 + }, + { + "epoch": 1.7565485362095532, + "grad_norm": 2.582592725753784, + "learning_rate": 2.331556909021698e-06, + "loss": 1.3743, + "step": 1710 + }, + { + "epoch": 1.7565485362095532, + "eval_loss": 1.7604093551635742, + "eval_runtime": 37.6917, + "eval_samples_per_second": 26.531, + "eval_steps_per_second": 1.114, + "step": 1710 + }, + { + "epoch": 1.7575757575757576, + "grad_norm": 2.4194061756134033, + "learning_rate": 2.3296535972592315e-06, + "loss": 1.4049, + "step": 1711 + }, + { + "epoch": 1.758602978941962, + "grad_norm": 2.4093074798583984, + "learning_rate": 2.3277502854967644e-06, + "loss": 1.352, + "step": 1712 + }, + { + "epoch": 1.7596302003081665, + "grad_norm": 2.509335517883301, + "learning_rate": 2.325846973734298e-06, + "loss": 1.4221, + "step": 1713 + }, + { + "epoch": 1.7606574216743707, + "grad_norm": 2.422558069229126, + "learning_rate": 2.323943661971831e-06, + "loss": 1.3508, + "step": 1714 + }, + { + "epoch": 1.7616846430405753, + "grad_norm": 2.3949813842773438, + "learning_rate": 2.3220403502093645e-06, + "loss": 1.4146, + "step": 1715 + }, + { + "epoch": 1.7627118644067796, + "grad_norm": 2.47383189201355, + "learning_rate": 2.320137038446898e-06, + "loss": 1.413, + "step": 1716 + }, + { + "epoch": 1.763739085772984, + "grad_norm": 2.403531312942505, + "learning_rate": 2.318233726684431e-06, + "loss": 1.4181, + "step": 1717 + }, + { + "epoch": 1.7647663071391886, + "grad_norm": 2.463175058364868, + "learning_rate": 2.3163304149219646e-06, + "loss": 1.4121, + "step": 1718 + }, + { + "epoch": 1.765793528505393, + "grad_norm": 2.4082279205322266, + "learning_rate": 2.314427103159498e-06, + "loss": 1.3844, + "step": 1719 + }, + { + "epoch": 1.7668207498715973, + "grad_norm": 2.2674102783203125, + "learning_rate": 2.312523791397031e-06, + "loss": 1.3647, + "step": 1720 + }, + { + "epoch": 1.7668207498715973, + "eval_loss": 1.7476056814193726, + "eval_runtime": 37.8818, + "eval_samples_per_second": 26.398, + "eval_steps_per_second": 1.109, + "step": 1720 + }, + { + "epoch": 1.767847971237802, + "grad_norm": 2.2334558963775635, + "learning_rate": 2.3106204796345643e-06, + "loss": 1.3425, + "step": 1721 + }, + { + "epoch": 1.768875192604006, + "grad_norm": 2.3458974361419678, + "learning_rate": 2.3087171678720976e-06, + "loss": 1.3707, + "step": 1722 + }, + { + "epoch": 1.7699024139702106, + "grad_norm": 2.437530040740967, + "learning_rate": 2.306813856109631e-06, + "loss": 1.422, + "step": 1723 + }, + { + "epoch": 1.770929635336415, + "grad_norm": 2.467456817626953, + "learning_rate": 2.3049105443471644e-06, + "loss": 1.3795, + "step": 1724 + }, + { + "epoch": 1.7719568567026194, + "grad_norm": 2.485546827316284, + "learning_rate": 2.3030072325846973e-06, + "loss": 1.3648, + "step": 1725 + }, + { + "epoch": 1.772984078068824, + "grad_norm": 2.4228715896606445, + "learning_rate": 2.3011039208222307e-06, + "loss": 1.3508, + "step": 1726 + }, + { + "epoch": 1.774011299435028, + "grad_norm": 2.3490238189697266, + "learning_rate": 2.299200609059764e-06, + "loss": 1.3587, + "step": 1727 + }, + { + "epoch": 1.7750385208012327, + "grad_norm": 2.333501100540161, + "learning_rate": 2.297297297297298e-06, + "loss": 1.3593, + "step": 1728 + }, + { + "epoch": 1.776065742167437, + "grad_norm": 2.459859609603882, + "learning_rate": 2.2953939855348308e-06, + "loss": 1.3333, + "step": 1729 + }, + { + "epoch": 1.7770929635336414, + "grad_norm": 2.476691961288452, + "learning_rate": 2.293490673772364e-06, + "loss": 1.4137, + "step": 1730 + }, + { + "epoch": 1.7770929635336414, + "eval_loss": 1.7561018466949463, + "eval_runtime": 37.9778, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1730 + }, + { + "epoch": 1.778120184899846, + "grad_norm": 2.443805694580078, + "learning_rate": 2.2915873620098975e-06, + "loss": 1.3981, + "step": 1731 + }, + { + "epoch": 1.7791474062660502, + "grad_norm": 2.4164700508117676, + "learning_rate": 2.289684050247431e-06, + "loss": 1.392, + "step": 1732 + }, + { + "epoch": 1.7801746276322548, + "grad_norm": 2.3005855083465576, + "learning_rate": 2.287780738484964e-06, + "loss": 1.3603, + "step": 1733 + }, + { + "epoch": 1.7812018489984591, + "grad_norm": 2.395570755004883, + "learning_rate": 2.285877426722497e-06, + "loss": 1.383, + "step": 1734 + }, + { + "epoch": 1.7822290703646635, + "grad_norm": 2.4991118907928467, + "learning_rate": 2.2839741149600305e-06, + "loss": 1.4121, + "step": 1735 + }, + { + "epoch": 1.783256291730868, + "grad_norm": 2.420830726623535, + "learning_rate": 2.282070803197564e-06, + "loss": 1.3997, + "step": 1736 + }, + { + "epoch": 1.7842835130970724, + "grad_norm": 2.3576571941375732, + "learning_rate": 2.2801674914350972e-06, + "loss": 1.3825, + "step": 1737 + }, + { + "epoch": 1.7853107344632768, + "grad_norm": 2.355811595916748, + "learning_rate": 2.2782641796726306e-06, + "loss": 1.3716, + "step": 1738 + }, + { + "epoch": 1.7863379558294814, + "grad_norm": 2.3596901893615723, + "learning_rate": 2.276360867910164e-06, + "loss": 1.4064, + "step": 1739 + }, + { + "epoch": 1.7873651771956856, + "grad_norm": 2.301748514175415, + "learning_rate": 2.2744575561476973e-06, + "loss": 1.3358, + "step": 1740 + }, + { + "epoch": 1.7873651771956856, + "eval_loss": 1.7705498933792114, + "eval_runtime": 37.9787, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1740 + }, + { + "epoch": 1.7883923985618901, + "grad_norm": 2.3325891494750977, + "learning_rate": 2.2725542443852307e-06, + "loss": 1.3749, + "step": 1741 + }, + { + "epoch": 1.7894196199280945, + "grad_norm": 2.3880693912506104, + "learning_rate": 2.2706509326227636e-06, + "loss": 1.4153, + "step": 1742 + }, + { + "epoch": 1.7904468412942989, + "grad_norm": 2.3157029151916504, + "learning_rate": 2.268747620860297e-06, + "loss": 1.3899, + "step": 1743 + }, + { + "epoch": 1.7914740626605035, + "grad_norm": 2.4292151927948, + "learning_rate": 2.2668443090978303e-06, + "loss": 1.3804, + "step": 1744 + }, + { + "epoch": 1.7925012840267076, + "grad_norm": 2.4811818599700928, + "learning_rate": 2.2649409973353637e-06, + "loss": 1.4036, + "step": 1745 + }, + { + "epoch": 1.7935285053929122, + "grad_norm": 2.260420799255371, + "learning_rate": 2.263037685572897e-06, + "loss": 1.3616, + "step": 1746 + }, + { + "epoch": 1.7945557267591166, + "grad_norm": 2.262769937515259, + "learning_rate": 2.26113437381043e-06, + "loss": 1.351, + "step": 1747 + }, + { + "epoch": 1.795582948125321, + "grad_norm": 2.481602191925049, + "learning_rate": 2.259231062047964e-06, + "loss": 1.3816, + "step": 1748 + }, + { + "epoch": 1.7966101694915255, + "grad_norm": 2.584700345993042, + "learning_rate": 2.257327750285497e-06, + "loss": 1.4026, + "step": 1749 + }, + { + "epoch": 1.7976373908577299, + "grad_norm": 2.3472707271575928, + "learning_rate": 2.2554244385230305e-06, + "loss": 1.37, + "step": 1750 + }, + { + "epoch": 1.7976373908577299, + "eval_loss": 1.7779643535614014, + "eval_runtime": 38.0042, + "eval_samples_per_second": 26.313, + "eval_steps_per_second": 1.105, + "step": 1750 + }, + { + "epoch": 1.7986646122239343, + "grad_norm": 2.3616456985473633, + "learning_rate": 2.2535211267605635e-06, + "loss": 1.3713, + "step": 1751 + }, + { + "epoch": 1.7996918335901386, + "grad_norm": 2.4747135639190674, + "learning_rate": 2.251617814998097e-06, + "loss": 1.3797, + "step": 1752 + }, + { + "epoch": 1.800719054956343, + "grad_norm": 2.396944284439087, + "learning_rate": 2.24971450323563e-06, + "loss": 1.3081, + "step": 1753 + }, + { + "epoch": 1.8017462763225476, + "grad_norm": 2.5272505283355713, + "learning_rate": 2.2478111914731636e-06, + "loss": 1.3888, + "step": 1754 + }, + { + "epoch": 1.802773497688752, + "grad_norm": 2.319617748260498, + "learning_rate": 2.245907879710697e-06, + "loss": 1.2698, + "step": 1755 + }, + { + "epoch": 1.8038007190549563, + "grad_norm": 2.4994258880615234, + "learning_rate": 2.24400456794823e-06, + "loss": 1.4026, + "step": 1756 + }, + { + "epoch": 1.804827940421161, + "grad_norm": 2.3463289737701416, + "learning_rate": 2.2421012561857632e-06, + "loss": 1.3682, + "step": 1757 + }, + { + "epoch": 1.805855161787365, + "grad_norm": 2.3547732830047607, + "learning_rate": 2.240197944423297e-06, + "loss": 1.3486, + "step": 1758 + }, + { + "epoch": 1.8068823831535696, + "grad_norm": 2.381570816040039, + "learning_rate": 2.23829463266083e-06, + "loss": 1.3531, + "step": 1759 + }, + { + "epoch": 1.807909604519774, + "grad_norm": 2.2834391593933105, + "learning_rate": 2.2363913208983633e-06, + "loss": 1.3467, + "step": 1760 + }, + { + "epoch": 1.807909604519774, + "eval_loss": 1.7836403846740723, + "eval_runtime": 37.966, + "eval_samples_per_second": 26.339, + "eval_steps_per_second": 1.106, + "step": 1760 + }, + { + "epoch": 1.8089368258859784, + "grad_norm": 2.3574178218841553, + "learning_rate": 2.2344880091358967e-06, + "loss": 1.3867, + "step": 1761 + }, + { + "epoch": 1.809964047252183, + "grad_norm": 2.3440022468566895, + "learning_rate": 2.23258469737343e-06, + "loss": 1.3652, + "step": 1762 + }, + { + "epoch": 1.810991268618387, + "grad_norm": 2.4115090370178223, + "learning_rate": 2.2306813856109634e-06, + "loss": 1.4059, + "step": 1763 + }, + { + "epoch": 1.8120184899845917, + "grad_norm": 2.3989944458007812, + "learning_rate": 2.2287780738484963e-06, + "loss": 1.3572, + "step": 1764 + }, + { + "epoch": 1.813045711350796, + "grad_norm": 2.394892930984497, + "learning_rate": 2.2268747620860297e-06, + "loss": 1.433, + "step": 1765 + }, + { + "epoch": 1.8140729327170004, + "grad_norm": 2.5118515491485596, + "learning_rate": 2.224971450323563e-06, + "loss": 1.3714, + "step": 1766 + }, + { + "epoch": 1.815100154083205, + "grad_norm": 2.392695426940918, + "learning_rate": 2.2230681385610964e-06, + "loss": 1.3061, + "step": 1767 + }, + { + "epoch": 1.8161273754494094, + "grad_norm": 2.4015321731567383, + "learning_rate": 2.2211648267986298e-06, + "loss": 1.3681, + "step": 1768 + }, + { + "epoch": 1.8171545968156138, + "grad_norm": 2.3518455028533936, + "learning_rate": 2.219261515036163e-06, + "loss": 1.3549, + "step": 1769 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 2.40037202835083, + "learning_rate": 2.2173582032736965e-06, + "loss": 1.3882, + "step": 1770 + }, + { + "epoch": 1.8181818181818183, + "eval_loss": 1.7858384847640991, + "eval_runtime": 37.9787, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1770 + }, + { + "epoch": 1.8192090395480225, + "grad_norm": 2.4189655780792236, + "learning_rate": 2.21545489151123e-06, + "loss": 1.4019, + "step": 1771 + }, + { + "epoch": 1.820236260914227, + "grad_norm": 2.4965226650238037, + "learning_rate": 2.2135515797487632e-06, + "loss": 1.3703, + "step": 1772 + }, + { + "epoch": 1.8212634822804314, + "grad_norm": 2.4330036640167236, + "learning_rate": 2.211648267986296e-06, + "loss": 1.3689, + "step": 1773 + }, + { + "epoch": 1.8222907036466358, + "grad_norm": 2.3654327392578125, + "learning_rate": 2.2097449562238295e-06, + "loss": 1.3677, + "step": 1774 + }, + { + "epoch": 1.8233179250128404, + "grad_norm": 2.4047489166259766, + "learning_rate": 2.207841644461363e-06, + "loss": 1.3626, + "step": 1775 + }, + { + "epoch": 1.8243451463790445, + "grad_norm": 2.563310146331787, + "learning_rate": 2.2059383326988963e-06, + "loss": 1.424, + "step": 1776 + }, + { + "epoch": 1.8253723677452491, + "grad_norm": 2.506540536880493, + "learning_rate": 2.2040350209364296e-06, + "loss": 1.3727, + "step": 1777 + }, + { + "epoch": 1.8263995891114535, + "grad_norm": 2.271742105484009, + "learning_rate": 2.2021317091739626e-06, + "loss": 1.3412, + "step": 1778 + }, + { + "epoch": 1.8274268104776579, + "grad_norm": 2.3172881603240967, + "learning_rate": 2.2002283974114963e-06, + "loss": 1.3812, + "step": 1779 + }, + { + "epoch": 1.8284540318438625, + "grad_norm": 2.3927204608917236, + "learning_rate": 2.1983250856490297e-06, + "loss": 1.3148, + "step": 1780 + }, + { + "epoch": 1.8284540318438625, + "eval_loss": 1.7755651473999023, + "eval_runtime": 37.9896, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1780 + }, + { + "epoch": 1.8294812532100666, + "grad_norm": 2.391972303390503, + "learning_rate": 2.1964217738865626e-06, + "loss": 1.3553, + "step": 1781 + }, + { + "epoch": 1.8305084745762712, + "grad_norm": 2.3727471828460693, + "learning_rate": 2.194518462124096e-06, + "loss": 1.377, + "step": 1782 + }, + { + "epoch": 1.8315356959424756, + "grad_norm": 2.5614147186279297, + "learning_rate": 2.1926151503616294e-06, + "loss": 1.4188, + "step": 1783 + }, + { + "epoch": 1.83256291730868, + "grad_norm": 2.344714641571045, + "learning_rate": 2.1907118385991627e-06, + "loss": 1.3877, + "step": 1784 + }, + { + "epoch": 1.8335901386748845, + "grad_norm": 2.34287428855896, + "learning_rate": 2.188808526836696e-06, + "loss": 1.351, + "step": 1785 + }, + { + "epoch": 1.8346173600410889, + "grad_norm": 2.343554735183716, + "learning_rate": 2.1869052150742295e-06, + "loss": 1.382, + "step": 1786 + }, + { + "epoch": 1.8356445814072933, + "grad_norm": 2.2418034076690674, + "learning_rate": 2.1850019033117624e-06, + "loss": 1.3062, + "step": 1787 + }, + { + "epoch": 1.8366718027734978, + "grad_norm": 2.4358294010162354, + "learning_rate": 2.1830985915492958e-06, + "loss": 1.3681, + "step": 1788 + }, + { + "epoch": 1.837699024139702, + "grad_norm": 2.460731029510498, + "learning_rate": 2.1811952797868295e-06, + "loss": 1.3791, + "step": 1789 + }, + { + "epoch": 1.8387262455059066, + "grad_norm": 2.4244463443756104, + "learning_rate": 2.1792919680243625e-06, + "loss": 1.3528, + "step": 1790 + }, + { + "epoch": 1.8387262455059066, + "eval_loss": 1.764306664466858, + "eval_runtime": 37.9906, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 1790 + }, + { + "epoch": 1.839753466872111, + "grad_norm": 2.3808743953704834, + "learning_rate": 2.177388656261896e-06, + "loss": 1.3882, + "step": 1791 + }, + { + "epoch": 1.8407806882383153, + "grad_norm": 2.443690538406372, + "learning_rate": 2.1754853444994292e-06, + "loss": 1.4542, + "step": 1792 + }, + { + "epoch": 1.84180790960452, + "grad_norm": 2.3968520164489746, + "learning_rate": 2.1735820327369626e-06, + "loss": 1.2815, + "step": 1793 + }, + { + "epoch": 1.842835130970724, + "grad_norm": 2.5139951705932617, + "learning_rate": 2.171678720974496e-06, + "loss": 1.3986, + "step": 1794 + }, + { + "epoch": 1.8438623523369286, + "grad_norm": 2.3268849849700928, + "learning_rate": 2.169775409212029e-06, + "loss": 1.3923, + "step": 1795 + }, + { + "epoch": 1.844889573703133, + "grad_norm": 2.424268960952759, + "learning_rate": 2.1678720974495622e-06, + "loss": 1.4046, + "step": 1796 + }, + { + "epoch": 1.8459167950693374, + "grad_norm": 2.3144543170928955, + "learning_rate": 2.1659687856870956e-06, + "loss": 1.391, + "step": 1797 + }, + { + "epoch": 1.846944016435542, + "grad_norm": 2.4157726764678955, + "learning_rate": 2.164065473924629e-06, + "loss": 1.4084, + "step": 1798 + }, + { + "epoch": 1.8479712378017463, + "grad_norm": 2.3993160724639893, + "learning_rate": 2.1621621621621623e-06, + "loss": 1.3372, + "step": 1799 + }, + { + "epoch": 1.8489984591679507, + "grad_norm": 2.3267669677734375, + "learning_rate": 2.1602588503996957e-06, + "loss": 1.3899, + "step": 1800 + }, + { + "epoch": 1.8489984591679507, + "eval_loss": 1.7611773014068604, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 1800 + }, + { + "epoch": 1.850025680534155, + "grad_norm": 2.488994836807251, + "learning_rate": 2.158355538637229e-06, + "loss": 1.3647, + "step": 1801 + }, + { + "epoch": 1.8510529019003594, + "grad_norm": 2.32352876663208, + "learning_rate": 2.1564522268747624e-06, + "loss": 1.3615, + "step": 1802 + }, + { + "epoch": 1.852080123266564, + "grad_norm": 2.667987585067749, + "learning_rate": 2.1545489151122958e-06, + "loss": 1.4388, + "step": 1803 + }, + { + "epoch": 1.8531073446327684, + "grad_norm": 2.4570565223693848, + "learning_rate": 2.1526456033498287e-06, + "loss": 1.3939, + "step": 1804 + }, + { + "epoch": 1.8541345659989728, + "grad_norm": 2.4037978649139404, + "learning_rate": 2.150742291587362e-06, + "loss": 1.3834, + "step": 1805 + }, + { + "epoch": 1.8551617873651773, + "grad_norm": 2.42080020904541, + "learning_rate": 2.1488389798248954e-06, + "loss": 1.3758, + "step": 1806 + }, + { + "epoch": 1.8561890087313815, + "grad_norm": 2.3361966609954834, + "learning_rate": 2.146935668062429e-06, + "loss": 1.4024, + "step": 1807 + }, + { + "epoch": 1.857216230097586, + "grad_norm": 2.251089334487915, + "learning_rate": 2.145032356299962e-06, + "loss": 1.3365, + "step": 1808 + }, + { + "epoch": 1.8582434514637904, + "grad_norm": 2.4046289920806885, + "learning_rate": 2.1431290445374955e-06, + "loss": 1.3831, + "step": 1809 + }, + { + "epoch": 1.8592706728299948, + "grad_norm": 2.2613766193389893, + "learning_rate": 2.141225732775029e-06, + "loss": 1.3423, + "step": 1810 + }, + { + "epoch": 1.8592706728299948, + "eval_loss": 1.752854347229004, + "eval_runtime": 37.6469, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 1810 + }, + { + "epoch": 1.8602978941961994, + "grad_norm": 2.377535581588745, + "learning_rate": 2.1393224210125623e-06, + "loss": 1.3796, + "step": 1811 + }, + { + "epoch": 1.8613251155624035, + "grad_norm": 2.3959462642669678, + "learning_rate": 2.137419109250095e-06, + "loss": 1.3664, + "step": 1812 + }, + { + "epoch": 1.8623523369286081, + "grad_norm": 2.5317859649658203, + "learning_rate": 2.1355157974876286e-06, + "loss": 1.418, + "step": 1813 + }, + { + "epoch": 1.8633795582948125, + "grad_norm": 2.5266010761260986, + "learning_rate": 2.133612485725162e-06, + "loss": 1.4117, + "step": 1814 + }, + { + "epoch": 1.8644067796610169, + "grad_norm": 2.3911185264587402, + "learning_rate": 2.1317091739626953e-06, + "loss": 1.3206, + "step": 1815 + }, + { + "epoch": 1.8654340010272215, + "grad_norm": 2.358596086502075, + "learning_rate": 2.1298058622002286e-06, + "loss": 1.3475, + "step": 1816 + }, + { + "epoch": 1.8664612223934258, + "grad_norm": 2.348175048828125, + "learning_rate": 2.1279025504377616e-06, + "loss": 1.3769, + "step": 1817 + }, + { + "epoch": 1.8674884437596302, + "grad_norm": 2.417470932006836, + "learning_rate": 2.125999238675295e-06, + "loss": 1.3667, + "step": 1818 + }, + { + "epoch": 1.8685156651258348, + "grad_norm": 2.5871050357818604, + "learning_rate": 2.1240959269128287e-06, + "loss": 1.4139, + "step": 1819 + }, + { + "epoch": 1.869542886492039, + "grad_norm": 2.384218692779541, + "learning_rate": 2.122192615150362e-06, + "loss": 1.3755, + "step": 1820 + }, + { + "epoch": 1.869542886492039, + "eval_loss": 1.7497111558914185, + "eval_runtime": 37.7655, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 1820 + }, + { + "epoch": 1.8705701078582435, + "grad_norm": 2.354877233505249, + "learning_rate": 2.120289303387895e-06, + "loss": 1.3174, + "step": 1821 + }, + { + "epoch": 1.8715973292244479, + "grad_norm": 2.4020509719848633, + "learning_rate": 2.1183859916254284e-06, + "loss": 1.3699, + "step": 1822 + }, + { + "epoch": 1.8726245505906522, + "grad_norm": 2.2876627445220947, + "learning_rate": 2.1164826798629618e-06, + "loss": 1.4053, + "step": 1823 + }, + { + "epoch": 1.8736517719568568, + "grad_norm": 2.2526464462280273, + "learning_rate": 2.114579368100495e-06, + "loss": 1.3569, + "step": 1824 + }, + { + "epoch": 1.874678993323061, + "grad_norm": 2.340773344039917, + "learning_rate": 2.1126760563380285e-06, + "loss": 1.3439, + "step": 1825 + }, + { + "epoch": 1.8757062146892656, + "grad_norm": 2.315812110900879, + "learning_rate": 2.1107727445755614e-06, + "loss": 1.3129, + "step": 1826 + }, + { + "epoch": 1.87673343605547, + "grad_norm": 2.480487108230591, + "learning_rate": 2.1088694328130948e-06, + "loss": 1.4372, + "step": 1827 + }, + { + "epoch": 1.8777606574216743, + "grad_norm": 2.267662286758423, + "learning_rate": 2.106966121050628e-06, + "loss": 1.3429, + "step": 1828 + }, + { + "epoch": 1.878787878787879, + "grad_norm": 2.383023977279663, + "learning_rate": 2.1050628092881615e-06, + "loss": 1.3963, + "step": 1829 + }, + { + "epoch": 1.879815100154083, + "grad_norm": 2.4784996509552, + "learning_rate": 2.103159497525695e-06, + "loss": 1.4405, + "step": 1830 + }, + { + "epoch": 1.879815100154083, + "eval_loss": 1.7609645128250122, + "eval_runtime": 37.904, + "eval_samples_per_second": 26.382, + "eval_steps_per_second": 1.108, + "step": 1830 + }, + { + "epoch": 1.8808423215202876, + "grad_norm": 2.359614133834839, + "learning_rate": 2.1012561857632282e-06, + "loss": 1.3577, + "step": 1831 + }, + { + "epoch": 1.881869542886492, + "grad_norm": 2.452862501144409, + "learning_rate": 2.0993528740007616e-06, + "loss": 1.4082, + "step": 1832 + }, + { + "epoch": 1.8828967642526964, + "grad_norm": 2.384011745452881, + "learning_rate": 2.097449562238295e-06, + "loss": 1.4175, + "step": 1833 + }, + { + "epoch": 1.883923985618901, + "grad_norm": 2.4906508922576904, + "learning_rate": 2.095546250475828e-06, + "loss": 1.3693, + "step": 1834 + }, + { + "epoch": 1.8849512069851053, + "grad_norm": 2.510040521621704, + "learning_rate": 2.0936429387133613e-06, + "loss": 1.3703, + "step": 1835 + }, + { + "epoch": 1.8859784283513097, + "grad_norm": 2.3438405990600586, + "learning_rate": 2.0917396269508946e-06, + "loss": 1.4311, + "step": 1836 + }, + { + "epoch": 1.8870056497175143, + "grad_norm": 2.532226800918579, + "learning_rate": 2.089836315188428e-06, + "loss": 1.3986, + "step": 1837 + }, + { + "epoch": 1.8880328710837184, + "grad_norm": 2.4368748664855957, + "learning_rate": 2.0879330034259614e-06, + "loss": 1.4025, + "step": 1838 + }, + { + "epoch": 1.889060092449923, + "grad_norm": 2.4487054347991943, + "learning_rate": 2.0860296916634947e-06, + "loss": 1.3562, + "step": 1839 + }, + { + "epoch": 1.8900873138161274, + "grad_norm": 2.6636273860931396, + "learning_rate": 2.084126379901028e-06, + "loss": 1.4384, + "step": 1840 + }, + { + "epoch": 1.8900873138161274, + "eval_loss": 1.7675000429153442, + "eval_runtime": 37.9371, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 1840 + }, + { + "epoch": 1.8911145351823317, + "grad_norm": 2.497738838195801, + "learning_rate": 2.0822230681385614e-06, + "loss": 1.3462, + "step": 1841 + }, + { + "epoch": 1.8921417565485363, + "grad_norm": 2.4419288635253906, + "learning_rate": 2.080319756376095e-06, + "loss": 1.3974, + "step": 1842 + }, + { + "epoch": 1.8931689779147405, + "grad_norm": 2.3671085834503174, + "learning_rate": 2.0784164446136277e-06, + "loss": 1.3917, + "step": 1843 + }, + { + "epoch": 1.894196199280945, + "grad_norm": 2.3837625980377197, + "learning_rate": 2.076513132851161e-06, + "loss": 1.3663, + "step": 1844 + }, + { + "epoch": 1.8952234206471494, + "grad_norm": 2.590935468673706, + "learning_rate": 2.0746098210886945e-06, + "loss": 1.3962, + "step": 1845 + }, + { + "epoch": 1.8962506420133538, + "grad_norm": 2.2917118072509766, + "learning_rate": 2.072706509326228e-06, + "loss": 1.3641, + "step": 1846 + }, + { + "epoch": 1.8972778633795584, + "grad_norm": 2.4366025924682617, + "learning_rate": 2.070803197563761e-06, + "loss": 1.3749, + "step": 1847 + }, + { + "epoch": 1.8983050847457628, + "grad_norm": 2.385152816772461, + "learning_rate": 2.068899885801294e-06, + "loss": 1.3655, + "step": 1848 + }, + { + "epoch": 1.8993323061119671, + "grad_norm": 2.473550796508789, + "learning_rate": 2.066996574038828e-06, + "loss": 1.3917, + "step": 1849 + }, + { + "epoch": 1.9003595274781715, + "grad_norm": 2.360003709793091, + "learning_rate": 2.0650932622763613e-06, + "loss": 1.4089, + "step": 1850 + }, + { + "epoch": 1.9003595274781715, + "eval_loss": 1.7553730010986328, + "eval_runtime": 37.9412, + "eval_samples_per_second": 26.357, + "eval_steps_per_second": 1.107, + "step": 1850 + }, + { + "epoch": 1.9013867488443759, + "grad_norm": 2.3891544342041016, + "learning_rate": 2.0631899505138942e-06, + "loss": 1.3033, + "step": 1851 + }, + { + "epoch": 1.9024139702105805, + "grad_norm": 2.330413341522217, + "learning_rate": 2.0612866387514276e-06, + "loss": 1.3446, + "step": 1852 + }, + { + "epoch": 1.9034411915767848, + "grad_norm": 2.3061635494232178, + "learning_rate": 2.059383326988961e-06, + "loss": 1.3872, + "step": 1853 + }, + { + "epoch": 1.9044684129429892, + "grad_norm": 2.4351041316986084, + "learning_rate": 2.0574800152264943e-06, + "loss": 1.362, + "step": 1854 + }, + { + "epoch": 1.9054956343091938, + "grad_norm": 2.447535991668701, + "learning_rate": 2.0555767034640277e-06, + "loss": 1.3417, + "step": 1855 + }, + { + "epoch": 1.906522855675398, + "grad_norm": 2.369133472442627, + "learning_rate": 2.053673391701561e-06, + "loss": 1.3181, + "step": 1856 + }, + { + "epoch": 1.9075500770416025, + "grad_norm": 2.3961751461029053, + "learning_rate": 2.051770079939094e-06, + "loss": 1.3913, + "step": 1857 + }, + { + "epoch": 1.9085772984078069, + "grad_norm": 2.40244460105896, + "learning_rate": 2.0498667681766273e-06, + "loss": 1.3447, + "step": 1858 + }, + { + "epoch": 1.9096045197740112, + "grad_norm": 2.4332613945007324, + "learning_rate": 2.047963456414161e-06, + "loss": 1.3427, + "step": 1859 + }, + { + "epoch": 1.9106317411402158, + "grad_norm": 2.3410897254943848, + "learning_rate": 2.046060144651694e-06, + "loss": 1.3697, + "step": 1860 + }, + { + "epoch": 1.9106317411402158, + "eval_loss": 1.74896240234375, + "eval_runtime": 37.9315, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 1860 + }, + { + "epoch": 1.91165896250642, + "grad_norm": 2.3965389728546143, + "learning_rate": 2.0441568328892274e-06, + "loss": 1.3752, + "step": 1861 + }, + { + "epoch": 1.9126861838726246, + "grad_norm": 2.3173274993896484, + "learning_rate": 2.0422535211267608e-06, + "loss": 1.3807, + "step": 1862 + }, + { + "epoch": 1.913713405238829, + "grad_norm": 2.2883224487304688, + "learning_rate": 2.040350209364294e-06, + "loss": 1.3551, + "step": 1863 + }, + { + "epoch": 1.9147406266050333, + "grad_norm": 2.2770445346832275, + "learning_rate": 2.0384468976018275e-06, + "loss": 1.3405, + "step": 1864 + }, + { + "epoch": 1.915767847971238, + "grad_norm": 2.3694701194763184, + "learning_rate": 2.0365435858393604e-06, + "loss": 1.3538, + "step": 1865 + }, + { + "epoch": 1.9167950693374423, + "grad_norm": 2.4186787605285645, + "learning_rate": 2.034640274076894e-06, + "loss": 1.353, + "step": 1866 + }, + { + "epoch": 1.9178222907036466, + "grad_norm": 2.4059510231018066, + "learning_rate": 2.032736962314427e-06, + "loss": 1.4149, + "step": 1867 + }, + { + "epoch": 1.9188495120698512, + "grad_norm": 2.5529062747955322, + "learning_rate": 2.0308336505519605e-06, + "loss": 1.3669, + "step": 1868 + }, + { + "epoch": 1.9198767334360554, + "grad_norm": 2.3925650119781494, + "learning_rate": 2.028930338789494e-06, + "loss": 1.4178, + "step": 1869 + }, + { + "epoch": 1.92090395480226, + "grad_norm": 2.5203816890716553, + "learning_rate": 2.0270270270270273e-06, + "loss": 1.3418, + "step": 1870 + }, + { + "epoch": 1.92090395480226, + "eval_loss": 1.7473065853118896, + "eval_runtime": 37.9634, + "eval_samples_per_second": 26.341, + "eval_steps_per_second": 1.106, + "step": 1870 + }, + { + "epoch": 1.9219311761684643, + "grad_norm": 2.4792747497558594, + "learning_rate": 2.0251237152645606e-06, + "loss": 1.3611, + "step": 1871 + }, + { + "epoch": 1.9229583975346687, + "grad_norm": 2.357645273208618, + "learning_rate": 2.023220403502094e-06, + "loss": 1.3433, + "step": 1872 + }, + { + "epoch": 1.9239856189008733, + "grad_norm": 2.284149408340454, + "learning_rate": 2.0213170917396273e-06, + "loss": 1.3157, + "step": 1873 + }, + { + "epoch": 1.9250128402670774, + "grad_norm": 2.3971638679504395, + "learning_rate": 2.0194137799771603e-06, + "loss": 1.3435, + "step": 1874 + }, + { + "epoch": 1.926040061633282, + "grad_norm": 2.605325222015381, + "learning_rate": 2.0175104682146937e-06, + "loss": 1.3619, + "step": 1875 + }, + { + "epoch": 1.9270672829994864, + "grad_norm": 2.3957643508911133, + "learning_rate": 2.015607156452227e-06, + "loss": 1.3078, + "step": 1876 + }, + { + "epoch": 1.9280945043656907, + "grad_norm": 2.3887429237365723, + "learning_rate": 2.0137038446897604e-06, + "loss": 1.3447, + "step": 1877 + }, + { + "epoch": 1.9291217257318953, + "grad_norm": 2.31253719329834, + "learning_rate": 2.0118005329272937e-06, + "loss": 1.3467, + "step": 1878 + }, + { + "epoch": 1.9301489470980995, + "grad_norm": 2.3084523677825928, + "learning_rate": 2.0098972211648267e-06, + "loss": 1.3466, + "step": 1879 + }, + { + "epoch": 1.931176168464304, + "grad_norm": 2.3024210929870605, + "learning_rate": 2.0079939094023605e-06, + "loss": 1.3525, + "step": 1880 + }, + { + "epoch": 1.931176168464304, + "eval_loss": 1.7498610019683838, + "eval_runtime": 37.9363, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 1880 + }, + { + "epoch": 1.9322033898305084, + "grad_norm": 2.342660665512085, + "learning_rate": 2.006090597639894e-06, + "loss": 1.4013, + "step": 1881 + }, + { + "epoch": 1.9332306111967128, + "grad_norm": 2.4165151119232178, + "learning_rate": 2.0041872858774268e-06, + "loss": 1.4154, + "step": 1882 + }, + { + "epoch": 1.9342578325629174, + "grad_norm": 2.3674707412719727, + "learning_rate": 2.00228397411496e-06, + "loss": 1.3903, + "step": 1883 + }, + { + "epoch": 1.9352850539291218, + "grad_norm": 2.2478187084198, + "learning_rate": 2.0003806623524935e-06, + "loss": 1.3708, + "step": 1884 + }, + { + "epoch": 1.9363122752953261, + "grad_norm": 2.328073740005493, + "learning_rate": 1.998477350590027e-06, + "loss": 1.3013, + "step": 1885 + }, + { + "epoch": 1.9373394966615307, + "grad_norm": 2.4772889614105225, + "learning_rate": 1.9965740388275602e-06, + "loss": 1.3818, + "step": 1886 + }, + { + "epoch": 1.9383667180277349, + "grad_norm": 2.456451892852783, + "learning_rate": 1.994670727065093e-06, + "loss": 1.3806, + "step": 1887 + }, + { + "epoch": 1.9393939393939394, + "grad_norm": 2.325561046600342, + "learning_rate": 1.9927674153026265e-06, + "loss": 1.3262, + "step": 1888 + }, + { + "epoch": 1.9404211607601438, + "grad_norm": 2.447673797607422, + "learning_rate": 1.99086410354016e-06, + "loss": 1.3776, + "step": 1889 + }, + { + "epoch": 1.9414483821263482, + "grad_norm": 2.4231903553009033, + "learning_rate": 1.9889607917776937e-06, + "loss": 1.3593, + "step": 1890 + }, + { + "epoch": 1.9414483821263482, + "eval_loss": 1.7612230777740479, + "eval_runtime": 37.9427, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 1890 + }, + { + "epoch": 1.9424756034925528, + "grad_norm": 2.5399789810180664, + "learning_rate": 1.9870574800152266e-06, + "loss": 1.3737, + "step": 1891 + }, + { + "epoch": 1.943502824858757, + "grad_norm": 2.3155572414398193, + "learning_rate": 1.98515416825276e-06, + "loss": 1.388, + "step": 1892 + }, + { + "epoch": 1.9445300462249615, + "grad_norm": 2.33693265914917, + "learning_rate": 1.9832508564902933e-06, + "loss": 1.3734, + "step": 1893 + }, + { + "epoch": 1.9455572675911659, + "grad_norm": 2.2426419258117676, + "learning_rate": 1.9813475447278267e-06, + "loss": 1.3011, + "step": 1894 + }, + { + "epoch": 1.9465844889573702, + "grad_norm": 2.4306788444519043, + "learning_rate": 1.97944423296536e-06, + "loss": 1.3721, + "step": 1895 + }, + { + "epoch": 1.9476117103235748, + "grad_norm": 2.329252004623413, + "learning_rate": 1.977540921202893e-06, + "loss": 1.363, + "step": 1896 + }, + { + "epoch": 1.9486389316897792, + "grad_norm": 2.3523757457733154, + "learning_rate": 1.9756376094404264e-06, + "loss": 1.3485, + "step": 1897 + }, + { + "epoch": 1.9496661530559836, + "grad_norm": 2.3240344524383545, + "learning_rate": 1.9737342976779597e-06, + "loss": 1.3258, + "step": 1898 + }, + { + "epoch": 1.950693374422188, + "grad_norm": 2.4068586826324463, + "learning_rate": 1.971830985915493e-06, + "loss": 1.406, + "step": 1899 + }, + { + "epoch": 1.9517205957883923, + "grad_norm": 2.385835886001587, + "learning_rate": 1.9699276741530264e-06, + "loss": 1.3641, + "step": 1900 + }, + { + "epoch": 1.9517205957883923, + "eval_loss": 1.747411847114563, + "eval_runtime": 37.9419, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 1900 + } + ], + "logging_steps": 1, + "max_steps": 2919, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 5.922887877787648e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/zero_to_fp32.py b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-1900/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/generation_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/latest b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/latest new file mode 100644 index 0000000000000000000000000000000000000000..ca3fd898b349457b6d1b62649e5bdfc274b799e1 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/latest @@ -0,0 +1 @@ +global_step2700 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/model.safetensors.index.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..12d8f134d0ace768aa9a4fc9a1c7787b2ce68b40 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1fd745605ae8f150c3645104631b2d3c3b452eecda3ab1f66d0a3f474af022ff +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..7255d4b96382a76a01d62b0f3fb8ed7ba214a8cf --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cab5d049f3d3fc102f2df0c1651bc9d13107a3c0c772fd671f399b9d7d1de545 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..cf2513c6be393ffd7f18b9966efe1ea94961ce76 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1e03f5804d1958af28b457d1c95eadeccb6d55a564daf236124a0bd65854c268 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..138ac5c44b7629d0888e04b5366cbbcd2662867e --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:93c6dc12af5b71d39808ec73fd5e1515807e922d6f0ea05f52389019207ceb08 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/special_tokens_map.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/tokenizer_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0b782d4602b361f7577abb52b710a8c0e088588d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/tokenizer_config.json @@ -0,0 +1,2078 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128257": { + "content": "🅁", + "lstrip": true, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/trainer_state.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..117ed506c0a7512c63962224d60ba91d0dfbfd75 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/trainer_state.json @@ -0,0 +1,21093 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.773497688751926, + "eval_steps": 10, + "global_step": 2700, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7525, + "step": 1 + }, + { + "epoch": 0.002054442732408834, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7605, + "step": 2 + }, + { + "epoch": 0.0030816640986132513, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8008, + "step": 3 + }, + { + "epoch": 0.004108885464817668, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.809, + "step": 4 + }, + { + "epoch": 0.005136106831022085, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7841, + "step": 5 + }, + { + "epoch": 0.0061633281972265025, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7924, + "step": 6 + }, + { + "epoch": 0.007190549563430919, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7732, + "step": 7 + }, + { + "epoch": 0.008217770929635337, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7953, + "step": 8 + }, + { + "epoch": 0.009244992295839754, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7747, + "step": 9 + }, + { + "epoch": 0.01027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7909, + "step": 10 + }, + { + "epoch": 0.01027221366204417, + "eval_loss": 2.813098192214966, + "eval_runtime": 38.1187, + "eval_samples_per_second": 26.234, + "eval_steps_per_second": 1.102, + "step": 10 + }, + { + "epoch": 0.011299435028248588, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7592, + "step": 11 + }, + { + "epoch": 0.012326656394453005, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7394, + "step": 12 + }, + { + "epoch": 0.01335387776065742, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8111, + "step": 13 + }, + { + "epoch": 0.014381099126861838, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7786, + "step": 14 + }, + { + "epoch": 0.015408320493066256, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7711, + "step": 15 + }, + { + "epoch": 0.016435541859270673, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7622, + "step": 16 + }, + { + "epoch": 0.01746276322547509, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7291, + "step": 17 + }, + { + "epoch": 0.01848998459167951, + "grad_norm": 6.8249125480651855, + "learning_rate": 3.424657534246575e-08, + "loss": 2.7995, + "step": 18 + }, + { + "epoch": 0.019517205957883924, + "grad_norm": 5.581949710845947, + "learning_rate": 5.136986301369863e-08, + "loss": 2.5998, + "step": 19 + }, + { + "epoch": 0.02054442732408834, + "grad_norm": 3.0888617038726807, + "learning_rate": 6.84931506849315e-08, + "loss": 2.6031, + "step": 20 + }, + { + "epoch": 0.02054442732408834, + "eval_loss": 2.60642147064209, + "eval_runtime": 38.2226, + "eval_samples_per_second": 26.163, + "eval_steps_per_second": 1.099, + "step": 20 + }, + { + "epoch": 0.02157164869029276, + "grad_norm": 4.42941427230835, + "learning_rate": 8.561643835616439e-08, + "loss": 2.5816, + "step": 21 + }, + { + "epoch": 0.022598870056497175, + "grad_norm": 2.359813690185547, + "learning_rate": 1.0273972602739726e-07, + "loss": 2.5352, + "step": 22 + }, + { + "epoch": 0.02362609142270159, + "grad_norm": 2.9022927284240723, + "learning_rate": 1.1986301369863014e-07, + "loss": 2.4641, + "step": 23 + }, + { + "epoch": 0.02465331278890601, + "grad_norm": 1.6895997524261475, + "learning_rate": 1.36986301369863e-07, + "loss": 2.4582, + "step": 24 + }, + { + "epoch": 0.025680534155110426, + "grad_norm": 2.4555776119232178, + "learning_rate": 1.541095890410959e-07, + "loss": 2.4171, + "step": 25 + }, + { + "epoch": 0.02670775552131484, + "grad_norm": 1.9735779762268066, + "learning_rate": 1.7123287671232878e-07, + "loss": 2.4249, + "step": 26 + }, + { + "epoch": 0.02773497688751926, + "grad_norm": 2.71161150932312, + "learning_rate": 1.8835616438356165e-07, + "loss": 2.3791, + "step": 27 + }, + { + "epoch": 0.028762198253723677, + "grad_norm": 1.8482733964920044, + "learning_rate": 2.0547945205479452e-07, + "loss": 2.374, + "step": 28 + }, + { + "epoch": 0.029789419619928096, + "grad_norm": 2.191246747970581, + "learning_rate": 2.226027397260274e-07, + "loss": 2.3557, + "step": 29 + }, + { + "epoch": 0.030816640986132512, + "grad_norm": 2.0575971603393555, + "learning_rate": 2.397260273972603e-07, + "loss": 2.3315, + "step": 30 + }, + { + "epoch": 0.030816640986132512, + "eval_loss": 2.399639844894409, + "eval_runtime": 37.737, + "eval_samples_per_second": 26.499, + "eval_steps_per_second": 1.113, + "step": 30 + }, + { + "epoch": 0.03184386235233693, + "grad_norm": 2.0346758365631104, + "learning_rate": 2.568493150684932e-07, + "loss": 2.2791, + "step": 31 + }, + { + "epoch": 0.03287108371854135, + "grad_norm": 1.8710602521896362, + "learning_rate": 2.73972602739726e-07, + "loss": 2.3219, + "step": 32 + }, + { + "epoch": 0.03389830508474576, + "grad_norm": 1.9982675313949585, + "learning_rate": 2.910958904109589e-07, + "loss": 2.273, + "step": 33 + }, + { + "epoch": 0.03492552645095018, + "grad_norm": 2.092076301574707, + "learning_rate": 3.082191780821918e-07, + "loss": 2.2555, + "step": 34 + }, + { + "epoch": 0.035952747817154594, + "grad_norm": 1.870373249053955, + "learning_rate": 3.2534246575342466e-07, + "loss": 2.2314, + "step": 35 + }, + { + "epoch": 0.03697996918335902, + "grad_norm": 1.8378052711486816, + "learning_rate": 3.4246575342465755e-07, + "loss": 2.1244, + "step": 36 + }, + { + "epoch": 0.03800719054956343, + "grad_norm": 2.0156054496765137, + "learning_rate": 3.595890410958904e-07, + "loss": 2.2502, + "step": 37 + }, + { + "epoch": 0.03903441191576785, + "grad_norm": 1.6175312995910645, + "learning_rate": 3.767123287671233e-07, + "loss": 2.1566, + "step": 38 + }, + { + "epoch": 0.040061633281972264, + "grad_norm": 1.9051450490951538, + "learning_rate": 3.938356164383562e-07, + "loss": 2.167, + "step": 39 + }, + { + "epoch": 0.04108885464817668, + "grad_norm": 2.383063793182373, + "learning_rate": 4.1095890410958903e-07, + "loss": 2.2099, + "step": 40 + }, + { + "epoch": 0.04108885464817668, + "eval_loss": 2.2763354778289795, + "eval_runtime": 37.8085, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 40 + }, + { + "epoch": 0.042116076014381096, + "grad_norm": 2.1310720443725586, + "learning_rate": 4.2808219178082193e-07, + "loss": 2.1714, + "step": 41 + }, + { + "epoch": 0.04314329738058552, + "grad_norm": 1.771941900253296, + "learning_rate": 4.452054794520548e-07, + "loss": 2.0895, + "step": 42 + }, + { + "epoch": 0.044170518746789934, + "grad_norm": 2.3784172534942627, + "learning_rate": 4.6232876712328767e-07, + "loss": 2.1555, + "step": 43 + }, + { + "epoch": 0.04519774011299435, + "grad_norm": 1.6692813634872437, + "learning_rate": 4.794520547945206e-07, + "loss": 2.11, + "step": 44 + }, + { + "epoch": 0.046224961479198766, + "grad_norm": 2.1780643463134766, + "learning_rate": 4.965753424657534e-07, + "loss": 2.1256, + "step": 45 + }, + { + "epoch": 0.04725218284540318, + "grad_norm": 1.9390568733215332, + "learning_rate": 5.136986301369864e-07, + "loss": 2.0971, + "step": 46 + }, + { + "epoch": 0.048279404211607604, + "grad_norm": 2.276216506958008, + "learning_rate": 5.308219178082192e-07, + "loss": 2.1479, + "step": 47 + }, + { + "epoch": 0.04930662557781202, + "grad_norm": 2.148573160171509, + "learning_rate": 5.47945205479452e-07, + "loss": 2.0808, + "step": 48 + }, + { + "epoch": 0.050333846944016436, + "grad_norm": 2.259422540664673, + "learning_rate": 5.65068493150685e-07, + "loss": 2.0823, + "step": 49 + }, + { + "epoch": 0.05136106831022085, + "grad_norm": 1.7703359127044678, + "learning_rate": 5.821917808219178e-07, + "loss": 2.0236, + "step": 50 + }, + { + "epoch": 0.05136106831022085, + "eval_loss": 2.2037274837493896, + "eval_runtime": 37.8385, + "eval_samples_per_second": 26.428, + "eval_steps_per_second": 1.11, + "step": 50 + }, + { + "epoch": 0.05238828967642527, + "grad_norm": 2.56268310546875, + "learning_rate": 5.993150684931507e-07, + "loss": 2.0228, + "step": 51 + }, + { + "epoch": 0.05341551104262968, + "grad_norm": 2.5256848335266113, + "learning_rate": 6.164383561643836e-07, + "loss": 2.1401, + "step": 52 + }, + { + "epoch": 0.054442732408834106, + "grad_norm": 2.6329638957977295, + "learning_rate": 6.335616438356165e-07, + "loss": 2.0107, + "step": 53 + }, + { + "epoch": 0.05546995377503852, + "grad_norm": 2.477762222290039, + "learning_rate": 6.506849315068493e-07, + "loss": 2.0648, + "step": 54 + }, + { + "epoch": 0.05649717514124294, + "grad_norm": 1.8869198560714722, + "learning_rate": 6.678082191780823e-07, + "loss": 2.0214, + "step": 55 + }, + { + "epoch": 0.05752439650744735, + "grad_norm": 2.707345485687256, + "learning_rate": 6.849315068493151e-07, + "loss": 1.9893, + "step": 56 + }, + { + "epoch": 0.05855161787365177, + "grad_norm": 2.067230701446533, + "learning_rate": 7.020547945205481e-07, + "loss": 2.0308, + "step": 57 + }, + { + "epoch": 0.05957883923985619, + "grad_norm": 2.1961982250213623, + "learning_rate": 7.191780821917808e-07, + "loss": 1.948, + "step": 58 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 2.1032280921936035, + "learning_rate": 7.363013698630137e-07, + "loss": 2.0098, + "step": 59 + }, + { + "epoch": 0.061633281972265024, + "grad_norm": 2.5191144943237305, + "learning_rate": 7.534246575342466e-07, + "loss": 2.0183, + "step": 60 + }, + { + "epoch": 0.061633281972265024, + "eval_loss": 2.1474344730377197, + "eval_runtime": 37.8631, + "eval_samples_per_second": 26.411, + "eval_steps_per_second": 1.109, + "step": 60 + }, + { + "epoch": 0.06266050333846944, + "grad_norm": 2.177739143371582, + "learning_rate": 7.705479452054795e-07, + "loss": 1.9702, + "step": 61 + }, + { + "epoch": 0.06368772470467386, + "grad_norm": 2.499784231185913, + "learning_rate": 7.876712328767124e-07, + "loss": 1.9825, + "step": 62 + }, + { + "epoch": 0.06471494607087827, + "grad_norm": 2.181591749191284, + "learning_rate": 8.047945205479453e-07, + "loss": 2.0414, + "step": 63 + }, + { + "epoch": 0.0657421674370827, + "grad_norm": 1.9575999975204468, + "learning_rate": 8.219178082191781e-07, + "loss": 2.0115, + "step": 64 + }, + { + "epoch": 0.0667693888032871, + "grad_norm": 2.3431336879730225, + "learning_rate": 8.39041095890411e-07, + "loss": 1.8781, + "step": 65 + }, + { + "epoch": 0.06779661016949153, + "grad_norm": 2.042747974395752, + "learning_rate": 8.561643835616439e-07, + "loss": 1.9798, + "step": 66 + }, + { + "epoch": 0.06882383153569595, + "grad_norm": 2.784104585647583, + "learning_rate": 8.732876712328768e-07, + "loss": 1.9274, + "step": 67 + }, + { + "epoch": 0.06985105290190036, + "grad_norm": 1.9781445264816284, + "learning_rate": 8.904109589041097e-07, + "loss": 2.0036, + "step": 68 + }, + { + "epoch": 0.07087827426810478, + "grad_norm": 2.631537437438965, + "learning_rate": 9.075342465753426e-07, + "loss": 1.938, + "step": 69 + }, + { + "epoch": 0.07190549563430919, + "grad_norm": 1.902701497077942, + "learning_rate": 9.246575342465753e-07, + "loss": 2.0545, + "step": 70 + }, + { + "epoch": 0.07190549563430919, + "eval_loss": 2.115269422531128, + "eval_runtime": 37.887, + "eval_samples_per_second": 26.394, + "eval_steps_per_second": 1.109, + "step": 70 + }, + { + "epoch": 0.07293271700051361, + "grad_norm": 1.9323904514312744, + "learning_rate": 9.417808219178083e-07, + "loss": 1.9463, + "step": 71 + }, + { + "epoch": 0.07395993836671803, + "grad_norm": 2.1631956100463867, + "learning_rate": 9.589041095890411e-07, + "loss": 1.8982, + "step": 72 + }, + { + "epoch": 0.07498715973292244, + "grad_norm": 1.959518313407898, + "learning_rate": 9.76027397260274e-07, + "loss": 1.9538, + "step": 73 + }, + { + "epoch": 0.07601438109912687, + "grad_norm": 1.9834569692611694, + "learning_rate": 9.931506849315068e-07, + "loss": 1.9257, + "step": 74 + }, + { + "epoch": 0.07704160246533127, + "grad_norm": 2.2818732261657715, + "learning_rate": 1.0102739726027399e-06, + "loss": 1.9318, + "step": 75 + }, + { + "epoch": 0.0780688238315357, + "grad_norm": 1.7696406841278076, + "learning_rate": 1.0273972602739727e-06, + "loss": 1.912, + "step": 76 + }, + { + "epoch": 0.07909604519774012, + "grad_norm": 2.396303176879883, + "learning_rate": 1.0445205479452056e-06, + "loss": 1.9482, + "step": 77 + }, + { + "epoch": 0.08012326656394453, + "grad_norm": 2.2371881008148193, + "learning_rate": 1.0616438356164384e-06, + "loss": 1.9448, + "step": 78 + }, + { + "epoch": 0.08115048793014895, + "grad_norm": 2.40077543258667, + "learning_rate": 1.0787671232876712e-06, + "loss": 1.8764, + "step": 79 + }, + { + "epoch": 0.08217770929635336, + "grad_norm": 2.267916679382324, + "learning_rate": 1.095890410958904e-06, + "loss": 1.9478, + "step": 80 + }, + { + "epoch": 0.08217770929635336, + "eval_loss": 2.0851893424987793, + "eval_runtime": 37.9176, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 80 + }, + { + "epoch": 0.08320493066255778, + "grad_norm": 2.286259651184082, + "learning_rate": 1.1130136986301371e-06, + "loss": 1.8856, + "step": 81 + }, + { + "epoch": 0.08423215202876219, + "grad_norm": 2.160848379135132, + "learning_rate": 1.13013698630137e-06, + "loss": 1.9196, + "step": 82 + }, + { + "epoch": 0.08525937339496661, + "grad_norm": 2.3618931770324707, + "learning_rate": 1.1472602739726028e-06, + "loss": 1.9229, + "step": 83 + }, + { + "epoch": 0.08628659476117104, + "grad_norm": 2.1963961124420166, + "learning_rate": 1.1643835616438357e-06, + "loss": 1.8998, + "step": 84 + }, + { + "epoch": 0.08731381612737545, + "grad_norm": 2.197981357574463, + "learning_rate": 1.1815068493150685e-06, + "loss": 1.8716, + "step": 85 + }, + { + "epoch": 0.08834103749357987, + "grad_norm": 2.3562612533569336, + "learning_rate": 1.1986301369863014e-06, + "loss": 1.9336, + "step": 86 + }, + { + "epoch": 0.08936825885978428, + "grad_norm": 2.49712872505188, + "learning_rate": 1.2157534246575344e-06, + "loss": 1.8827, + "step": 87 + }, + { + "epoch": 0.0903954802259887, + "grad_norm": 2.315244436264038, + "learning_rate": 1.2328767123287673e-06, + "loss": 1.8801, + "step": 88 + }, + { + "epoch": 0.09142270159219312, + "grad_norm": 2.774587869644165, + "learning_rate": 1.25e-06, + "loss": 1.8972, + "step": 89 + }, + { + "epoch": 0.09244992295839753, + "grad_norm": 2.262364387512207, + "learning_rate": 1.267123287671233e-06, + "loss": 1.8426, + "step": 90 + }, + { + "epoch": 0.09244992295839753, + "eval_loss": 2.069462299346924, + "eval_runtime": 37.9319, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 90 + }, + { + "epoch": 0.09347714432460195, + "grad_norm": 2.6195597648620605, + "learning_rate": 1.284246575342466e-06, + "loss": 1.871, + "step": 91 + }, + { + "epoch": 0.09450436569080636, + "grad_norm": 2.051847219467163, + "learning_rate": 1.3013698630136986e-06, + "loss": 1.8922, + "step": 92 + }, + { + "epoch": 0.09553158705701079, + "grad_norm": 2.5930886268615723, + "learning_rate": 1.3184931506849317e-06, + "loss": 1.8818, + "step": 93 + }, + { + "epoch": 0.09655880842321521, + "grad_norm": 2.304164171218872, + "learning_rate": 1.3356164383561645e-06, + "loss": 1.8779, + "step": 94 + }, + { + "epoch": 0.09758602978941962, + "grad_norm": 2.187354564666748, + "learning_rate": 1.3527397260273976e-06, + "loss": 1.9272, + "step": 95 + }, + { + "epoch": 0.09861325115562404, + "grad_norm": 2.2492458820343018, + "learning_rate": 1.3698630136986302e-06, + "loss": 1.805, + "step": 96 + }, + { + "epoch": 0.09964047252182845, + "grad_norm": 2.3325352668762207, + "learning_rate": 1.386986301369863e-06, + "loss": 1.8817, + "step": 97 + }, + { + "epoch": 0.10066769388803287, + "grad_norm": 2.1708617210388184, + "learning_rate": 1.4041095890410961e-06, + "loss": 1.7956, + "step": 98 + }, + { + "epoch": 0.1016949152542373, + "grad_norm": 2.4267706871032715, + "learning_rate": 1.421232876712329e-06, + "loss": 1.8116, + "step": 99 + }, + { + "epoch": 0.1027221366204417, + "grad_norm": 2.3286805152893066, + "learning_rate": 1.4383561643835616e-06, + "loss": 1.8299, + "step": 100 + }, + { + "epoch": 0.1027221366204417, + "eval_loss": 2.0549135208129883, + "eval_runtime": 37.935, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 100 + }, + { + "epoch": 0.10374935798664613, + "grad_norm": 2.5391643047332764, + "learning_rate": 1.4554794520547946e-06, + "loss": 1.8521, + "step": 101 + }, + { + "epoch": 0.10477657935285054, + "grad_norm": 2.018380880355835, + "learning_rate": 1.4726027397260275e-06, + "loss": 1.8209, + "step": 102 + }, + { + "epoch": 0.10580380071905496, + "grad_norm": 2.3706746101379395, + "learning_rate": 1.4897260273972605e-06, + "loss": 1.79, + "step": 103 + }, + { + "epoch": 0.10683102208525937, + "grad_norm": 2.1290764808654785, + "learning_rate": 1.5068493150684932e-06, + "loss": 1.8141, + "step": 104 + }, + { + "epoch": 0.10785824345146379, + "grad_norm": 2.2768120765686035, + "learning_rate": 1.5239726027397262e-06, + "loss": 1.8826, + "step": 105 + }, + { + "epoch": 0.10888546481766821, + "grad_norm": 2.316584348678589, + "learning_rate": 1.541095890410959e-06, + "loss": 1.8604, + "step": 106 + }, + { + "epoch": 0.10991268618387262, + "grad_norm": 2.14679217338562, + "learning_rate": 1.5582191780821921e-06, + "loss": 1.865, + "step": 107 + }, + { + "epoch": 0.11093990755007704, + "grad_norm": 2.3288021087646484, + "learning_rate": 1.5753424657534248e-06, + "loss": 1.8512, + "step": 108 + }, + { + "epoch": 0.11196712891628145, + "grad_norm": 2.300482988357544, + "learning_rate": 1.5924657534246576e-06, + "loss": 1.7455, + "step": 109 + }, + { + "epoch": 0.11299435028248588, + "grad_norm": 2.4226229190826416, + "learning_rate": 1.6095890410958907e-06, + "loss": 1.8058, + "step": 110 + }, + { + "epoch": 0.11299435028248588, + "eval_loss": 2.025261640548706, + "eval_runtime": 37.6779, + "eval_samples_per_second": 26.541, + "eval_steps_per_second": 1.115, + "step": 110 + }, + { + "epoch": 0.1140215716486903, + "grad_norm": 2.0930111408233643, + "learning_rate": 1.6267123287671235e-06, + "loss": 1.7375, + "step": 111 + }, + { + "epoch": 0.1150487930148947, + "grad_norm": 2.4973645210266113, + "learning_rate": 1.6438356164383561e-06, + "loss": 1.8268, + "step": 112 + }, + { + "epoch": 0.11607601438109913, + "grad_norm": 2.3295114040374756, + "learning_rate": 1.6609589041095892e-06, + "loss": 1.7882, + "step": 113 + }, + { + "epoch": 0.11710323574730354, + "grad_norm": 2.3693318367004395, + "learning_rate": 1.678082191780822e-06, + "loss": 1.8192, + "step": 114 + }, + { + "epoch": 0.11813045711350796, + "grad_norm": 2.553637742996216, + "learning_rate": 1.695205479452055e-06, + "loss": 1.7491, + "step": 115 + }, + { + "epoch": 0.11915767847971238, + "grad_norm": 2.303945541381836, + "learning_rate": 1.7123287671232877e-06, + "loss": 1.8737, + "step": 116 + }, + { + "epoch": 0.12018489984591679, + "grad_norm": 2.250356435775757, + "learning_rate": 1.7294520547945206e-06, + "loss": 1.8039, + "step": 117 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 2.3854403495788574, + "learning_rate": 1.7465753424657536e-06, + "loss": 1.8357, + "step": 118 + }, + { + "epoch": 0.12223934257832562, + "grad_norm": 2.4049572944641113, + "learning_rate": 1.7636986301369865e-06, + "loss": 1.8209, + "step": 119 + }, + { + "epoch": 0.12326656394453005, + "grad_norm": 2.313551425933838, + "learning_rate": 1.7808219178082193e-06, + "loss": 1.9026, + "step": 120 + }, + { + "epoch": 0.12326656394453005, + "eval_loss": 2.012171745300293, + "eval_runtime": 37.7908, + "eval_samples_per_second": 26.461, + "eval_steps_per_second": 1.111, + "step": 120 + }, + { + "epoch": 0.12429378531073447, + "grad_norm": 2.518320322036743, + "learning_rate": 1.7979452054794521e-06, + "loss": 1.8241, + "step": 121 + }, + { + "epoch": 0.12532100667693888, + "grad_norm": 2.6504592895507812, + "learning_rate": 1.8150684931506852e-06, + "loss": 1.8091, + "step": 122 + }, + { + "epoch": 0.1263482280431433, + "grad_norm": 2.556610107421875, + "learning_rate": 1.832191780821918e-06, + "loss": 1.7547, + "step": 123 + }, + { + "epoch": 0.12737544940934772, + "grad_norm": 2.247633695602417, + "learning_rate": 1.8493150684931507e-06, + "loss": 1.7923, + "step": 124 + }, + { + "epoch": 0.12840267077555212, + "grad_norm": 2.195596933364868, + "learning_rate": 1.8664383561643837e-06, + "loss": 1.747, + "step": 125 + }, + { + "epoch": 0.12942989214175654, + "grad_norm": 2.121396780014038, + "learning_rate": 1.8835616438356166e-06, + "loss": 1.8088, + "step": 126 + }, + { + "epoch": 0.13045711350796096, + "grad_norm": 2.4572501182556152, + "learning_rate": 1.9006849315068496e-06, + "loss": 1.7285, + "step": 127 + }, + { + "epoch": 0.1314843348741654, + "grad_norm": 2.2590177059173584, + "learning_rate": 1.9178082191780823e-06, + "loss": 1.7929, + "step": 128 + }, + { + "epoch": 0.1325115562403698, + "grad_norm": 2.5099234580993652, + "learning_rate": 1.9349315068493153e-06, + "loss": 1.8365, + "step": 129 + }, + { + "epoch": 0.1335387776065742, + "grad_norm": 2.3064465522766113, + "learning_rate": 1.952054794520548e-06, + "loss": 1.8351, + "step": 130 + }, + { + "epoch": 0.1335387776065742, + "eval_loss": 2.0043866634368896, + "eval_runtime": 37.8661, + "eval_samples_per_second": 26.409, + "eval_steps_per_second": 1.109, + "step": 130 + }, + { + "epoch": 0.13456599897277863, + "grad_norm": 2.5977718830108643, + "learning_rate": 1.969178082191781e-06, + "loss": 1.7603, + "step": 131 + }, + { + "epoch": 0.13559322033898305, + "grad_norm": 2.23221755027771, + "learning_rate": 1.9863013698630136e-06, + "loss": 1.8237, + "step": 132 + }, + { + "epoch": 0.13662044170518747, + "grad_norm": 2.247044324874878, + "learning_rate": 2.0034246575342467e-06, + "loss": 1.8006, + "step": 133 + }, + { + "epoch": 0.1376476630713919, + "grad_norm": 2.1796107292175293, + "learning_rate": 2.0205479452054797e-06, + "loss": 1.7354, + "step": 134 + }, + { + "epoch": 0.1386748844375963, + "grad_norm": 2.3269693851470947, + "learning_rate": 2.037671232876713e-06, + "loss": 1.783, + "step": 135 + }, + { + "epoch": 0.1397021058038007, + "grad_norm": 2.2903053760528564, + "learning_rate": 2.0547945205479454e-06, + "loss": 1.8072, + "step": 136 + }, + { + "epoch": 0.14072932717000514, + "grad_norm": 2.634150981903076, + "learning_rate": 2.071917808219178e-06, + "loss": 1.7251, + "step": 137 + }, + { + "epoch": 0.14175654853620956, + "grad_norm": 2.42579984664917, + "learning_rate": 2.089041095890411e-06, + "loss": 1.7997, + "step": 138 + }, + { + "epoch": 0.14278376990241398, + "grad_norm": 2.615828514099121, + "learning_rate": 2.106164383561644e-06, + "loss": 1.766, + "step": 139 + }, + { + "epoch": 0.14381099126861838, + "grad_norm": 2.312714099884033, + "learning_rate": 2.123287671232877e-06, + "loss": 1.7538, + "step": 140 + }, + { + "epoch": 0.14381099126861838, + "eval_loss": 1.9918214082717896, + "eval_runtime": 37.8698, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 140 + }, + { + "epoch": 0.1448382126348228, + "grad_norm": 2.7540030479431152, + "learning_rate": 2.14041095890411e-06, + "loss": 1.7856, + "step": 141 + }, + { + "epoch": 0.14586543400102722, + "grad_norm": 2.289191246032715, + "learning_rate": 2.1575342465753425e-06, + "loss": 1.7288, + "step": 142 + }, + { + "epoch": 0.14689265536723164, + "grad_norm": 2.788001775741577, + "learning_rate": 2.1746575342465755e-06, + "loss": 1.8703, + "step": 143 + }, + { + "epoch": 0.14791987673343607, + "grad_norm": 2.4128854274749756, + "learning_rate": 2.191780821917808e-06, + "loss": 1.7461, + "step": 144 + }, + { + "epoch": 0.14894709809964046, + "grad_norm": 2.4108099937438965, + "learning_rate": 2.2089041095890412e-06, + "loss": 1.73, + "step": 145 + }, + { + "epoch": 0.14997431946584489, + "grad_norm": 2.4315948486328125, + "learning_rate": 2.2260273972602743e-06, + "loss": 1.7303, + "step": 146 + }, + { + "epoch": 0.1510015408320493, + "grad_norm": 2.494056224822998, + "learning_rate": 2.243150684931507e-06, + "loss": 1.7616, + "step": 147 + }, + { + "epoch": 0.15202876219825373, + "grad_norm": 2.4557394981384277, + "learning_rate": 2.26027397260274e-06, + "loss": 1.7655, + "step": 148 + }, + { + "epoch": 0.15305598356445815, + "grad_norm": 2.4114091396331787, + "learning_rate": 2.2773972602739726e-06, + "loss": 1.8538, + "step": 149 + }, + { + "epoch": 0.15408320493066255, + "grad_norm": 2.609219789505005, + "learning_rate": 2.2945205479452057e-06, + "loss": 1.7767, + "step": 150 + }, + { + "epoch": 0.15408320493066255, + "eval_loss": 1.981367588043213, + "eval_runtime": 37.9562, + "eval_samples_per_second": 26.346, + "eval_steps_per_second": 1.107, + "step": 150 + }, + { + "epoch": 0.15511042629686697, + "grad_norm": 2.213622808456421, + "learning_rate": 2.3116438356164387e-06, + "loss": 1.687, + "step": 151 + }, + { + "epoch": 0.1561376476630714, + "grad_norm": 2.6517975330352783, + "learning_rate": 2.3287671232876713e-06, + "loss": 1.7311, + "step": 152 + }, + { + "epoch": 0.15716486902927582, + "grad_norm": 2.5174458026885986, + "learning_rate": 2.3458904109589044e-06, + "loss": 1.7835, + "step": 153 + }, + { + "epoch": 0.15819209039548024, + "grad_norm": 2.485480308532715, + "learning_rate": 2.363013698630137e-06, + "loss": 1.7749, + "step": 154 + }, + { + "epoch": 0.15921931176168463, + "grad_norm": 2.5386061668395996, + "learning_rate": 2.38013698630137e-06, + "loss": 1.7223, + "step": 155 + }, + { + "epoch": 0.16024653312788906, + "grad_norm": 2.2895941734313965, + "learning_rate": 2.3972602739726027e-06, + "loss": 1.6845, + "step": 156 + }, + { + "epoch": 0.16127375449409348, + "grad_norm": 2.5453789234161377, + "learning_rate": 2.4143835616438358e-06, + "loss": 1.7627, + "step": 157 + }, + { + "epoch": 0.1623009758602979, + "grad_norm": 2.5322303771972656, + "learning_rate": 2.431506849315069e-06, + "loss": 1.764, + "step": 158 + }, + { + "epoch": 0.1633281972265023, + "grad_norm": 2.2669899463653564, + "learning_rate": 2.4486301369863015e-06, + "loss": 1.7487, + "step": 159 + }, + { + "epoch": 0.16435541859270672, + "grad_norm": 2.4577853679656982, + "learning_rate": 2.4657534246575345e-06, + "loss": 1.7635, + "step": 160 + }, + { + "epoch": 0.16435541859270672, + "eval_loss": 1.9793355464935303, + "eval_runtime": 37.8576, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 160 + }, + { + "epoch": 0.16538263995891114, + "grad_norm": 2.354893922805786, + "learning_rate": 2.482876712328767e-06, + "loss": 1.7538, + "step": 161 + }, + { + "epoch": 0.16640986132511557, + "grad_norm": 2.3255856037139893, + "learning_rate": 2.5e-06, + "loss": 1.7277, + "step": 162 + }, + { + "epoch": 0.16743708269132, + "grad_norm": 2.3501994609832764, + "learning_rate": 2.5171232876712333e-06, + "loss": 1.6985, + "step": 163 + }, + { + "epoch": 0.16846430405752438, + "grad_norm": 2.386627674102783, + "learning_rate": 2.534246575342466e-06, + "loss": 1.7552, + "step": 164 + }, + { + "epoch": 0.1694915254237288, + "grad_norm": 2.2238457202911377, + "learning_rate": 2.551369863013699e-06, + "loss": 1.7805, + "step": 165 + }, + { + "epoch": 0.17051874678993323, + "grad_norm": 2.2653403282165527, + "learning_rate": 2.568493150684932e-06, + "loss": 1.7656, + "step": 166 + }, + { + "epoch": 0.17154596815613765, + "grad_norm": 2.233132839202881, + "learning_rate": 2.585616438356164e-06, + "loss": 1.6997, + "step": 167 + }, + { + "epoch": 0.17257318952234207, + "grad_norm": 2.1696043014526367, + "learning_rate": 2.6027397260273973e-06, + "loss": 1.7487, + "step": 168 + }, + { + "epoch": 0.17360041088854647, + "grad_norm": 2.4479899406433105, + "learning_rate": 2.6198630136986303e-06, + "loss": 1.7957, + "step": 169 + }, + { + "epoch": 0.1746276322547509, + "grad_norm": 2.3054702281951904, + "learning_rate": 2.6369863013698634e-06, + "loss": 1.766, + "step": 170 + }, + { + "epoch": 0.1746276322547509, + "eval_loss": 1.9543790817260742, + "eval_runtime": 37.8837, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 170 + }, + { + "epoch": 0.17565485362095531, + "grad_norm": 2.3591110706329346, + "learning_rate": 2.654109589041096e-06, + "loss": 1.7149, + "step": 171 + }, + { + "epoch": 0.17668207498715974, + "grad_norm": 2.450526237487793, + "learning_rate": 2.671232876712329e-06, + "loss": 1.7074, + "step": 172 + }, + { + "epoch": 0.17770929635336416, + "grad_norm": 2.1511754989624023, + "learning_rate": 2.688356164383562e-06, + "loss": 1.7036, + "step": 173 + }, + { + "epoch": 0.17873651771956856, + "grad_norm": 2.5944275856018066, + "learning_rate": 2.705479452054795e-06, + "loss": 1.6566, + "step": 174 + }, + { + "epoch": 0.17976373908577298, + "grad_norm": 2.4301369190216064, + "learning_rate": 2.7226027397260274e-06, + "loss": 1.7977, + "step": 175 + }, + { + "epoch": 0.1807909604519774, + "grad_norm": 2.287665367126465, + "learning_rate": 2.7397260273972604e-06, + "loss": 1.7526, + "step": 176 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.380032777786255, + "learning_rate": 2.7568493150684935e-06, + "loss": 1.6753, + "step": 177 + }, + { + "epoch": 0.18284540318438625, + "grad_norm": 2.4040510654449463, + "learning_rate": 2.773972602739726e-06, + "loss": 1.758, + "step": 178 + }, + { + "epoch": 0.18387262455059064, + "grad_norm": 2.5074045658111572, + "learning_rate": 2.791095890410959e-06, + "loss": 1.6498, + "step": 179 + }, + { + "epoch": 0.18489984591679506, + "grad_norm": 2.3926949501037598, + "learning_rate": 2.8082191780821922e-06, + "loss": 1.8006, + "step": 180 + }, + { + "epoch": 0.18489984591679506, + "eval_loss": 1.9446762800216675, + "eval_runtime": 37.8572, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 180 + }, + { + "epoch": 0.1859270672829995, + "grad_norm": 2.4991633892059326, + "learning_rate": 2.8253424657534253e-06, + "loss": 1.7649, + "step": 181 + }, + { + "epoch": 0.1869542886492039, + "grad_norm": 2.2558324337005615, + "learning_rate": 2.842465753424658e-06, + "loss": 1.7743, + "step": 182 + }, + { + "epoch": 0.18798151001540833, + "grad_norm": 2.219717025756836, + "learning_rate": 2.8595890410958905e-06, + "loss": 1.7807, + "step": 183 + }, + { + "epoch": 0.18900873138161273, + "grad_norm": 2.406562566757202, + "learning_rate": 2.876712328767123e-06, + "loss": 1.6289, + "step": 184 + }, + { + "epoch": 0.19003595274781715, + "grad_norm": 2.356579303741455, + "learning_rate": 2.8938356164383562e-06, + "loss": 1.6747, + "step": 185 + }, + { + "epoch": 0.19106317411402157, + "grad_norm": 2.4844322204589844, + "learning_rate": 2.9109589041095893e-06, + "loss": 1.6731, + "step": 186 + }, + { + "epoch": 0.192090395480226, + "grad_norm": 2.8179149627685547, + "learning_rate": 2.9280821917808223e-06, + "loss": 1.6786, + "step": 187 + }, + { + "epoch": 0.19311761684643042, + "grad_norm": 2.395416259765625, + "learning_rate": 2.945205479452055e-06, + "loss": 1.7313, + "step": 188 + }, + { + "epoch": 0.1941448382126348, + "grad_norm": 2.848740816116333, + "learning_rate": 2.962328767123288e-06, + "loss": 1.7258, + "step": 189 + }, + { + "epoch": 0.19517205957883924, + "grad_norm": 2.427395820617676, + "learning_rate": 2.979452054794521e-06, + "loss": 1.6633, + "step": 190 + }, + { + "epoch": 0.19517205957883924, + "eval_loss": 1.9443999528884888, + "eval_runtime": 37.8579, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 190 + }, + { + "epoch": 0.19619928094504366, + "grad_norm": 2.5347256660461426, + "learning_rate": 2.9965753424657533e-06, + "loss": 1.7342, + "step": 191 + }, + { + "epoch": 0.19722650231124808, + "grad_norm": 2.8696866035461426, + "learning_rate": 3.0136986301369864e-06, + "loss": 1.7037, + "step": 192 + }, + { + "epoch": 0.1982537236774525, + "grad_norm": 2.427468776702881, + "learning_rate": 3.0308219178082194e-06, + "loss": 1.672, + "step": 193 + }, + { + "epoch": 0.1992809450436569, + "grad_norm": 2.4253737926483154, + "learning_rate": 3.0479452054794525e-06, + "loss": 1.7252, + "step": 194 + }, + { + "epoch": 0.20030816640986132, + "grad_norm": 2.801668643951416, + "learning_rate": 3.065068493150685e-06, + "loss": 1.6694, + "step": 195 + }, + { + "epoch": 0.20133538777606574, + "grad_norm": 2.442263603210449, + "learning_rate": 3.082191780821918e-06, + "loss": 1.6866, + "step": 196 + }, + { + "epoch": 0.20236260914227017, + "grad_norm": 2.777463674545288, + "learning_rate": 3.099315068493151e-06, + "loss": 1.686, + "step": 197 + }, + { + "epoch": 0.2033898305084746, + "grad_norm": 2.491227388381958, + "learning_rate": 3.1164383561643843e-06, + "loss": 1.6414, + "step": 198 + }, + { + "epoch": 0.20441705187467898, + "grad_norm": 2.33585262298584, + "learning_rate": 3.1335616438356165e-06, + "loss": 1.7808, + "step": 199 + }, + { + "epoch": 0.2054442732408834, + "grad_norm": 2.518160581588745, + "learning_rate": 3.1506849315068495e-06, + "loss": 1.6592, + "step": 200 + }, + { + "epoch": 0.2054442732408834, + "eval_loss": 1.9371097087860107, + "eval_runtime": 37.8602, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 200 + }, + { + "epoch": 0.20647149460708783, + "grad_norm": 2.460204839706421, + "learning_rate": 3.167808219178082e-06, + "loss": 1.6668, + "step": 201 + }, + { + "epoch": 0.20749871597329225, + "grad_norm": 2.377824068069458, + "learning_rate": 3.184931506849315e-06, + "loss": 1.6462, + "step": 202 + }, + { + "epoch": 0.20852593733949668, + "grad_norm": 2.426177740097046, + "learning_rate": 3.2020547945205483e-06, + "loss": 1.6272, + "step": 203 + }, + { + "epoch": 0.20955315870570107, + "grad_norm": 2.4532034397125244, + "learning_rate": 3.2191780821917813e-06, + "loss": 1.6169, + "step": 204 + }, + { + "epoch": 0.2105803800719055, + "grad_norm": 2.517059564590454, + "learning_rate": 3.236301369863014e-06, + "loss": 1.5991, + "step": 205 + }, + { + "epoch": 0.21160760143810992, + "grad_norm": 2.22658109664917, + "learning_rate": 3.253424657534247e-06, + "loss": 1.6754, + "step": 206 + }, + { + "epoch": 0.21263482280431434, + "grad_norm": 2.344799518585205, + "learning_rate": 3.2705479452054796e-06, + "loss": 1.6648, + "step": 207 + }, + { + "epoch": 0.21366204417051873, + "grad_norm": 2.331875801086426, + "learning_rate": 3.2876712328767123e-06, + "loss": 1.6112, + "step": 208 + }, + { + "epoch": 0.21468926553672316, + "grad_norm": 2.262343645095825, + "learning_rate": 3.3047945205479453e-06, + "loss": 1.6479, + "step": 209 + }, + { + "epoch": 0.21571648690292758, + "grad_norm": 2.282278537750244, + "learning_rate": 3.3219178082191784e-06, + "loss": 1.6753, + "step": 210 + }, + { + "epoch": 0.21571648690292758, + "eval_loss": 1.9338418245315552, + "eval_runtime": 37.6083, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 210 + }, + { + "epoch": 0.216743708269132, + "grad_norm": 2.578476905822754, + "learning_rate": 3.3390410958904114e-06, + "loss": 1.7043, + "step": 211 + }, + { + "epoch": 0.21777092963533642, + "grad_norm": 2.3118813037872314, + "learning_rate": 3.356164383561644e-06, + "loss": 1.7541, + "step": 212 + }, + { + "epoch": 0.21879815100154082, + "grad_norm": 2.2418692111968994, + "learning_rate": 3.373287671232877e-06, + "loss": 1.6886, + "step": 213 + }, + { + "epoch": 0.21982537236774524, + "grad_norm": 2.3583438396453857, + "learning_rate": 3.39041095890411e-06, + "loss": 1.7241, + "step": 214 + }, + { + "epoch": 0.22085259373394966, + "grad_norm": 2.345090866088867, + "learning_rate": 3.4075342465753424e-06, + "loss": 1.6194, + "step": 215 + }, + { + "epoch": 0.2218798151001541, + "grad_norm": 2.3996200561523438, + "learning_rate": 3.4246575342465754e-06, + "loss": 1.6889, + "step": 216 + }, + { + "epoch": 0.2229070364663585, + "grad_norm": 2.3552374839782715, + "learning_rate": 3.4417808219178085e-06, + "loss": 1.6713, + "step": 217 + }, + { + "epoch": 0.2239342578325629, + "grad_norm": 2.2653744220733643, + "learning_rate": 3.458904109589041e-06, + "loss": 1.673, + "step": 218 + }, + { + "epoch": 0.22496147919876733, + "grad_norm": 2.263288974761963, + "learning_rate": 3.476027397260274e-06, + "loss": 1.673, + "step": 219 + }, + { + "epoch": 0.22598870056497175, + "grad_norm": 2.3173880577087402, + "learning_rate": 3.4931506849315072e-06, + "loss": 1.6803, + "step": 220 + }, + { + "epoch": 0.22598870056497175, + "eval_loss": 1.9166476726531982, + "eval_runtime": 37.7659, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 220 + }, + { + "epoch": 0.22701592193117617, + "grad_norm": 2.4005839824676514, + "learning_rate": 3.5102739726027403e-06, + "loss": 1.6809, + "step": 221 + }, + { + "epoch": 0.2280431432973806, + "grad_norm": 2.5032525062561035, + "learning_rate": 3.527397260273973e-06, + "loss": 1.6411, + "step": 222 + }, + { + "epoch": 0.229070364663585, + "grad_norm": 2.533583879470825, + "learning_rate": 3.5445205479452056e-06, + "loss": 1.6608, + "step": 223 + }, + { + "epoch": 0.2300975860297894, + "grad_norm": 2.3575572967529297, + "learning_rate": 3.5616438356164386e-06, + "loss": 1.7236, + "step": 224 + }, + { + "epoch": 0.23112480739599384, + "grad_norm": 2.3408942222595215, + "learning_rate": 3.5787671232876712e-06, + "loss": 1.7101, + "step": 225 + }, + { + "epoch": 0.23215202876219826, + "grad_norm": 2.3819596767425537, + "learning_rate": 3.5958904109589043e-06, + "loss": 1.6414, + "step": 226 + }, + { + "epoch": 0.23317925012840268, + "grad_norm": 2.2874338626861572, + "learning_rate": 3.6130136986301373e-06, + "loss": 1.6767, + "step": 227 + }, + { + "epoch": 0.23420647149460708, + "grad_norm": 2.3523948192596436, + "learning_rate": 3.6301369863013704e-06, + "loss": 1.6932, + "step": 228 + }, + { + "epoch": 0.2352336928608115, + "grad_norm": 2.476128578186035, + "learning_rate": 3.647260273972603e-06, + "loss": 1.7125, + "step": 229 + }, + { + "epoch": 0.23626091422701592, + "grad_norm": 2.3841593265533447, + "learning_rate": 3.664383561643836e-06, + "loss": 1.6469, + "step": 230 + }, + { + "epoch": 0.23626091422701592, + "eval_loss": 1.925522804260254, + "eval_runtime": 37.8581, + "eval_samples_per_second": 26.414, + "eval_steps_per_second": 1.109, + "step": 230 + }, + { + "epoch": 0.23728813559322035, + "grad_norm": 2.4401519298553467, + "learning_rate": 3.6815068493150687e-06, + "loss": 1.6726, + "step": 231 + }, + { + "epoch": 0.23831535695942477, + "grad_norm": 2.329674005508423, + "learning_rate": 3.6986301369863014e-06, + "loss": 1.6349, + "step": 232 + }, + { + "epoch": 0.23934257832562916, + "grad_norm": 2.654290199279785, + "learning_rate": 3.7157534246575344e-06, + "loss": 1.7309, + "step": 233 + }, + { + "epoch": 0.24036979969183359, + "grad_norm": 2.1599373817443848, + "learning_rate": 3.7328767123287675e-06, + "loss": 1.6437, + "step": 234 + }, + { + "epoch": 0.241397021058038, + "grad_norm": 2.6170263290405273, + "learning_rate": 3.7500000000000005e-06, + "loss": 1.6862, + "step": 235 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 2.2836415767669678, + "learning_rate": 3.767123287671233e-06, + "loss": 1.7073, + "step": 236 + }, + { + "epoch": 0.24345146379044685, + "grad_norm": 2.3001456260681152, + "learning_rate": 3.784246575342466e-06, + "loss": 1.6455, + "step": 237 + }, + { + "epoch": 0.24447868515665125, + "grad_norm": 2.4549131393432617, + "learning_rate": 3.8013698630136993e-06, + "loss": 1.6566, + "step": 238 + }, + { + "epoch": 0.24550590652285567, + "grad_norm": 2.4387831687927246, + "learning_rate": 3.818493150684932e-06, + "loss": 1.627, + "step": 239 + }, + { + "epoch": 0.2465331278890601, + "grad_norm": 2.136474847793579, + "learning_rate": 3.8356164383561645e-06, + "loss": 1.7286, + "step": 240 + }, + { + "epoch": 0.2465331278890601, + "eval_loss": 1.9188542366027832, + "eval_runtime": 37.9079, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 240 + }, + { + "epoch": 0.24756034925526452, + "grad_norm": 2.430297374725342, + "learning_rate": 3.852739726027397e-06, + "loss": 1.7014, + "step": 241 + }, + { + "epoch": 0.24858757062146894, + "grad_norm": 2.302480697631836, + "learning_rate": 3.869863013698631e-06, + "loss": 1.6412, + "step": 242 + }, + { + "epoch": 0.24961479198767333, + "grad_norm": 2.2905426025390625, + "learning_rate": 3.886986301369863e-06, + "loss": 1.6445, + "step": 243 + }, + { + "epoch": 0.25064201335387776, + "grad_norm": 2.55385684967041, + "learning_rate": 3.904109589041096e-06, + "loss": 1.6265, + "step": 244 + }, + { + "epoch": 0.2516692347200822, + "grad_norm": 2.3053576946258545, + "learning_rate": 3.921232876712329e-06, + "loss": 1.6338, + "step": 245 + }, + { + "epoch": 0.2526964560862866, + "grad_norm": 2.4894707202911377, + "learning_rate": 3.938356164383562e-06, + "loss": 1.5966, + "step": 246 + }, + { + "epoch": 0.253723677452491, + "grad_norm": 2.6198930740356445, + "learning_rate": 3.9554794520547955e-06, + "loss": 1.6935, + "step": 247 + }, + { + "epoch": 0.25475089881869545, + "grad_norm": 2.3548998832702637, + "learning_rate": 3.972602739726027e-06, + "loss": 1.6227, + "step": 248 + }, + { + "epoch": 0.25577812018489987, + "grad_norm": 2.535460948944092, + "learning_rate": 3.989726027397261e-06, + "loss": 1.6699, + "step": 249 + }, + { + "epoch": 0.25680534155110424, + "grad_norm": 2.3680567741394043, + "learning_rate": 4.006849315068493e-06, + "loss": 1.6666, + "step": 250 + }, + { + "epoch": 0.25680534155110424, + "eval_loss": 1.9149967432022095, + "eval_runtime": 37.9423, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 250 + }, + { + "epoch": 0.25783256291730866, + "grad_norm": 2.3914358615875244, + "learning_rate": 4.023972602739726e-06, + "loss": 1.6633, + "step": 251 + }, + { + "epoch": 0.2588597842835131, + "grad_norm": 2.4293038845062256, + "learning_rate": 4.0410958904109595e-06, + "loss": 1.6573, + "step": 252 + }, + { + "epoch": 0.2598870056497175, + "grad_norm": 2.3409900665283203, + "learning_rate": 4.058219178082192e-06, + "loss": 1.6815, + "step": 253 + }, + { + "epoch": 0.26091422701592193, + "grad_norm": 2.5807456970214844, + "learning_rate": 4.075342465753426e-06, + "loss": 1.6412, + "step": 254 + }, + { + "epoch": 0.26194144838212635, + "grad_norm": 2.317861795425415, + "learning_rate": 4.092465753424658e-06, + "loss": 1.6311, + "step": 255 + }, + { + "epoch": 0.2629686697483308, + "grad_norm": 2.231682777404785, + "learning_rate": 4.109589041095891e-06, + "loss": 1.7021, + "step": 256 + }, + { + "epoch": 0.2639958911145352, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.1267123287671235e-06, + "loss": 1.6446, + "step": 257 + }, + { + "epoch": 0.2650231124807396, + "grad_norm": 2.2619071006774902, + "learning_rate": 4.143835616438356e-06, + "loss": 1.7101, + "step": 258 + }, + { + "epoch": 0.26605033384694404, + "grad_norm": 2.461945056915283, + "learning_rate": 4.16095890410959e-06, + "loss": 1.6901, + "step": 259 + }, + { + "epoch": 0.2670775552131484, + "grad_norm": 2.516348123550415, + "learning_rate": 4.178082191780822e-06, + "loss": 1.644, + "step": 260 + }, + { + "epoch": 0.2670775552131484, + "eval_loss": 1.904897928237915, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 260 + }, + { + "epoch": 0.26810477657935283, + "grad_norm": 2.2311813831329346, + "learning_rate": 4.195205479452055e-06, + "loss": 1.6301, + "step": 261 + }, + { + "epoch": 0.26913199794555726, + "grad_norm": 2.605109214782715, + "learning_rate": 4.212328767123288e-06, + "loss": 1.6393, + "step": 262 + }, + { + "epoch": 0.2701592193117617, + "grad_norm": 2.6103134155273438, + "learning_rate": 4.229452054794521e-06, + "loss": 1.6468, + "step": 263 + }, + { + "epoch": 0.2711864406779661, + "grad_norm": 2.5538523197174072, + "learning_rate": 4.246575342465754e-06, + "loss": 1.6294, + "step": 264 + }, + { + "epoch": 0.2722136620441705, + "grad_norm": 2.5841965675354004, + "learning_rate": 4.263698630136986e-06, + "loss": 1.5837, + "step": 265 + }, + { + "epoch": 0.27324088341037495, + "grad_norm": 2.4604523181915283, + "learning_rate": 4.28082191780822e-06, + "loss": 1.7473, + "step": 266 + }, + { + "epoch": 0.27426810477657937, + "grad_norm": 2.6701765060424805, + "learning_rate": 4.297945205479452e-06, + "loss": 1.6539, + "step": 267 + }, + { + "epoch": 0.2752953261427838, + "grad_norm": 2.360069990158081, + "learning_rate": 4.315068493150685e-06, + "loss": 1.6385, + "step": 268 + }, + { + "epoch": 0.2763225475089882, + "grad_norm": 2.4922797679901123, + "learning_rate": 4.3321917808219185e-06, + "loss": 1.6586, + "step": 269 + }, + { + "epoch": 0.2773497688751926, + "grad_norm": 2.5443379878997803, + "learning_rate": 4.349315068493151e-06, + "loss": 1.6318, + "step": 270 + }, + { + "epoch": 0.2773497688751926, + "eval_loss": 1.9062609672546387, + "eval_runtime": 37.92, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 270 + }, + { + "epoch": 0.278376990241397, + "grad_norm": 2.4064481258392334, + "learning_rate": 4.3664383561643846e-06, + "loss": 1.6178, + "step": 271 + }, + { + "epoch": 0.2794042116076014, + "grad_norm": 2.681637763977051, + "learning_rate": 4.383561643835616e-06, + "loss": 1.7071, + "step": 272 + }, + { + "epoch": 0.28043143297380585, + "grad_norm": 2.5805184841156006, + "learning_rate": 4.40068493150685e-06, + "loss": 1.6553, + "step": 273 + }, + { + "epoch": 0.2814586543400103, + "grad_norm": 2.4169883728027344, + "learning_rate": 4.4178082191780825e-06, + "loss": 1.6249, + "step": 274 + }, + { + "epoch": 0.2824858757062147, + "grad_norm": 2.557570457458496, + "learning_rate": 4.434931506849315e-06, + "loss": 1.5932, + "step": 275 + }, + { + "epoch": 0.2835130970724191, + "grad_norm": 2.3808107376098633, + "learning_rate": 4.4520547945205486e-06, + "loss": 1.6377, + "step": 276 + }, + { + "epoch": 0.28454031843862354, + "grad_norm": 2.402404546737671, + "learning_rate": 4.469178082191781e-06, + "loss": 1.6285, + "step": 277 + }, + { + "epoch": 0.28556753980482796, + "grad_norm": 2.4676456451416016, + "learning_rate": 4.486301369863014e-06, + "loss": 1.6838, + "step": 278 + }, + { + "epoch": 0.28659476117103233, + "grad_norm": 2.659083366394043, + "learning_rate": 4.503424657534247e-06, + "loss": 1.5872, + "step": 279 + }, + { + "epoch": 0.28762198253723675, + "grad_norm": 2.326733350753784, + "learning_rate": 4.52054794520548e-06, + "loss": 1.6142, + "step": 280 + }, + { + "epoch": 0.28762198253723675, + "eval_loss": 1.904759407043457, + "eval_runtime": 37.9651, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 280 + }, + { + "epoch": 0.2886492039034412, + "grad_norm": 2.299504518508911, + "learning_rate": 4.537671232876713e-06, + "loss": 1.6673, + "step": 281 + }, + { + "epoch": 0.2896764252696456, + "grad_norm": 2.4932825565338135, + "learning_rate": 4.554794520547945e-06, + "loss": 1.6586, + "step": 282 + }, + { + "epoch": 0.29070364663585, + "grad_norm": 2.2599050998687744, + "learning_rate": 4.571917808219179e-06, + "loss": 1.6568, + "step": 283 + }, + { + "epoch": 0.29173086800205444, + "grad_norm": 2.435098886489868, + "learning_rate": 4.589041095890411e-06, + "loss": 1.6113, + "step": 284 + }, + { + "epoch": 0.29275808936825887, + "grad_norm": 2.539246082305908, + "learning_rate": 4.606164383561644e-06, + "loss": 1.6306, + "step": 285 + }, + { + "epoch": 0.2937853107344633, + "grad_norm": 2.596198320388794, + "learning_rate": 4.6232876712328774e-06, + "loss": 1.6473, + "step": 286 + }, + { + "epoch": 0.2948125321006677, + "grad_norm": 2.349133014678955, + "learning_rate": 4.64041095890411e-06, + "loss": 1.6042, + "step": 287 + }, + { + "epoch": 0.29583975346687214, + "grad_norm": 2.3572661876678467, + "learning_rate": 4.657534246575343e-06, + "loss": 1.6635, + "step": 288 + }, + { + "epoch": 0.2968669748330765, + "grad_norm": 2.4210307598114014, + "learning_rate": 4.674657534246575e-06, + "loss": 1.624, + "step": 289 + }, + { + "epoch": 0.2978941961992809, + "grad_norm": 2.427403688430786, + "learning_rate": 4.691780821917809e-06, + "loss": 1.6824, + "step": 290 + }, + { + "epoch": 0.2978941961992809, + "eval_loss": 1.8888994455337524, + "eval_runtime": 37.9777, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 290 + }, + { + "epoch": 0.29892141756548535, + "grad_norm": 2.534029722213745, + "learning_rate": 4.7089041095890414e-06, + "loss": 1.6456, + "step": 291 + }, + { + "epoch": 0.29994863893168977, + "grad_norm": 2.329305648803711, + "learning_rate": 4.726027397260274e-06, + "loss": 1.646, + "step": 292 + }, + { + "epoch": 0.3009758602978942, + "grad_norm": 2.561548948287964, + "learning_rate": 4.7431506849315075e-06, + "loss": 1.6511, + "step": 293 + }, + { + "epoch": 0.3020030816640986, + "grad_norm": 2.707712411880493, + "learning_rate": 4.76027397260274e-06, + "loss": 1.5984, + "step": 294 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 2.403089761734009, + "learning_rate": 4.777397260273973e-06, + "loss": 1.6164, + "step": 295 + }, + { + "epoch": 0.30405752439650746, + "grad_norm": 2.424910306930542, + "learning_rate": 4.7945205479452054e-06, + "loss": 1.6291, + "step": 296 + }, + { + "epoch": 0.3050847457627119, + "grad_norm": 2.3474302291870117, + "learning_rate": 4.811643835616439e-06, + "loss": 1.553, + "step": 297 + }, + { + "epoch": 0.3061119671289163, + "grad_norm": 2.2211227416992188, + "learning_rate": 4.8287671232876716e-06, + "loss": 1.6108, + "step": 298 + }, + { + "epoch": 0.3071391884951207, + "grad_norm": 2.495983600616455, + "learning_rate": 4.845890410958904e-06, + "loss": 1.6135, + "step": 299 + }, + { + "epoch": 0.3081664098613251, + "grad_norm": 2.525495767593384, + "learning_rate": 4.863013698630138e-06, + "loss": 1.6061, + "step": 300 + }, + { + "epoch": 0.3081664098613251, + "eval_loss": 1.8982168436050415, + "eval_runtime": 37.9848, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 300 + }, + { + "epoch": 0.3091936312275295, + "grad_norm": 2.449064254760742, + "learning_rate": 4.88013698630137e-06, + "loss": 1.5827, + "step": 301 + }, + { + "epoch": 0.31022085259373394, + "grad_norm": 2.288095712661743, + "learning_rate": 4.897260273972603e-06, + "loss": 1.704, + "step": 302 + }, + { + "epoch": 0.31124807395993837, + "grad_norm": 2.5301458835601807, + "learning_rate": 4.914383561643836e-06, + "loss": 1.6342, + "step": 303 + }, + { + "epoch": 0.3122752953261428, + "grad_norm": 2.2793986797332764, + "learning_rate": 4.931506849315069e-06, + "loss": 1.6688, + "step": 304 + }, + { + "epoch": 0.3133025166923472, + "grad_norm": 2.38783597946167, + "learning_rate": 4.948630136986302e-06, + "loss": 1.7014, + "step": 305 + }, + { + "epoch": 0.31432973805855163, + "grad_norm": 2.5224838256835938, + "learning_rate": 4.965753424657534e-06, + "loss": 1.5812, + "step": 306 + }, + { + "epoch": 0.31535695942475606, + "grad_norm": 2.434432029724121, + "learning_rate": 4.982876712328768e-06, + "loss": 1.6186, + "step": 307 + }, + { + "epoch": 0.3163841807909605, + "grad_norm": 2.3708460330963135, + "learning_rate": 5e-06, + "loss": 1.5669, + "step": 308 + }, + { + "epoch": 0.31741140215716485, + "grad_norm": 2.5524990558624268, + "learning_rate": 4.998096688237534e-06, + "loss": 1.6688, + "step": 309 + }, + { + "epoch": 0.31843862352336927, + "grad_norm": 2.322441577911377, + "learning_rate": 4.996193376475067e-06, + "loss": 1.5571, + "step": 310 + }, + { + "epoch": 0.31843862352336927, + "eval_loss": 1.8914265632629395, + "eval_runtime": 37.6505, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 310 + }, + { + "epoch": 0.3194658448895737, + "grad_norm": 2.5248687267303467, + "learning_rate": 4.9942900647126005e-06, + "loss": 1.5585, + "step": 311 + }, + { + "epoch": 0.3204930662557781, + "grad_norm": 2.2464914321899414, + "learning_rate": 4.992386752950134e-06, + "loss": 1.5921, + "step": 312 + }, + { + "epoch": 0.32152028762198254, + "grad_norm": 2.5355331897735596, + "learning_rate": 4.990483441187667e-06, + "loss": 1.6263, + "step": 313 + }, + { + "epoch": 0.32254750898818696, + "grad_norm": 2.5611143112182617, + "learning_rate": 4.9885801294252e-06, + "loss": 1.6267, + "step": 314 + }, + { + "epoch": 0.3235747303543914, + "grad_norm": 2.364356517791748, + "learning_rate": 4.986676817662733e-06, + "loss": 1.5847, + "step": 315 + }, + { + "epoch": 0.3246019517205958, + "grad_norm": 2.455165147781372, + "learning_rate": 4.9847735059002665e-06, + "loss": 1.661, + "step": 316 + }, + { + "epoch": 0.3256291730868002, + "grad_norm": 2.4312195777893066, + "learning_rate": 4.9828701941378e-06, + "loss": 1.5995, + "step": 317 + }, + { + "epoch": 0.3266563944530046, + "grad_norm": 2.4081499576568604, + "learning_rate": 4.980966882375333e-06, + "loss": 1.5756, + "step": 318 + }, + { + "epoch": 0.327683615819209, + "grad_norm": 2.2608511447906494, + "learning_rate": 4.979063570612867e-06, + "loss": 1.543, + "step": 319 + }, + { + "epoch": 0.32871083718541344, + "grad_norm": 2.6551144123077393, + "learning_rate": 4.9771602588504e-06, + "loss": 1.6467, + "step": 320 + }, + { + "epoch": 0.32871083718541344, + "eval_loss": 1.89058518409729, + "eval_runtime": 37.8192, + "eval_samples_per_second": 26.442, + "eval_steps_per_second": 1.111, + "step": 320 + }, + { + "epoch": 0.32973805855161786, + "grad_norm": 2.3897273540496826, + "learning_rate": 4.975256947087933e-06, + "loss": 1.5607, + "step": 321 + }, + { + "epoch": 0.3307652799178223, + "grad_norm": 2.3869469165802, + "learning_rate": 4.973353635325467e-06, + "loss": 1.6079, + "step": 322 + }, + { + "epoch": 0.3317925012840267, + "grad_norm": 2.159193515777588, + "learning_rate": 4.971450323563e-06, + "loss": 1.5475, + "step": 323 + }, + { + "epoch": 0.33281972265023113, + "grad_norm": 2.3069653511047363, + "learning_rate": 4.969547011800533e-06, + "loss": 1.616, + "step": 324 + }, + { + "epoch": 0.33384694401643555, + "grad_norm": 2.3131422996520996, + "learning_rate": 4.967643700038067e-06, + "loss": 1.6106, + "step": 325 + }, + { + "epoch": 0.33487416538264, + "grad_norm": 2.4398200511932373, + "learning_rate": 4.9657403882756e-06, + "loss": 1.5779, + "step": 326 + }, + { + "epoch": 0.3359013867488444, + "grad_norm": 2.278024435043335, + "learning_rate": 4.9638370765131335e-06, + "loss": 1.5724, + "step": 327 + }, + { + "epoch": 0.33692860811504877, + "grad_norm": 2.4989173412323, + "learning_rate": 4.961933764750667e-06, + "loss": 1.6337, + "step": 328 + }, + { + "epoch": 0.3379558294812532, + "grad_norm": 2.46319580078125, + "learning_rate": 4.9600304529882e-06, + "loss": 1.6519, + "step": 329 + }, + { + "epoch": 0.3389830508474576, + "grad_norm": 2.2703049182891846, + "learning_rate": 4.9581271412257335e-06, + "loss": 1.6081, + "step": 330 + }, + { + "epoch": 0.3389830508474576, + "eval_loss": 1.8755545616149902, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 330 + }, + { + "epoch": 0.34001027221366203, + "grad_norm": 2.502196788787842, + "learning_rate": 4.956223829463266e-06, + "loss": 1.5221, + "step": 331 + }, + { + "epoch": 0.34103749357986646, + "grad_norm": 2.284280776977539, + "learning_rate": 4.954320517700799e-06, + "loss": 1.6268, + "step": 332 + }, + { + "epoch": 0.3420647149460709, + "grad_norm": 2.3551881313323975, + "learning_rate": 4.952417205938333e-06, + "loss": 1.6078, + "step": 333 + }, + { + "epoch": 0.3430919363122753, + "grad_norm": 2.343780755996704, + "learning_rate": 4.950513894175866e-06, + "loss": 1.591, + "step": 334 + }, + { + "epoch": 0.3441191576784797, + "grad_norm": 2.3524727821350098, + "learning_rate": 4.9486105824133995e-06, + "loss": 1.605, + "step": 335 + }, + { + "epoch": 0.34514637904468415, + "grad_norm": 2.349896192550659, + "learning_rate": 4.946707270650933e-06, + "loss": 1.6106, + "step": 336 + }, + { + "epoch": 0.34617360041088857, + "grad_norm": 2.5154407024383545, + "learning_rate": 4.944803958888466e-06, + "loss": 1.6429, + "step": 337 + }, + { + "epoch": 0.34720082177709294, + "grad_norm": 2.2692947387695312, + "learning_rate": 4.942900647126e-06, + "loss": 1.5865, + "step": 338 + }, + { + "epoch": 0.34822804314329736, + "grad_norm": 2.2672805786132812, + "learning_rate": 4.940997335363533e-06, + "loss": 1.5464, + "step": 339 + }, + { + "epoch": 0.3492552645095018, + "grad_norm": 2.5147039890289307, + "learning_rate": 4.939094023601066e-06, + "loss": 1.5837, + "step": 340 + }, + { + "epoch": 0.3492552645095018, + "eval_loss": 1.8672455549240112, + "eval_runtime": 37.9644, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 340 + }, + { + "epoch": 0.3502824858757062, + "grad_norm": 2.3416433334350586, + "learning_rate": 4.9371907118386e-06, + "loss": 1.5937, + "step": 341 + }, + { + "epoch": 0.35130970724191063, + "grad_norm": 2.3044960498809814, + "learning_rate": 4.935287400076133e-06, + "loss": 1.6132, + "step": 342 + }, + { + "epoch": 0.35233692860811505, + "grad_norm": 2.4360430240631104, + "learning_rate": 4.933384088313666e-06, + "loss": 1.5731, + "step": 343 + }, + { + "epoch": 0.3533641499743195, + "grad_norm": 2.392050266265869, + "learning_rate": 4.9314807765512e-06, + "loss": 1.5098, + "step": 344 + }, + { + "epoch": 0.3543913713405239, + "grad_norm": 2.3815808296203613, + "learning_rate": 4.929577464788733e-06, + "loss": 1.5827, + "step": 345 + }, + { + "epoch": 0.3554185927067283, + "grad_norm": 2.423077344894409, + "learning_rate": 4.9276741530262665e-06, + "loss": 1.5857, + "step": 346 + }, + { + "epoch": 0.35644581407293274, + "grad_norm": 2.5535898208618164, + "learning_rate": 4.9257708412638e-06, + "loss": 1.5884, + "step": 347 + }, + { + "epoch": 0.3574730354391371, + "grad_norm": 2.2665445804595947, + "learning_rate": 4.923867529501332e-06, + "loss": 1.5561, + "step": 348 + }, + { + "epoch": 0.35850025680534153, + "grad_norm": 2.3357481956481934, + "learning_rate": 4.921964217738866e-06, + "loss": 1.6444, + "step": 349 + }, + { + "epoch": 0.35952747817154596, + "grad_norm": 2.4537885189056396, + "learning_rate": 4.920060905976399e-06, + "loss": 1.6262, + "step": 350 + }, + { + "epoch": 0.35952747817154596, + "eval_loss": 1.8621584177017212, + "eval_runtime": 38.0027, + "eval_samples_per_second": 26.314, + "eval_steps_per_second": 1.105, + "step": 350 + }, + { + "epoch": 0.3605546995377504, + "grad_norm": 2.379754066467285, + "learning_rate": 4.9181575942139325e-06, + "loss": 1.6227, + "step": 351 + }, + { + "epoch": 0.3615819209039548, + "grad_norm": 2.432905435562134, + "learning_rate": 4.916254282451466e-06, + "loss": 1.6554, + "step": 352 + }, + { + "epoch": 0.3626091422701592, + "grad_norm": 2.34236216545105, + "learning_rate": 4.914350970688999e-06, + "loss": 1.5613, + "step": 353 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.3019444942474365, + "learning_rate": 4.9124476589265325e-06, + "loss": 1.5535, + "step": 354 + }, + { + "epoch": 0.36466358500256807, + "grad_norm": 2.3354644775390625, + "learning_rate": 4.910544347164066e-06, + "loss": 1.6384, + "step": 355 + }, + { + "epoch": 0.3656908063687725, + "grad_norm": 2.5154788494110107, + "learning_rate": 4.908641035401599e-06, + "loss": 1.6986, + "step": 356 + }, + { + "epoch": 0.3667180277349769, + "grad_norm": 2.2707178592681885, + "learning_rate": 4.906737723639133e-06, + "loss": 1.5705, + "step": 357 + }, + { + "epoch": 0.3677452491011813, + "grad_norm": 2.3058297634124756, + "learning_rate": 4.904834411876665e-06, + "loss": 1.6033, + "step": 358 + }, + { + "epoch": 0.3687724704673857, + "grad_norm": 2.3923370838165283, + "learning_rate": 4.902931100114199e-06, + "loss": 1.5333, + "step": 359 + }, + { + "epoch": 0.3697996918335901, + "grad_norm": 2.4023196697235107, + "learning_rate": 4.901027788351733e-06, + "loss": 1.5775, + "step": 360 + }, + { + "epoch": 0.3697996918335901, + "eval_loss": 1.867856740951538, + "eval_runtime": 38.0125, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 360 + }, + { + "epoch": 0.37082691319979455, + "grad_norm": 2.4903581142425537, + "learning_rate": 4.899124476589266e-06, + "loss": 1.5491, + "step": 361 + }, + { + "epoch": 0.371854134565999, + "grad_norm": 2.556683301925659, + "learning_rate": 4.8972211648267994e-06, + "loss": 1.5967, + "step": 362 + }, + { + "epoch": 0.3728813559322034, + "grad_norm": 2.3017377853393555, + "learning_rate": 4.895317853064333e-06, + "loss": 1.5661, + "step": 363 + }, + { + "epoch": 0.3739085772984078, + "grad_norm": 2.286191940307617, + "learning_rate": 4.893414541301866e-06, + "loss": 1.5679, + "step": 364 + }, + { + "epoch": 0.37493579866461224, + "grad_norm": 2.260812520980835, + "learning_rate": 4.891511229539399e-06, + "loss": 1.5965, + "step": 365 + }, + { + "epoch": 0.37596302003081666, + "grad_norm": 2.4569408893585205, + "learning_rate": 4.889607917776932e-06, + "loss": 1.5923, + "step": 366 + }, + { + "epoch": 0.37699024139702103, + "grad_norm": 2.571552038192749, + "learning_rate": 4.887704606014465e-06, + "loss": 1.6021, + "step": 367 + }, + { + "epoch": 0.37801746276322545, + "grad_norm": 2.287754774093628, + "learning_rate": 4.885801294251999e-06, + "loss": 1.5304, + "step": 368 + }, + { + "epoch": 0.3790446841294299, + "grad_norm": 2.317784070968628, + "learning_rate": 4.883897982489532e-06, + "loss": 1.6233, + "step": 369 + }, + { + "epoch": 0.3800719054956343, + "grad_norm": 2.290601968765259, + "learning_rate": 4.8819946707270655e-06, + "loss": 1.624, + "step": 370 + }, + { + "epoch": 0.3800719054956343, + "eval_loss": 1.8604528903961182, + "eval_runtime": 38.0176, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 370 + }, + { + "epoch": 0.3810991268618387, + "grad_norm": 2.28519868850708, + "learning_rate": 4.880091358964599e-06, + "loss": 1.5286, + "step": 371 + }, + { + "epoch": 0.38212634822804314, + "grad_norm": 2.271810293197632, + "learning_rate": 4.878188047202132e-06, + "loss": 1.6879, + "step": 372 + }, + { + "epoch": 0.38315356959424757, + "grad_norm": 2.391848564147949, + "learning_rate": 4.876284735439666e-06, + "loss": 1.6458, + "step": 373 + }, + { + "epoch": 0.384180790960452, + "grad_norm": 2.3787031173706055, + "learning_rate": 4.874381423677199e-06, + "loss": 1.5738, + "step": 374 + }, + { + "epoch": 0.3852080123266564, + "grad_norm": 2.404219150543213, + "learning_rate": 4.8724781119147315e-06, + "loss": 1.6184, + "step": 375 + }, + { + "epoch": 0.38623523369286084, + "grad_norm": 2.372661590576172, + "learning_rate": 4.870574800152265e-06, + "loss": 1.5316, + "step": 376 + }, + { + "epoch": 0.3872624550590652, + "grad_norm": 2.4397103786468506, + "learning_rate": 4.868671488389798e-06, + "loss": 1.5605, + "step": 377 + }, + { + "epoch": 0.3882896764252696, + "grad_norm": 2.371006727218628, + "learning_rate": 4.8667681766273316e-06, + "loss": 1.5184, + "step": 378 + }, + { + "epoch": 0.38931689779147405, + "grad_norm": 2.2621910572052, + "learning_rate": 4.864864864864866e-06, + "loss": 1.5788, + "step": 379 + }, + { + "epoch": 0.39034411915767847, + "grad_norm": 2.2888615131378174, + "learning_rate": 4.862961553102399e-06, + "loss": 1.6244, + "step": 380 + }, + { + "epoch": 0.39034411915767847, + "eval_loss": 1.8586912155151367, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 380 + }, + { + "epoch": 0.3913713405238829, + "grad_norm": 2.4439220428466797, + "learning_rate": 4.8610582413399325e-06, + "loss": 1.5749, + "step": 381 + }, + { + "epoch": 0.3923985618900873, + "grad_norm": 2.2423763275146484, + "learning_rate": 4.859154929577465e-06, + "loss": 1.5738, + "step": 382 + }, + { + "epoch": 0.39342578325629174, + "grad_norm": 2.267467737197876, + "learning_rate": 4.857251617814998e-06, + "loss": 1.6212, + "step": 383 + }, + { + "epoch": 0.39445300462249616, + "grad_norm": 2.3680620193481445, + "learning_rate": 4.855348306052532e-06, + "loss": 1.6129, + "step": 384 + }, + { + "epoch": 0.3954802259887006, + "grad_norm": 2.3782172203063965, + "learning_rate": 4.853444994290065e-06, + "loss": 1.5605, + "step": 385 + }, + { + "epoch": 0.396507447354905, + "grad_norm": 2.5745720863342285, + "learning_rate": 4.8515416825275985e-06, + "loss": 1.634, + "step": 386 + }, + { + "epoch": 0.3975346687211094, + "grad_norm": 2.38212513923645, + "learning_rate": 4.849638370765132e-06, + "loss": 1.5671, + "step": 387 + }, + { + "epoch": 0.3985618900873138, + "grad_norm": 2.3395450115203857, + "learning_rate": 4.847735059002665e-06, + "loss": 1.5088, + "step": 388 + }, + { + "epoch": 0.3995891114535182, + "grad_norm": 2.474440574645996, + "learning_rate": 4.8458317472401985e-06, + "loss": 1.5997, + "step": 389 + }, + { + "epoch": 0.40061633281972264, + "grad_norm": 2.429133176803589, + "learning_rate": 4.843928435477732e-06, + "loss": 1.5611, + "step": 390 + }, + { + "epoch": 0.40061633281972264, + "eval_loss": 1.8613603115081787, + "eval_runtime": 37.9862, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 390 + }, + { + "epoch": 0.40164355418592707, + "grad_norm": 2.403207778930664, + "learning_rate": 4.842025123715265e-06, + "loss": 1.5976, + "step": 391 + }, + { + "epoch": 0.4026707755521315, + "grad_norm": 2.383018732070923, + "learning_rate": 4.840121811952798e-06, + "loss": 1.5791, + "step": 392 + }, + { + "epoch": 0.4036979969183359, + "grad_norm": 2.5303306579589844, + "learning_rate": 4.838218500190331e-06, + "loss": 1.6035, + "step": 393 + }, + { + "epoch": 0.40472521828454033, + "grad_norm": 2.391075372695923, + "learning_rate": 4.8363151884278645e-06, + "loss": 1.5795, + "step": 394 + }, + { + "epoch": 0.40575243965074476, + "grad_norm": 2.396934747695923, + "learning_rate": 4.834411876665398e-06, + "loss": 1.5965, + "step": 395 + }, + { + "epoch": 0.4067796610169492, + "grad_norm": 2.4631423950195312, + "learning_rate": 4.832508564902931e-06, + "loss": 1.5372, + "step": 396 + }, + { + "epoch": 0.40780688238315355, + "grad_norm": 2.3457326889038086, + "learning_rate": 4.830605253140465e-06, + "loss": 1.5729, + "step": 397 + }, + { + "epoch": 0.40883410374935797, + "grad_norm": 2.3763277530670166, + "learning_rate": 4.828701941377998e-06, + "loss": 1.5309, + "step": 398 + }, + { + "epoch": 0.4098613251155624, + "grad_norm": 2.442439556121826, + "learning_rate": 4.826798629615531e-06, + "loss": 1.5465, + "step": 399 + }, + { + "epoch": 0.4108885464817668, + "grad_norm": 2.3464581966400146, + "learning_rate": 4.824895317853065e-06, + "loss": 1.5665, + "step": 400 + }, + { + "epoch": 0.4108885464817668, + "eval_loss": 1.8671027421951294, + "eval_runtime": 38.0166, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 400 + }, + { + "epoch": 0.41191576784797124, + "grad_norm": 2.5441887378692627, + "learning_rate": 4.822992006090598e-06, + "loss": 1.5232, + "step": 401 + }, + { + "epoch": 0.41294298921417566, + "grad_norm": 2.4234976768493652, + "learning_rate": 4.821088694328131e-06, + "loss": 1.5569, + "step": 402 + }, + { + "epoch": 0.4139702105803801, + "grad_norm": 2.3603928089141846, + "learning_rate": 4.819185382565665e-06, + "loss": 1.612, + "step": 403 + }, + { + "epoch": 0.4149974319465845, + "grad_norm": 2.337862491607666, + "learning_rate": 4.817282070803198e-06, + "loss": 1.5882, + "step": 404 + }, + { + "epoch": 0.41602465331278893, + "grad_norm": 2.3805742263793945, + "learning_rate": 4.8153787590407315e-06, + "loss": 1.5204, + "step": 405 + }, + { + "epoch": 0.41705187467899335, + "grad_norm": 2.3335204124450684, + "learning_rate": 4.813475447278265e-06, + "loss": 1.6041, + "step": 406 + }, + { + "epoch": 0.4180790960451977, + "grad_norm": 2.406137466430664, + "learning_rate": 4.811572135515798e-06, + "loss": 1.5348, + "step": 407 + }, + { + "epoch": 0.41910631741140214, + "grad_norm": 2.2823917865753174, + "learning_rate": 4.809668823753332e-06, + "loss": 1.5351, + "step": 408 + }, + { + "epoch": 0.42013353877760656, + "grad_norm": 2.3577091693878174, + "learning_rate": 4.807765511990864e-06, + "loss": 1.5733, + "step": 409 + }, + { + "epoch": 0.421160760143811, + "grad_norm": 2.536691665649414, + "learning_rate": 4.8058622002283975e-06, + "loss": 1.6153, + "step": 410 + }, + { + "epoch": 0.421160760143811, + "eval_loss": 1.8613901138305664, + "eval_runtime": 37.6689, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 410 + }, + { + "epoch": 0.4221879815100154, + "grad_norm": 2.4020118713378906, + "learning_rate": 4.803958888465931e-06, + "loss": 1.5722, + "step": 411 + }, + { + "epoch": 0.42321520287621983, + "grad_norm": 2.5284810066223145, + "learning_rate": 4.802055576703464e-06, + "loss": 1.6546, + "step": 412 + }, + { + "epoch": 0.42424242424242425, + "grad_norm": 2.270094156265259, + "learning_rate": 4.8001522649409976e-06, + "loss": 1.5345, + "step": 413 + }, + { + "epoch": 0.4252696456086287, + "grad_norm": 2.241417407989502, + "learning_rate": 4.798248953178531e-06, + "loss": 1.5167, + "step": 414 + }, + { + "epoch": 0.4262968669748331, + "grad_norm": 2.3733017444610596, + "learning_rate": 4.796345641416064e-06, + "loss": 1.5532, + "step": 415 + }, + { + "epoch": 0.42732408834103747, + "grad_norm": 2.452587842941284, + "learning_rate": 4.794442329653598e-06, + "loss": 1.6136, + "step": 416 + }, + { + "epoch": 0.4283513097072419, + "grad_norm": 2.417217969894409, + "learning_rate": 4.792539017891131e-06, + "loss": 1.5398, + "step": 417 + }, + { + "epoch": 0.4293785310734463, + "grad_norm": 2.4502453804016113, + "learning_rate": 4.790635706128664e-06, + "loss": 1.5741, + "step": 418 + }, + { + "epoch": 0.43040575243965074, + "grad_norm": 2.2996811866760254, + "learning_rate": 4.788732394366197e-06, + "loss": 1.511, + "step": 419 + }, + { + "epoch": 0.43143297380585516, + "grad_norm": 2.3573555946350098, + "learning_rate": 4.786829082603731e-06, + "loss": 1.5117, + "step": 420 + }, + { + "epoch": 0.43143297380585516, + "eval_loss": 1.8690649271011353, + "eval_runtime": 37.7813, + "eval_samples_per_second": 26.468, + "eval_steps_per_second": 1.112, + "step": 420 + }, + { + "epoch": 0.4324601951720596, + "grad_norm": 2.37233567237854, + "learning_rate": 4.7849257708412645e-06, + "loss": 1.5873, + "step": 421 + }, + { + "epoch": 0.433487416538264, + "grad_norm": 2.146909713745117, + "learning_rate": 4.783022459078798e-06, + "loss": 1.5291, + "step": 422 + }, + { + "epoch": 0.4345146379044684, + "grad_norm": 2.5237205028533936, + "learning_rate": 4.781119147316331e-06, + "loss": 1.5545, + "step": 423 + }, + { + "epoch": 0.43554185927067285, + "grad_norm": 2.356842279434204, + "learning_rate": 4.7792158355538645e-06, + "loss": 1.5635, + "step": 424 + }, + { + "epoch": 0.43656908063687727, + "grad_norm": 2.1963610649108887, + "learning_rate": 4.777312523791398e-06, + "loss": 1.5355, + "step": 425 + }, + { + "epoch": 0.43759630200308164, + "grad_norm": 2.6425886154174805, + "learning_rate": 4.77540921202893e-06, + "loss": 1.5803, + "step": 426 + }, + { + "epoch": 0.43862352336928606, + "grad_norm": 2.4072272777557373, + "learning_rate": 4.773505900266464e-06, + "loss": 1.5649, + "step": 427 + }, + { + "epoch": 0.4396507447354905, + "grad_norm": 2.592374801635742, + "learning_rate": 4.771602588503997e-06, + "loss": 1.5693, + "step": 428 + }, + { + "epoch": 0.4406779661016949, + "grad_norm": 2.4534225463867188, + "learning_rate": 4.7696992767415305e-06, + "loss": 1.591, + "step": 429 + }, + { + "epoch": 0.44170518746789933, + "grad_norm": 2.434457778930664, + "learning_rate": 4.767795964979064e-06, + "loss": 1.5991, + "step": 430 + }, + { + "epoch": 0.44170518746789933, + "eval_loss": 1.8584965467453003, + "eval_runtime": 37.8854, + "eval_samples_per_second": 26.395, + "eval_steps_per_second": 1.109, + "step": 430 + }, + { + "epoch": 0.44273240883410375, + "grad_norm": 2.3150746822357178, + "learning_rate": 4.765892653216597e-06, + "loss": 1.6032, + "step": 431 + }, + { + "epoch": 0.4437596302003082, + "grad_norm": 2.3040223121643066, + "learning_rate": 4.763989341454131e-06, + "loss": 1.5115, + "step": 432 + }, + { + "epoch": 0.4447868515665126, + "grad_norm": 2.413007974624634, + "learning_rate": 4.762086029691664e-06, + "loss": 1.5305, + "step": 433 + }, + { + "epoch": 0.445814072932717, + "grad_norm": 2.526289463043213, + "learning_rate": 4.760182717929197e-06, + "loss": 1.564, + "step": 434 + }, + { + "epoch": 0.44684129429892144, + "grad_norm": 2.4122776985168457, + "learning_rate": 4.758279406166731e-06, + "loss": 1.4918, + "step": 435 + }, + { + "epoch": 0.4478685156651258, + "grad_norm": 2.474274158477783, + "learning_rate": 4.756376094404263e-06, + "loss": 1.5583, + "step": 436 + }, + { + "epoch": 0.44889573703133023, + "grad_norm": 2.5322892665863037, + "learning_rate": 4.7544727826417966e-06, + "loss": 1.6081, + "step": 437 + }, + { + "epoch": 0.44992295839753466, + "grad_norm": 2.391660690307617, + "learning_rate": 4.75256947087933e-06, + "loss": 1.5628, + "step": 438 + }, + { + "epoch": 0.4509501797637391, + "grad_norm": 2.426671028137207, + "learning_rate": 4.750666159116863e-06, + "loss": 1.531, + "step": 439 + }, + { + "epoch": 0.4519774011299435, + "grad_norm": 2.7095694541931152, + "learning_rate": 4.7487628473543975e-06, + "loss": 1.4811, + "step": 440 + }, + { + "epoch": 0.4519774011299435, + "eval_loss": 1.8348888158798218, + "eval_runtime": 37.9114, + "eval_samples_per_second": 26.377, + "eval_steps_per_second": 1.108, + "step": 440 + }, + { + "epoch": 0.4530046224961479, + "grad_norm": 2.5425240993499756, + "learning_rate": 4.746859535591931e-06, + "loss": 1.5453, + "step": 441 + }, + { + "epoch": 0.45403184386235235, + "grad_norm": 2.6848669052124023, + "learning_rate": 4.744956223829464e-06, + "loss": 1.526, + "step": 442 + }, + { + "epoch": 0.45505906522855677, + "grad_norm": 2.634068489074707, + "learning_rate": 4.743052912066997e-06, + "loss": 1.5748, + "step": 443 + }, + { + "epoch": 0.4560862865947612, + "grad_norm": 2.492030382156372, + "learning_rate": 4.74114960030453e-06, + "loss": 1.5858, + "step": 444 + }, + { + "epoch": 0.4571135079609656, + "grad_norm": 2.787740707397461, + "learning_rate": 4.7392462885420635e-06, + "loss": 1.4774, + "step": 445 + }, + { + "epoch": 0.45814072932717, + "grad_norm": 2.395914077758789, + "learning_rate": 4.737342976779597e-06, + "loss": 1.558, + "step": 446 + }, + { + "epoch": 0.4591679506933744, + "grad_norm": 2.6642563343048096, + "learning_rate": 4.73543966501713e-06, + "loss": 1.5558, + "step": 447 + }, + { + "epoch": 0.4601951720595788, + "grad_norm": 2.619943857192993, + "learning_rate": 4.7335363532546636e-06, + "loss": 1.5442, + "step": 448 + }, + { + "epoch": 0.46122239342578325, + "grad_norm": 2.355133056640625, + "learning_rate": 4.731633041492197e-06, + "loss": 1.5613, + "step": 449 + }, + { + "epoch": 0.4622496147919877, + "grad_norm": 2.6572225093841553, + "learning_rate": 4.72972972972973e-06, + "loss": 1.5951, + "step": 450 + }, + { + "epoch": 0.4622496147919877, + "eval_loss": 1.83199143409729, + "eval_runtime": 37.8996, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 450 + }, + { + "epoch": 0.4632768361581921, + "grad_norm": 2.692873001098633, + "learning_rate": 4.727826417967264e-06, + "loss": 1.5531, + "step": 451 + }, + { + "epoch": 0.4643040575243965, + "grad_norm": 2.3903489112854004, + "learning_rate": 4.725923106204797e-06, + "loss": 1.5411, + "step": 452 + }, + { + "epoch": 0.46533127889060094, + "grad_norm": 2.453263521194458, + "learning_rate": 4.72401979444233e-06, + "loss": 1.5632, + "step": 453 + }, + { + "epoch": 0.46635850025680536, + "grad_norm": 2.5637760162353516, + "learning_rate": 4.722116482679863e-06, + "loss": 1.5629, + "step": 454 + }, + { + "epoch": 0.46738572162300973, + "grad_norm": 2.387033700942993, + "learning_rate": 4.720213170917396e-06, + "loss": 1.5786, + "step": 455 + }, + { + "epoch": 0.46841294298921415, + "grad_norm": 2.4563071727752686, + "learning_rate": 4.71830985915493e-06, + "loss": 1.5321, + "step": 456 + }, + { + "epoch": 0.4694401643554186, + "grad_norm": 2.607451915740967, + "learning_rate": 4.716406547392463e-06, + "loss": 1.5958, + "step": 457 + }, + { + "epoch": 0.470467385721623, + "grad_norm": 2.4846854209899902, + "learning_rate": 4.714503235629996e-06, + "loss": 1.5205, + "step": 458 + }, + { + "epoch": 0.4714946070878274, + "grad_norm": 2.377629041671753, + "learning_rate": 4.71259992386753e-06, + "loss": 1.5403, + "step": 459 + }, + { + "epoch": 0.47252182845403184, + "grad_norm": 2.4470160007476807, + "learning_rate": 4.710696612105063e-06, + "loss": 1.5456, + "step": 460 + }, + { + "epoch": 0.47252182845403184, + "eval_loss": 1.83836829662323, + "eval_runtime": 37.88, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 460 + }, + { + "epoch": 0.47354904982023627, + "grad_norm": 2.383418560028076, + "learning_rate": 4.708793300342596e-06, + "loss": 1.5468, + "step": 461 + }, + { + "epoch": 0.4745762711864407, + "grad_norm": 2.3672895431518555, + "learning_rate": 4.70688998858013e-06, + "loss": 1.5313, + "step": 462 + }, + { + "epoch": 0.4756034925526451, + "grad_norm": 2.415717124938965, + "learning_rate": 4.704986676817663e-06, + "loss": 1.5645, + "step": 463 + }, + { + "epoch": 0.47663071391884954, + "grad_norm": 2.4690067768096924, + "learning_rate": 4.7030833650551965e-06, + "loss": 1.5285, + "step": 464 + }, + { + "epoch": 0.4776579352850539, + "grad_norm": 2.3845672607421875, + "learning_rate": 4.70118005329273e-06, + "loss": 1.596, + "step": 465 + }, + { + "epoch": 0.4786851566512583, + "grad_norm": 2.2801802158355713, + "learning_rate": 4.699276741530263e-06, + "loss": 1.5777, + "step": 466 + }, + { + "epoch": 0.47971237801746275, + "grad_norm": 2.4668991565704346, + "learning_rate": 4.697373429767797e-06, + "loss": 1.5566, + "step": 467 + }, + { + "epoch": 0.48073959938366717, + "grad_norm": 2.446078062057495, + "learning_rate": 4.69547011800533e-06, + "loss": 1.5664, + "step": 468 + }, + { + "epoch": 0.4817668207498716, + "grad_norm": 2.2310681343078613, + "learning_rate": 4.693566806242863e-06, + "loss": 1.5425, + "step": 469 + }, + { + "epoch": 0.482794042116076, + "grad_norm": 2.3938229084014893, + "learning_rate": 4.691663494480397e-06, + "loss": 1.576, + "step": 470 + }, + { + "epoch": 0.482794042116076, + "eval_loss": 1.83778977394104, + "eval_runtime": 37.9276, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 470 + }, + { + "epoch": 0.48382126348228044, + "grad_norm": 2.23827862739563, + "learning_rate": 4.689760182717929e-06, + "loss": 1.4763, + "step": 471 + }, + { + "epoch": 0.48484848484848486, + "grad_norm": 2.3148930072784424, + "learning_rate": 4.6878568709554626e-06, + "loss": 1.5386, + "step": 472 + }, + { + "epoch": 0.4858757062146893, + "grad_norm": 2.407036781311035, + "learning_rate": 4.685953559192996e-06, + "loss": 1.5576, + "step": 473 + }, + { + "epoch": 0.4869029275808937, + "grad_norm": 2.4675374031066895, + "learning_rate": 4.684050247430529e-06, + "loss": 1.4931, + "step": 474 + }, + { + "epoch": 0.4879301489470981, + "grad_norm": 2.4593775272369385, + "learning_rate": 4.682146935668063e-06, + "loss": 1.5348, + "step": 475 + }, + { + "epoch": 0.4889573703133025, + "grad_norm": 2.4169375896453857, + "learning_rate": 4.680243623905596e-06, + "loss": 1.5614, + "step": 476 + }, + { + "epoch": 0.4899845916795069, + "grad_norm": 2.6177303791046143, + "learning_rate": 4.678340312143129e-06, + "loss": 1.5138, + "step": 477 + }, + { + "epoch": 0.49101181304571134, + "grad_norm": 2.6418113708496094, + "learning_rate": 4.676437000380663e-06, + "loss": 1.6377, + "step": 478 + }, + { + "epoch": 0.49203903441191577, + "grad_norm": 2.338491439819336, + "learning_rate": 4.674533688618196e-06, + "loss": 1.5682, + "step": 479 + }, + { + "epoch": 0.4930662557781202, + "grad_norm": 2.5901622772216797, + "learning_rate": 4.6726303768557295e-06, + "loss": 1.4899, + "step": 480 + }, + { + "epoch": 0.4930662557781202, + "eval_loss": 1.8238886594772339, + "eval_runtime": 37.9605, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 480 + }, + { + "epoch": 0.4940934771443246, + "grad_norm": 2.470827102661133, + "learning_rate": 4.670727065093263e-06, + "loss": 1.5016, + "step": 481 + }, + { + "epoch": 0.49512069851052903, + "grad_norm": 2.371201992034912, + "learning_rate": 4.668823753330796e-06, + "loss": 1.5399, + "step": 482 + }, + { + "epoch": 0.49614791987673346, + "grad_norm": 2.532802104949951, + "learning_rate": 4.6669204415683295e-06, + "loss": 1.5985, + "step": 483 + }, + { + "epoch": 0.4971751412429379, + "grad_norm": 2.6294727325439453, + "learning_rate": 4.665017129805863e-06, + "loss": 1.5802, + "step": 484 + }, + { + "epoch": 0.49820236260914225, + "grad_norm": 2.2164878845214844, + "learning_rate": 4.663113818043396e-06, + "loss": 1.5384, + "step": 485 + }, + { + "epoch": 0.49922958397534667, + "grad_norm": 2.311232805252075, + "learning_rate": 4.66121050628093e-06, + "loss": 1.5785, + "step": 486 + }, + { + "epoch": 0.5002568053415511, + "grad_norm": 2.2916736602783203, + "learning_rate": 4.659307194518463e-06, + "loss": 1.5401, + "step": 487 + }, + { + "epoch": 0.5012840267077555, + "grad_norm": 2.3978800773620605, + "learning_rate": 4.6574038827559955e-06, + "loss": 1.5619, + "step": 488 + }, + { + "epoch": 0.50231124807396, + "grad_norm": 2.438917398452759, + "learning_rate": 4.655500570993529e-06, + "loss": 1.5824, + "step": 489 + }, + { + "epoch": 0.5033384694401644, + "grad_norm": 2.528862237930298, + "learning_rate": 4.653597259231062e-06, + "loss": 1.4802, + "step": 490 + }, + { + "epoch": 0.5033384694401644, + "eval_loss": 1.83504056930542, + "eval_runtime": 37.9357, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 490 + }, + { + "epoch": 0.5043656908063687, + "grad_norm": 2.54522442817688, + "learning_rate": 4.651693947468596e-06, + "loss": 1.6186, + "step": 491 + }, + { + "epoch": 0.5053929121725732, + "grad_norm": 2.5037031173706055, + "learning_rate": 4.649790635706129e-06, + "loss": 1.5159, + "step": 492 + }, + { + "epoch": 0.5064201335387776, + "grad_norm": 2.4600465297698975, + "learning_rate": 4.647887323943662e-06, + "loss": 1.5378, + "step": 493 + }, + { + "epoch": 0.507447354904982, + "grad_norm": 2.393695831298828, + "learning_rate": 4.645984012181196e-06, + "loss": 1.4485, + "step": 494 + }, + { + "epoch": 0.5084745762711864, + "grad_norm": 2.488626718521118, + "learning_rate": 4.644080700418729e-06, + "loss": 1.57, + "step": 495 + }, + { + "epoch": 0.5095017976373909, + "grad_norm": 2.409858465194702, + "learning_rate": 4.642177388656262e-06, + "loss": 1.5343, + "step": 496 + }, + { + "epoch": 0.5105290190035953, + "grad_norm": 2.5607573986053467, + "learning_rate": 4.640274076893796e-06, + "loss": 1.5825, + "step": 497 + }, + { + "epoch": 0.5115562403697997, + "grad_norm": 2.5953354835510254, + "learning_rate": 4.638370765131328e-06, + "loss": 1.5129, + "step": 498 + }, + { + "epoch": 0.5125834617360041, + "grad_norm": 2.3321897983551025, + "learning_rate": 4.636467453368862e-06, + "loss": 1.4972, + "step": 499 + }, + { + "epoch": 0.5136106831022085, + "grad_norm": 2.332052707672119, + "learning_rate": 4.634564141606396e-06, + "loss": 1.567, + "step": 500 + }, + { + "epoch": 0.5136106831022085, + "eval_loss": 1.824558138847351, + "eval_runtime": 37.9157, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 500 + }, + { + "epoch": 0.514637904468413, + "grad_norm": 2.431809902191162, + "learning_rate": 4.632660829843929e-06, + "loss": 1.494, + "step": 501 + }, + { + "epoch": 0.5156651258346173, + "grad_norm": 2.7118983268737793, + "learning_rate": 4.630757518081463e-06, + "loss": 1.6289, + "step": 502 + }, + { + "epoch": 0.5166923472008218, + "grad_norm": 2.3549370765686035, + "learning_rate": 4.628854206318996e-06, + "loss": 1.5363, + "step": 503 + }, + { + "epoch": 0.5177195685670262, + "grad_norm": 2.3343000411987305, + "learning_rate": 4.626950894556529e-06, + "loss": 1.5749, + "step": 504 + }, + { + "epoch": 0.5187467899332306, + "grad_norm": 2.4281466007232666, + "learning_rate": 4.625047582794062e-06, + "loss": 1.4958, + "step": 505 + }, + { + "epoch": 0.519774011299435, + "grad_norm": 2.3858845233917236, + "learning_rate": 4.623144271031595e-06, + "loss": 1.5031, + "step": 506 + }, + { + "epoch": 0.5208012326656395, + "grad_norm": 2.3714635372161865, + "learning_rate": 4.6212409592691286e-06, + "loss": 1.5808, + "step": 507 + }, + { + "epoch": 0.5218284540318439, + "grad_norm": 2.5290520191192627, + "learning_rate": 4.619337647506662e-06, + "loss": 1.541, + "step": 508 + }, + { + "epoch": 0.5228556753980482, + "grad_norm": 2.5521368980407715, + "learning_rate": 4.617434335744195e-06, + "loss": 1.4992, + "step": 509 + }, + { + "epoch": 0.5238828967642527, + "grad_norm": 2.436588764190674, + "learning_rate": 4.615531023981729e-06, + "loss": 1.5283, + "step": 510 + }, + { + "epoch": 0.5238828967642527, + "eval_loss": 1.8220165967941284, + "eval_runtime": 37.6123, + "eval_samples_per_second": 26.587, + "eval_steps_per_second": 1.117, + "step": 510 + }, + { + "epoch": 0.5249101181304571, + "grad_norm": 2.5240023136138916, + "learning_rate": 4.613627712219262e-06, + "loss": 1.5426, + "step": 511 + }, + { + "epoch": 0.5259373394966615, + "grad_norm": 2.5292775630950928, + "learning_rate": 4.611724400456795e-06, + "loss": 1.5411, + "step": 512 + }, + { + "epoch": 0.5269645608628659, + "grad_norm": 2.3622210025787354, + "learning_rate": 4.609821088694329e-06, + "loss": 1.5083, + "step": 513 + }, + { + "epoch": 0.5279917822290704, + "grad_norm": 2.545267105102539, + "learning_rate": 4.607917776931862e-06, + "loss": 1.5541, + "step": 514 + }, + { + "epoch": 0.5290190035952748, + "grad_norm": 2.3417961597442627, + "learning_rate": 4.606014465169395e-06, + "loss": 1.5686, + "step": 515 + }, + { + "epoch": 0.5300462249614792, + "grad_norm": 2.2782580852508545, + "learning_rate": 4.604111153406928e-06, + "loss": 1.4761, + "step": 516 + }, + { + "epoch": 0.5310734463276836, + "grad_norm": 2.4437527656555176, + "learning_rate": 4.602207841644461e-06, + "loss": 1.5818, + "step": 517 + }, + { + "epoch": 0.5321006676938881, + "grad_norm": 2.3594319820404053, + "learning_rate": 4.600304529881995e-06, + "loss": 1.5556, + "step": 518 + }, + { + "epoch": 0.5331278890600925, + "grad_norm": 2.325979471206665, + "learning_rate": 4.598401218119528e-06, + "loss": 1.5726, + "step": 519 + }, + { + "epoch": 0.5341551104262968, + "grad_norm": 2.2765676975250244, + "learning_rate": 4.5964979063570614e-06, + "loss": 1.5151, + "step": 520 + }, + { + "epoch": 0.5341551104262968, + "eval_loss": 1.825669288635254, + "eval_runtime": 37.7491, + "eval_samples_per_second": 26.491, + "eval_steps_per_second": 1.113, + "step": 520 + }, + { + "epoch": 0.5351823317925013, + "grad_norm": 2.357647657394409, + "learning_rate": 4.594594594594596e-06, + "loss": 1.5505, + "step": 521 + }, + { + "epoch": 0.5362095531587057, + "grad_norm": 2.3652896881103516, + "learning_rate": 4.592691282832128e-06, + "loss": 1.5621, + "step": 522 + }, + { + "epoch": 0.5372367745249101, + "grad_norm": 2.234096050262451, + "learning_rate": 4.5907879710696615e-06, + "loss": 1.4851, + "step": 523 + }, + { + "epoch": 0.5382639958911145, + "grad_norm": 2.3414828777313232, + "learning_rate": 4.588884659307195e-06, + "loss": 1.5594, + "step": 524 + }, + { + "epoch": 0.539291217257319, + "grad_norm": 2.3934929370880127, + "learning_rate": 4.586981347544728e-06, + "loss": 1.4994, + "step": 525 + }, + { + "epoch": 0.5403184386235234, + "grad_norm": 2.400479555130005, + "learning_rate": 4.585078035782262e-06, + "loss": 1.5346, + "step": 526 + }, + { + "epoch": 0.5413456599897278, + "grad_norm": 2.3817873001098633, + "learning_rate": 4.583174724019795e-06, + "loss": 1.4803, + "step": 527 + }, + { + "epoch": 0.5423728813559322, + "grad_norm": 2.4503185749053955, + "learning_rate": 4.581271412257328e-06, + "loss": 1.4711, + "step": 528 + }, + { + "epoch": 0.5434001027221366, + "grad_norm": 2.452578067779541, + "learning_rate": 4.579368100494862e-06, + "loss": 1.5645, + "step": 529 + }, + { + "epoch": 0.544427324088341, + "grad_norm": 2.3278186321258545, + "learning_rate": 4.577464788732395e-06, + "loss": 1.5428, + "step": 530 + }, + { + "epoch": 0.544427324088341, + "eval_loss": 1.8242785930633545, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 530 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 2.5192365646362305, + "learning_rate": 4.575561476969928e-06, + "loss": 1.5896, + "step": 531 + }, + { + "epoch": 0.5464817668207499, + "grad_norm": 2.308701276779175, + "learning_rate": 4.573658165207461e-06, + "loss": 1.5518, + "step": 532 + }, + { + "epoch": 0.5475089881869543, + "grad_norm": 2.4221107959747314, + "learning_rate": 4.571754853444994e-06, + "loss": 1.5387, + "step": 533 + }, + { + "epoch": 0.5485362095531587, + "grad_norm": 2.457606077194214, + "learning_rate": 4.569851541682528e-06, + "loss": 1.4997, + "step": 534 + }, + { + "epoch": 0.5495634309193631, + "grad_norm": 2.3406503200531006, + "learning_rate": 4.567948229920061e-06, + "loss": 1.4761, + "step": 535 + }, + { + "epoch": 0.5505906522855676, + "grad_norm": 2.555712938308716, + "learning_rate": 4.566044918157594e-06, + "loss": 1.5114, + "step": 536 + }, + { + "epoch": 0.551617873651772, + "grad_norm": 2.373013734817505, + "learning_rate": 4.564141606395128e-06, + "loss": 1.5516, + "step": 537 + }, + { + "epoch": 0.5526450950179764, + "grad_norm": 2.3915255069732666, + "learning_rate": 4.562238294632661e-06, + "loss": 1.5145, + "step": 538 + }, + { + "epoch": 0.5536723163841808, + "grad_norm": 2.3272323608398438, + "learning_rate": 4.5603349828701945e-06, + "loss": 1.5472, + "step": 539 + }, + { + "epoch": 0.5546995377503852, + "grad_norm": 2.416072130203247, + "learning_rate": 4.558431671107728e-06, + "loss": 1.5842, + "step": 540 + }, + { + "epoch": 0.5546995377503852, + "eval_loss": 1.825427532196045, + "eval_runtime": 37.9905, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 540 + }, + { + "epoch": 0.5557267591165896, + "grad_norm": 2.3225271701812744, + "learning_rate": 4.556528359345261e-06, + "loss": 1.5188, + "step": 541 + }, + { + "epoch": 0.556753980482794, + "grad_norm": 2.2387030124664307, + "learning_rate": 4.5546250475827946e-06, + "loss": 1.4778, + "step": 542 + }, + { + "epoch": 0.5577812018489985, + "grad_norm": 2.3054494857788086, + "learning_rate": 4.552721735820328e-06, + "loss": 1.5229, + "step": 543 + }, + { + "epoch": 0.5588084232152029, + "grad_norm": 2.4956626892089844, + "learning_rate": 4.550818424057861e-06, + "loss": 1.5367, + "step": 544 + }, + { + "epoch": 0.5598356445814073, + "grad_norm": 2.454986810684204, + "learning_rate": 4.548915112295395e-06, + "loss": 1.5444, + "step": 545 + }, + { + "epoch": 0.5608628659476117, + "grad_norm": 2.466810703277588, + "learning_rate": 4.547011800532928e-06, + "loss": 1.4914, + "step": 546 + }, + { + "epoch": 0.5618900873138162, + "grad_norm": 2.51777982711792, + "learning_rate": 4.545108488770461e-06, + "loss": 1.5775, + "step": 547 + }, + { + "epoch": 0.5629173086800205, + "grad_norm": 2.340421199798584, + "learning_rate": 4.543205177007995e-06, + "loss": 1.5185, + "step": 548 + }, + { + "epoch": 0.5639445300462249, + "grad_norm": 2.376772880554199, + "learning_rate": 4.541301865245527e-06, + "loss": 1.5588, + "step": 549 + }, + { + "epoch": 0.5649717514124294, + "grad_norm": 2.4004335403442383, + "learning_rate": 4.539398553483061e-06, + "loss": 1.5283, + "step": 550 + }, + { + "epoch": 0.5649717514124294, + "eval_loss": 1.8294963836669922, + "eval_runtime": 37.9391, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 550 + }, + { + "epoch": 0.5659989727786338, + "grad_norm": 2.424459934234619, + "learning_rate": 4.537495241720594e-06, + "loss": 1.5576, + "step": 551 + }, + { + "epoch": 0.5670261941448382, + "grad_norm": 2.4413998126983643, + "learning_rate": 4.535591929958127e-06, + "loss": 1.5444, + "step": 552 + }, + { + "epoch": 0.5680534155110426, + "grad_norm": 2.408897876739502, + "learning_rate": 4.533688618195661e-06, + "loss": 1.5269, + "step": 553 + }, + { + "epoch": 0.5690806368772471, + "grad_norm": 2.471024990081787, + "learning_rate": 4.531785306433194e-06, + "loss": 1.5109, + "step": 554 + }, + { + "epoch": 0.5701078582434514, + "grad_norm": 2.374882459640503, + "learning_rate": 4.529881994670727e-06, + "loss": 1.5205, + "step": 555 + }, + { + "epoch": 0.5711350796096559, + "grad_norm": 2.4166030883789062, + "learning_rate": 4.527978682908261e-06, + "loss": 1.5276, + "step": 556 + }, + { + "epoch": 0.5721623009758603, + "grad_norm": 2.2681620121002197, + "learning_rate": 4.526075371145794e-06, + "loss": 1.5374, + "step": 557 + }, + { + "epoch": 0.5731895223420647, + "grad_norm": 2.3933286666870117, + "learning_rate": 4.5241720593833275e-06, + "loss": 1.5349, + "step": 558 + }, + { + "epoch": 0.5742167437082691, + "grad_norm": 2.4086451530456543, + "learning_rate": 4.52226874762086e-06, + "loss": 1.5512, + "step": 559 + }, + { + "epoch": 0.5752439650744735, + "grad_norm": 2.559021472930908, + "learning_rate": 4.520365435858393e-06, + "loss": 1.5789, + "step": 560 + }, + { + "epoch": 0.5752439650744735, + "eval_loss": 1.8243234157562256, + "eval_runtime": 37.934, + "eval_samples_per_second": 26.362, + "eval_steps_per_second": 1.107, + "step": 560 + }, + { + "epoch": 0.576271186440678, + "grad_norm": 2.483494520187378, + "learning_rate": 4.518462124095928e-06, + "loss": 1.5348, + "step": 561 + }, + { + "epoch": 0.5772984078068824, + "grad_norm": 2.4039993286132812, + "learning_rate": 4.516558812333461e-06, + "loss": 1.5201, + "step": 562 + }, + { + "epoch": 0.5783256291730868, + "grad_norm": 2.5080959796905518, + "learning_rate": 4.514655500570994e-06, + "loss": 1.5864, + "step": 563 + }, + { + "epoch": 0.5793528505392912, + "grad_norm": 2.620887279510498, + "learning_rate": 4.512752188808528e-06, + "loss": 1.5901, + "step": 564 + }, + { + "epoch": 0.5803800719054957, + "grad_norm": 2.5691096782684326, + "learning_rate": 4.510848877046061e-06, + "loss": 1.5506, + "step": 565 + }, + { + "epoch": 0.5814072932717, + "grad_norm": 2.3777835369110107, + "learning_rate": 4.5089455652835936e-06, + "loss": 1.5149, + "step": 566 + }, + { + "epoch": 0.5824345146379045, + "grad_norm": 2.2791666984558105, + "learning_rate": 4.507042253521127e-06, + "loss": 1.4771, + "step": 567 + }, + { + "epoch": 0.5834617360041089, + "grad_norm": 2.4953627586364746, + "learning_rate": 4.50513894175866e-06, + "loss": 1.556, + "step": 568 + }, + { + "epoch": 0.5844889573703133, + "grad_norm": 2.3563215732574463, + "learning_rate": 4.503235629996194e-06, + "loss": 1.549, + "step": 569 + }, + { + "epoch": 0.5855161787365177, + "grad_norm": 2.49676513671875, + "learning_rate": 4.501332318233727e-06, + "loss": 1.513, + "step": 570 + }, + { + "epoch": 0.5855161787365177, + "eval_loss": 1.8353043794631958, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 570 + }, + { + "epoch": 0.5865434001027221, + "grad_norm": 2.385275363922119, + "learning_rate": 4.49942900647126e-06, + "loss": 1.5622, + "step": 571 + }, + { + "epoch": 0.5875706214689266, + "grad_norm": 2.3089914321899414, + "learning_rate": 4.497525694708794e-06, + "loss": 1.5426, + "step": 572 + }, + { + "epoch": 0.588597842835131, + "grad_norm": 2.307323455810547, + "learning_rate": 4.495622382946327e-06, + "loss": 1.5088, + "step": 573 + }, + { + "epoch": 0.5896250642013354, + "grad_norm": 2.597288131713867, + "learning_rate": 4.4937190711838605e-06, + "loss": 1.5788, + "step": 574 + }, + { + "epoch": 0.5906522855675398, + "grad_norm": 2.3923256397247314, + "learning_rate": 4.491815759421394e-06, + "loss": 1.5038, + "step": 575 + }, + { + "epoch": 0.5916795069337443, + "grad_norm": 2.4199156761169434, + "learning_rate": 4.489912447658926e-06, + "loss": 1.5057, + "step": 576 + }, + { + "epoch": 0.5927067282999486, + "grad_norm": 2.5643393993377686, + "learning_rate": 4.48800913589646e-06, + "loss": 1.5274, + "step": 577 + }, + { + "epoch": 0.593733949666153, + "grad_norm": 2.5854485034942627, + "learning_rate": 4.486105824133993e-06, + "loss": 1.5341, + "step": 578 + }, + { + "epoch": 0.5947611710323575, + "grad_norm": 2.5363597869873047, + "learning_rate": 4.4842025123715264e-06, + "loss": 1.4953, + "step": 579 + }, + { + "epoch": 0.5957883923985619, + "grad_norm": 2.3880693912506104, + "learning_rate": 4.48229920060906e-06, + "loss": 1.4868, + "step": 580 + }, + { + "epoch": 0.5957883923985619, + "eval_loss": 1.8251410722732544, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 580 + }, + { + "epoch": 0.5968156137647663, + "grad_norm": 2.5272583961486816, + "learning_rate": 4.480395888846594e-06, + "loss": 1.5851, + "step": 581 + }, + { + "epoch": 0.5978428351309707, + "grad_norm": 2.4496943950653076, + "learning_rate": 4.478492577084127e-06, + "loss": 1.4896, + "step": 582 + }, + { + "epoch": 0.5988700564971752, + "grad_norm": 2.3996055126190186, + "learning_rate": 4.47658926532166e-06, + "loss": 1.5253, + "step": 583 + }, + { + "epoch": 0.5998972778633795, + "grad_norm": 2.5964202880859375, + "learning_rate": 4.474685953559193e-06, + "loss": 1.6294, + "step": 584 + }, + { + "epoch": 0.600924499229584, + "grad_norm": 2.5567739009857178, + "learning_rate": 4.472782641796727e-06, + "loss": 1.4834, + "step": 585 + }, + { + "epoch": 0.6019517205957884, + "grad_norm": 2.440291166305542, + "learning_rate": 4.47087933003426e-06, + "loss": 1.4966, + "step": 586 + }, + { + "epoch": 0.6029789419619929, + "grad_norm": 2.525763988494873, + "learning_rate": 4.468976018271793e-06, + "loss": 1.5005, + "step": 587 + }, + { + "epoch": 0.6040061633281972, + "grad_norm": 2.4822962284088135, + "learning_rate": 4.467072706509327e-06, + "loss": 1.5358, + "step": 588 + }, + { + "epoch": 0.6050333846944016, + "grad_norm": 2.591149091720581, + "learning_rate": 4.46516939474686e-06, + "loss": 1.4975, + "step": 589 + }, + { + "epoch": 0.6060606060606061, + "grad_norm": 2.4884238243103027, + "learning_rate": 4.463266082984393e-06, + "loss": 1.6107, + "step": 590 + }, + { + "epoch": 0.6060606060606061, + "eval_loss": 1.8383466005325317, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 590 + }, + { + "epoch": 0.6070878274268104, + "grad_norm": 2.4965732097625732, + "learning_rate": 4.461362771221927e-06, + "loss": 1.4986, + "step": 591 + }, + { + "epoch": 0.6081150487930149, + "grad_norm": 2.544792413711548, + "learning_rate": 4.45945945945946e-06, + "loss": 1.5067, + "step": 592 + }, + { + "epoch": 0.6091422701592193, + "grad_norm": 2.5726170539855957, + "learning_rate": 4.457556147696993e-06, + "loss": 1.5861, + "step": 593 + }, + { + "epoch": 0.6101694915254238, + "grad_norm": 2.493027448654175, + "learning_rate": 4.455652835934526e-06, + "loss": 1.5857, + "step": 594 + }, + { + "epoch": 0.6111967128916281, + "grad_norm": 2.470771074295044, + "learning_rate": 4.453749524172059e-06, + "loss": 1.5419, + "step": 595 + }, + { + "epoch": 0.6122239342578326, + "grad_norm": 2.568052053451538, + "learning_rate": 4.451846212409593e-06, + "loss": 1.5391, + "step": 596 + }, + { + "epoch": 0.613251155624037, + "grad_norm": 2.4346213340759277, + "learning_rate": 4.449942900647126e-06, + "loss": 1.4761, + "step": 597 + }, + { + "epoch": 0.6142783769902413, + "grad_norm": 2.418133497238159, + "learning_rate": 4.4480395888846595e-06, + "loss": 1.5383, + "step": 598 + }, + { + "epoch": 0.6153055983564458, + "grad_norm": 2.4795567989349365, + "learning_rate": 4.446136277122193e-06, + "loss": 1.4777, + "step": 599 + }, + { + "epoch": 0.6163328197226502, + "grad_norm": 2.5122742652893066, + "learning_rate": 4.444232965359726e-06, + "loss": 1.5353, + "step": 600 + }, + { + "epoch": 0.6163328197226502, + "eval_loss": 1.8325241804122925, + "eval_runtime": 37.9546, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 600 + }, + { + "epoch": 0.6173600410888547, + "grad_norm": 2.604067325592041, + "learning_rate": 4.4423296535972596e-06, + "loss": 1.489, + "step": 601 + }, + { + "epoch": 0.618387262455059, + "grad_norm": 2.3741753101348877, + "learning_rate": 4.440426341834793e-06, + "loss": 1.532, + "step": 602 + }, + { + "epoch": 0.6194144838212635, + "grad_norm": 2.5373036861419678, + "learning_rate": 4.438523030072326e-06, + "loss": 1.561, + "step": 603 + }, + { + "epoch": 0.6204417051874679, + "grad_norm": 2.2958099842071533, + "learning_rate": 4.43661971830986e-06, + "loss": 1.5025, + "step": 604 + }, + { + "epoch": 0.6214689265536724, + "grad_norm": 2.417006015777588, + "learning_rate": 4.434716406547393e-06, + "loss": 1.4977, + "step": 605 + }, + { + "epoch": 0.6224961479198767, + "grad_norm": 2.335296392440796, + "learning_rate": 4.432813094784926e-06, + "loss": 1.5504, + "step": 606 + }, + { + "epoch": 0.6235233692860811, + "grad_norm": 2.4086828231811523, + "learning_rate": 4.43090978302246e-06, + "loss": 1.5217, + "step": 607 + }, + { + "epoch": 0.6245505906522856, + "grad_norm": 2.3816819190979004, + "learning_rate": 4.429006471259993e-06, + "loss": 1.5126, + "step": 608 + }, + { + "epoch": 0.6255778120184899, + "grad_norm": 2.307671546936035, + "learning_rate": 4.4271031594975265e-06, + "loss": 1.4876, + "step": 609 + }, + { + "epoch": 0.6266050333846944, + "grad_norm": 2.355837821960449, + "learning_rate": 4.425199847735059e-06, + "loss": 1.4827, + "step": 610 + }, + { + "epoch": 0.6266050333846944, + "eval_loss": 1.8269455432891846, + "eval_runtime": 37.6459, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 610 + }, + { + "epoch": 0.6276322547508988, + "grad_norm": 2.558631658554077, + "learning_rate": 4.423296535972592e-06, + "loss": 1.5172, + "step": 611 + }, + { + "epoch": 0.6286594761171033, + "grad_norm": 2.3170175552368164, + "learning_rate": 4.421393224210126e-06, + "loss": 1.4784, + "step": 612 + }, + { + "epoch": 0.6296866974833076, + "grad_norm": 2.4064524173736572, + "learning_rate": 4.419489912447659e-06, + "loss": 1.5616, + "step": 613 + }, + { + "epoch": 0.6307139188495121, + "grad_norm": 2.3719160556793213, + "learning_rate": 4.4175866006851924e-06, + "loss": 1.5292, + "step": 614 + }, + { + "epoch": 0.6317411402157165, + "grad_norm": 2.3808202743530273, + "learning_rate": 4.415683288922726e-06, + "loss": 1.5122, + "step": 615 + }, + { + "epoch": 0.632768361581921, + "grad_norm": 2.4208037853240967, + "learning_rate": 4.413779977160259e-06, + "loss": 1.5088, + "step": 616 + }, + { + "epoch": 0.6337955829481253, + "grad_norm": 2.5558979511260986, + "learning_rate": 4.4118766653977925e-06, + "loss": 1.5678, + "step": 617 + }, + { + "epoch": 0.6348228043143297, + "grad_norm": 2.443908452987671, + "learning_rate": 4.409973353635326e-06, + "loss": 1.5234, + "step": 618 + }, + { + "epoch": 0.6358500256805342, + "grad_norm": 2.420408010482788, + "learning_rate": 4.408070041872859e-06, + "loss": 1.4708, + "step": 619 + }, + { + "epoch": 0.6368772470467385, + "grad_norm": 2.3884904384613037, + "learning_rate": 4.406166730110393e-06, + "loss": 1.4821, + "step": 620 + }, + { + "epoch": 0.6368772470467385, + "eval_loss": 1.8217289447784424, + "eval_runtime": 37.7936, + "eval_samples_per_second": 26.459, + "eval_steps_per_second": 1.111, + "step": 620 + }, + { + "epoch": 0.637904468412943, + "grad_norm": 2.313720703125, + "learning_rate": 4.404263418347925e-06, + "loss": 1.4962, + "step": 621 + }, + { + "epoch": 0.6389316897791474, + "grad_norm": 2.3173866271972656, + "learning_rate": 4.402360106585459e-06, + "loss": 1.4857, + "step": 622 + }, + { + "epoch": 0.6399589111453519, + "grad_norm": 2.541757822036743, + "learning_rate": 4.400456794822993e-06, + "loss": 1.5177, + "step": 623 + }, + { + "epoch": 0.6409861325115562, + "grad_norm": 2.4382357597351074, + "learning_rate": 4.398553483060526e-06, + "loss": 1.4718, + "step": 624 + }, + { + "epoch": 0.6420133538777607, + "grad_norm": 2.2988059520721436, + "learning_rate": 4.396650171298059e-06, + "loss": 1.4883, + "step": 625 + }, + { + "epoch": 0.6430405752439651, + "grad_norm": 2.4817495346069336, + "learning_rate": 4.394746859535593e-06, + "loss": 1.4823, + "step": 626 + }, + { + "epoch": 0.6440677966101694, + "grad_norm": 2.4578585624694824, + "learning_rate": 4.392843547773125e-06, + "loss": 1.4882, + "step": 627 + }, + { + "epoch": 0.6450950179763739, + "grad_norm": 2.184972047805786, + "learning_rate": 4.390940236010659e-06, + "loss": 1.514, + "step": 628 + }, + { + "epoch": 0.6461222393425783, + "grad_norm": 2.4040279388427734, + "learning_rate": 4.389036924248192e-06, + "loss": 1.5343, + "step": 629 + }, + { + "epoch": 0.6471494607087828, + "grad_norm": 2.3676459789276123, + "learning_rate": 4.387133612485725e-06, + "loss": 1.5007, + "step": 630 + }, + { + "epoch": 0.6471494607087828, + "eval_loss": 1.8106164932250977, + "eval_runtime": 37.9383, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 630 + }, + { + "epoch": 0.6481766820749871, + "grad_norm": 2.234689235687256, + "learning_rate": 4.385230300723259e-06, + "loss": 1.5064, + "step": 631 + }, + { + "epoch": 0.6492039034411916, + "grad_norm": 2.281742572784424, + "learning_rate": 4.383326988960792e-06, + "loss": 1.4777, + "step": 632 + }, + { + "epoch": 0.650231124807396, + "grad_norm": 2.4314322471618652, + "learning_rate": 4.3814236771983255e-06, + "loss": 1.5331, + "step": 633 + }, + { + "epoch": 0.6512583461736005, + "grad_norm": 2.4092588424682617, + "learning_rate": 4.379520365435859e-06, + "loss": 1.5122, + "step": 634 + }, + { + "epoch": 0.6522855675398048, + "grad_norm": 2.517362117767334, + "learning_rate": 4.377617053673392e-06, + "loss": 1.52, + "step": 635 + }, + { + "epoch": 0.6533127889060092, + "grad_norm": 2.4784109592437744, + "learning_rate": 4.3757137419109256e-06, + "loss": 1.5771, + "step": 636 + }, + { + "epoch": 0.6543400102722137, + "grad_norm": 2.32181715965271, + "learning_rate": 4.373810430148459e-06, + "loss": 1.4718, + "step": 637 + }, + { + "epoch": 0.655367231638418, + "grad_norm": 2.5146260261535645, + "learning_rate": 4.3719071183859914e-06, + "loss": 1.5307, + "step": 638 + }, + { + "epoch": 0.6563944530046225, + "grad_norm": 2.2364494800567627, + "learning_rate": 4.370003806623525e-06, + "loss": 1.4803, + "step": 639 + }, + { + "epoch": 0.6574216743708269, + "grad_norm": 2.3735806941986084, + "learning_rate": 4.368100494861058e-06, + "loss": 1.5249, + "step": 640 + }, + { + "epoch": 0.6574216743708269, + "eval_loss": 1.8134880065917969, + "eval_runtime": 38.0376, + "eval_samples_per_second": 26.29, + "eval_steps_per_second": 1.104, + "step": 640 + }, + { + "epoch": 0.6584488957370314, + "grad_norm": 2.4600815773010254, + "learning_rate": 4.3661971830985915e-06, + "loss": 1.4893, + "step": 641 + }, + { + "epoch": 0.6594761171032357, + "grad_norm": 2.4106783866882324, + "learning_rate": 4.364293871336126e-06, + "loss": 1.5163, + "step": 642 + }, + { + "epoch": 0.6605033384694402, + "grad_norm": 2.276624917984009, + "learning_rate": 4.362390559573659e-06, + "loss": 1.482, + "step": 643 + }, + { + "epoch": 0.6615305598356446, + "grad_norm": 2.336341381072998, + "learning_rate": 4.3604872478111925e-06, + "loss": 1.4152, + "step": 644 + }, + { + "epoch": 0.662557781201849, + "grad_norm": 2.3329105377197266, + "learning_rate": 4.358583936048725e-06, + "loss": 1.5411, + "step": 645 + }, + { + "epoch": 0.6635850025680534, + "grad_norm": 2.2877161502838135, + "learning_rate": 4.356680624286258e-06, + "loss": 1.4997, + "step": 646 + }, + { + "epoch": 0.6646122239342578, + "grad_norm": 2.410820484161377, + "learning_rate": 4.354777312523792e-06, + "loss": 1.5446, + "step": 647 + }, + { + "epoch": 0.6656394453004623, + "grad_norm": 2.403038501739502, + "learning_rate": 4.352874000761325e-06, + "loss": 1.4916, + "step": 648 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 2.3261778354644775, + "learning_rate": 4.3509706889988584e-06, + "loss": 1.5561, + "step": 649 + }, + { + "epoch": 0.6676938880328711, + "grad_norm": 2.4512174129486084, + "learning_rate": 4.349067377236392e-06, + "loss": 1.556, + "step": 650 + }, + { + "epoch": 0.6676938880328711, + "eval_loss": 1.8238279819488525, + "eval_runtime": 37.9988, + "eval_samples_per_second": 26.317, + "eval_steps_per_second": 1.105, + "step": 650 + }, + { + "epoch": 0.6687211093990755, + "grad_norm": 2.4523823261260986, + "learning_rate": 4.347164065473925e-06, + "loss": 1.5676, + "step": 651 + }, + { + "epoch": 0.66974833076528, + "grad_norm": 2.3770365715026855, + "learning_rate": 4.3452607537114585e-06, + "loss": 1.4815, + "step": 652 + }, + { + "epoch": 0.6707755521314843, + "grad_norm": 2.2440760135650635, + "learning_rate": 4.343357441948992e-06, + "loss": 1.5141, + "step": 653 + }, + { + "epoch": 0.6718027734976888, + "grad_norm": 2.3341948986053467, + "learning_rate": 4.341454130186525e-06, + "loss": 1.4859, + "step": 654 + }, + { + "epoch": 0.6728299948638932, + "grad_norm": 2.507354497909546, + "learning_rate": 4.339550818424058e-06, + "loss": 1.4734, + "step": 655 + }, + { + "epoch": 0.6738572162300975, + "grad_norm": 2.395954132080078, + "learning_rate": 4.337647506661591e-06, + "loss": 1.5069, + "step": 656 + }, + { + "epoch": 0.674884437596302, + "grad_norm": 2.362211227416992, + "learning_rate": 4.3357441948991245e-06, + "loss": 1.5299, + "step": 657 + }, + { + "epoch": 0.6759116589625064, + "grad_norm": 2.460008144378662, + "learning_rate": 4.333840883136658e-06, + "loss": 1.4795, + "step": 658 + }, + { + "epoch": 0.6769388803287109, + "grad_norm": 2.3578217029571533, + "learning_rate": 4.331937571374191e-06, + "loss": 1.4906, + "step": 659 + }, + { + "epoch": 0.6779661016949152, + "grad_norm": 2.398843288421631, + "learning_rate": 4.3300342596117246e-06, + "loss": 1.4986, + "step": 660 + }, + { + "epoch": 0.6779661016949152, + "eval_loss": 1.8206881284713745, + "eval_runtime": 38.044, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 660 + }, + { + "epoch": 0.6789933230611197, + "grad_norm": 2.3587260246276855, + "learning_rate": 4.328130947849258e-06, + "loss": 1.461, + "step": 661 + }, + { + "epoch": 0.6800205444273241, + "grad_norm": 2.311952590942383, + "learning_rate": 4.326227636086791e-06, + "loss": 1.5041, + "step": 662 + }, + { + "epoch": 0.6810477657935285, + "grad_norm": 2.376188278198242, + "learning_rate": 4.324324324324325e-06, + "loss": 1.5042, + "step": 663 + }, + { + "epoch": 0.6820749871597329, + "grad_norm": 2.5807230472564697, + "learning_rate": 4.322421012561858e-06, + "loss": 1.4935, + "step": 664 + }, + { + "epoch": 0.6831022085259374, + "grad_norm": 2.368128538131714, + "learning_rate": 4.320517700799391e-06, + "loss": 1.5851, + "step": 665 + }, + { + "epoch": 0.6841294298921418, + "grad_norm": 2.4517524242401123, + "learning_rate": 4.318614389036925e-06, + "loss": 1.5528, + "step": 666 + }, + { + "epoch": 0.6851566512583461, + "grad_norm": 2.5436811447143555, + "learning_rate": 4.316711077274458e-06, + "loss": 1.4633, + "step": 667 + }, + { + "epoch": 0.6861838726245506, + "grad_norm": 2.4287915229797363, + "learning_rate": 4.3148077655119915e-06, + "loss": 1.5454, + "step": 668 + }, + { + "epoch": 0.687211093990755, + "grad_norm": 2.386528730392456, + "learning_rate": 4.312904453749525e-06, + "loss": 1.5116, + "step": 669 + }, + { + "epoch": 0.6882383153569595, + "grad_norm": 2.6602251529693604, + "learning_rate": 4.311001141987058e-06, + "loss": 1.5199, + "step": 670 + }, + { + "epoch": 0.6882383153569595, + "eval_loss": 1.8252663612365723, + "eval_runtime": 38.0791, + "eval_samples_per_second": 26.261, + "eval_steps_per_second": 1.103, + "step": 670 + }, + { + "epoch": 0.6892655367231638, + "grad_norm": 2.351128339767456, + "learning_rate": 4.3090978302245916e-06, + "loss": 1.507, + "step": 671 + }, + { + "epoch": 0.6902927580893683, + "grad_norm": 2.393441677093506, + "learning_rate": 4.307194518462124e-06, + "loss": 1.579, + "step": 672 + }, + { + "epoch": 0.6913199794555727, + "grad_norm": 2.3730578422546387, + "learning_rate": 4.3052912066996574e-06, + "loss": 1.5446, + "step": 673 + }, + { + "epoch": 0.6923472008217771, + "grad_norm": 2.2522945404052734, + "learning_rate": 4.303387894937191e-06, + "loss": 1.4705, + "step": 674 + }, + { + "epoch": 0.6933744221879815, + "grad_norm": 2.384563446044922, + "learning_rate": 4.301484583174724e-06, + "loss": 1.4636, + "step": 675 + }, + { + "epoch": 0.6944016435541859, + "grad_norm": 2.2879116535186768, + "learning_rate": 4.2995812714122575e-06, + "loss": 1.4914, + "step": 676 + }, + { + "epoch": 0.6954288649203904, + "grad_norm": 2.494788646697998, + "learning_rate": 4.297677959649791e-06, + "loss": 1.5126, + "step": 677 + }, + { + "epoch": 0.6964560862865947, + "grad_norm": 2.700012683868408, + "learning_rate": 4.295774647887324e-06, + "loss": 1.5362, + "step": 678 + }, + { + "epoch": 0.6974833076527992, + "grad_norm": 2.5130302906036377, + "learning_rate": 4.293871336124858e-06, + "loss": 1.5075, + "step": 679 + }, + { + "epoch": 0.6985105290190036, + "grad_norm": 2.486236572265625, + "learning_rate": 4.291968024362391e-06, + "loss": 1.5338, + "step": 680 + }, + { + "epoch": 0.6985105290190036, + "eval_loss": 1.8229657411575317, + "eval_runtime": 38.0642, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 680 + }, + { + "epoch": 0.699537750385208, + "grad_norm": 2.441683530807495, + "learning_rate": 4.290064712599924e-06, + "loss": 1.537, + "step": 681 + }, + { + "epoch": 0.7005649717514124, + "grad_norm": 2.396798610687256, + "learning_rate": 4.288161400837458e-06, + "loss": 1.4553, + "step": 682 + }, + { + "epoch": 0.7015921931176169, + "grad_norm": 2.3636221885681152, + "learning_rate": 4.286258089074991e-06, + "loss": 1.5138, + "step": 683 + }, + { + "epoch": 0.7026194144838213, + "grad_norm": 2.4584708213806152, + "learning_rate": 4.284354777312524e-06, + "loss": 1.4725, + "step": 684 + }, + { + "epoch": 0.7036466358500256, + "grad_norm": 2.850332498550415, + "learning_rate": 4.282451465550058e-06, + "loss": 1.5075, + "step": 685 + }, + { + "epoch": 0.7046738572162301, + "grad_norm": 2.6345887184143066, + "learning_rate": 4.280548153787591e-06, + "loss": 1.5054, + "step": 686 + }, + { + "epoch": 0.7057010785824345, + "grad_norm": 2.3535311222076416, + "learning_rate": 4.2786448420251245e-06, + "loss": 1.4607, + "step": 687 + }, + { + "epoch": 0.706728299948639, + "grad_norm": 2.490431547164917, + "learning_rate": 4.276741530262658e-06, + "loss": 1.4852, + "step": 688 + }, + { + "epoch": 0.7077555213148433, + "grad_norm": 2.4353036880493164, + "learning_rate": 4.27483821850019e-06, + "loss": 1.5503, + "step": 689 + }, + { + "epoch": 0.7087827426810478, + "grad_norm": 2.3356940746307373, + "learning_rate": 4.272934906737724e-06, + "loss": 1.5416, + "step": 690 + }, + { + "epoch": 0.7087827426810478, + "eval_loss": 1.8164218664169312, + "eval_runtime": 38.0592, + "eval_samples_per_second": 26.275, + "eval_steps_per_second": 1.104, + "step": 690 + }, + { + "epoch": 0.7098099640472522, + "grad_norm": 2.4772775173187256, + "learning_rate": 4.271031594975257e-06, + "loss": 1.5392, + "step": 691 + }, + { + "epoch": 0.7108371854134566, + "grad_norm": 2.2921438217163086, + "learning_rate": 4.2691282832127905e-06, + "loss": 1.4935, + "step": 692 + }, + { + "epoch": 0.711864406779661, + "grad_norm": 2.3176305294036865, + "learning_rate": 4.267224971450324e-06, + "loss": 1.4451, + "step": 693 + }, + { + "epoch": 0.7128916281458655, + "grad_norm": 2.6825685501098633, + "learning_rate": 4.265321659687857e-06, + "loss": 1.4878, + "step": 694 + }, + { + "epoch": 0.7139188495120699, + "grad_norm": 2.6141414642333984, + "learning_rate": 4.2634183479253906e-06, + "loss": 1.4695, + "step": 695 + }, + { + "epoch": 0.7149460708782742, + "grad_norm": 2.5179429054260254, + "learning_rate": 4.261515036162924e-06, + "loss": 1.492, + "step": 696 + }, + { + "epoch": 0.7159732922444787, + "grad_norm": 2.4744884967803955, + "learning_rate": 4.259611724400457e-06, + "loss": 1.4752, + "step": 697 + }, + { + "epoch": 0.7170005136106831, + "grad_norm": 2.4911048412323, + "learning_rate": 4.257708412637991e-06, + "loss": 1.423, + "step": 698 + }, + { + "epoch": 0.7180277349768875, + "grad_norm": 2.5116214752197266, + "learning_rate": 4.255805100875523e-06, + "loss": 1.4535, + "step": 699 + }, + { + "epoch": 0.7190549563430919, + "grad_norm": 2.476651430130005, + "learning_rate": 4.2539017891130565e-06, + "loss": 1.5349, + "step": 700 + }, + { + "epoch": 0.7190549563430919, + "eval_loss": 1.8069179058074951, + "eval_runtime": 38.0292, + "eval_samples_per_second": 26.296, + "eval_steps_per_second": 1.104, + "step": 700 + }, + { + "epoch": 0.7200821777092964, + "grad_norm": 2.5234575271606445, + "learning_rate": 4.25199847735059e-06, + "loss": 1.4596, + "step": 701 + }, + { + "epoch": 0.7211093990755008, + "grad_norm": 2.33290433883667, + "learning_rate": 4.250095165588123e-06, + "loss": 1.5147, + "step": 702 + }, + { + "epoch": 0.7221366204417052, + "grad_norm": 2.479883909225464, + "learning_rate": 4.2481918538256575e-06, + "loss": 1.47, + "step": 703 + }, + { + "epoch": 0.7231638418079096, + "grad_norm": 2.432253837585449, + "learning_rate": 4.246288542063191e-06, + "loss": 1.5152, + "step": 704 + }, + { + "epoch": 0.724191063174114, + "grad_norm": 2.358304738998413, + "learning_rate": 4.244385230300724e-06, + "loss": 1.4562, + "step": 705 + }, + { + "epoch": 0.7252182845403184, + "grad_norm": 2.480877161026001, + "learning_rate": 4.242481918538257e-06, + "loss": 1.4906, + "step": 706 + }, + { + "epoch": 0.7262455059065228, + "grad_norm": 2.4287192821502686, + "learning_rate": 4.24057860677579e-06, + "loss": 1.5294, + "step": 707 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 2.4782888889312744, + "learning_rate": 4.2386752950133234e-06, + "loss": 1.4767, + "step": 708 + }, + { + "epoch": 0.7282999486389317, + "grad_norm": 2.4948253631591797, + "learning_rate": 4.236771983250857e-06, + "loss": 1.4664, + "step": 709 + }, + { + "epoch": 0.7293271700051361, + "grad_norm": 2.4625205993652344, + "learning_rate": 4.23486867148839e-06, + "loss": 1.495, + "step": 710 + }, + { + "epoch": 0.7293271700051361, + "eval_loss": 1.8140079975128174, + "eval_runtime": 37.6964, + "eval_samples_per_second": 26.528, + "eval_steps_per_second": 1.114, + "step": 710 + }, + { + "epoch": 0.7303543913713405, + "grad_norm": 2.669027805328369, + "learning_rate": 4.2329653597259235e-06, + "loss": 1.5204, + "step": 711 + }, + { + "epoch": 0.731381612737545, + "grad_norm": 2.5024707317352295, + "learning_rate": 4.231062047963457e-06, + "loss": 1.5283, + "step": 712 + }, + { + "epoch": 0.7324088341037494, + "grad_norm": 2.498258590698242, + "learning_rate": 4.22915873620099e-06, + "loss": 1.5139, + "step": 713 + }, + { + "epoch": 0.7334360554699538, + "grad_norm": 2.4305667877197266, + "learning_rate": 4.227255424438524e-06, + "loss": 1.5337, + "step": 714 + }, + { + "epoch": 0.7344632768361582, + "grad_norm": 2.3153419494628906, + "learning_rate": 4.225352112676057e-06, + "loss": 1.4646, + "step": 715 + }, + { + "epoch": 0.7354904982023626, + "grad_norm": 2.2835583686828613, + "learning_rate": 4.2234488009135895e-06, + "loss": 1.4328, + "step": 716 + }, + { + "epoch": 0.736517719568567, + "grad_norm": 2.401418924331665, + "learning_rate": 4.221545489151123e-06, + "loss": 1.4685, + "step": 717 + }, + { + "epoch": 0.7375449409347714, + "grad_norm": 2.454864978790283, + "learning_rate": 4.219642177388656e-06, + "loss": 1.4925, + "step": 718 + }, + { + "epoch": 0.7385721623009759, + "grad_norm": 2.2332873344421387, + "learning_rate": 4.2177388656261896e-06, + "loss": 1.4481, + "step": 719 + }, + { + "epoch": 0.7395993836671803, + "grad_norm": 2.46628475189209, + "learning_rate": 4.215835553863723e-06, + "loss": 1.4173, + "step": 720 + }, + { + "epoch": 0.7395993836671803, + "eval_loss": 1.817592740058899, + "eval_runtime": 37.8086, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 720 + }, + { + "epoch": 0.7406266050333847, + "grad_norm": 2.346245765686035, + "learning_rate": 4.213932242101256e-06, + "loss": 1.4688, + "step": 721 + }, + { + "epoch": 0.7416538263995891, + "grad_norm": 2.2708640098571777, + "learning_rate": 4.21202893033879e-06, + "loss": 1.4937, + "step": 722 + }, + { + "epoch": 0.7426810477657936, + "grad_norm": 2.3716516494750977, + "learning_rate": 4.210125618576323e-06, + "loss": 1.4869, + "step": 723 + }, + { + "epoch": 0.743708269131998, + "grad_norm": 2.443058490753174, + "learning_rate": 4.208222306813856e-06, + "loss": 1.542, + "step": 724 + }, + { + "epoch": 0.7447354904982023, + "grad_norm": 2.327270984649658, + "learning_rate": 4.20631899505139e-06, + "loss": 1.3933, + "step": 725 + }, + { + "epoch": 0.7457627118644068, + "grad_norm": 2.3734512329101562, + "learning_rate": 4.204415683288923e-06, + "loss": 1.5109, + "step": 726 + }, + { + "epoch": 0.7467899332306112, + "grad_norm": 2.3631174564361572, + "learning_rate": 4.2025123715264565e-06, + "loss": 1.474, + "step": 727 + }, + { + "epoch": 0.7478171545968156, + "grad_norm": 2.306068181991577, + "learning_rate": 4.20060905976399e-06, + "loss": 1.42, + "step": 728 + }, + { + "epoch": 0.74884437596302, + "grad_norm": 2.391889810562134, + "learning_rate": 4.198705748001523e-06, + "loss": 1.5178, + "step": 729 + }, + { + "epoch": 0.7498715973292245, + "grad_norm": 2.3151705265045166, + "learning_rate": 4.1968024362390566e-06, + "loss": 1.4808, + "step": 730 + }, + { + "epoch": 0.7498715973292245, + "eval_loss": 1.8157070875167847, + "eval_runtime": 37.8799, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 730 + }, + { + "epoch": 0.7508988186954288, + "grad_norm": 2.336479902267456, + "learning_rate": 4.19489912447659e-06, + "loss": 1.5438, + "step": 731 + }, + { + "epoch": 0.7519260400616333, + "grad_norm": 2.338242292404175, + "learning_rate": 4.192995812714123e-06, + "loss": 1.4869, + "step": 732 + }, + { + "epoch": 0.7529532614278377, + "grad_norm": 2.268650770187378, + "learning_rate": 4.191092500951656e-06, + "loss": 1.4787, + "step": 733 + }, + { + "epoch": 0.7539804827940421, + "grad_norm": 2.3010005950927734, + "learning_rate": 4.189189189189189e-06, + "loss": 1.4711, + "step": 734 + }, + { + "epoch": 0.7550077041602465, + "grad_norm": 2.54770827293396, + "learning_rate": 4.1872858774267225e-06, + "loss": 1.5351, + "step": 735 + }, + { + "epoch": 0.7560349255264509, + "grad_norm": 2.3412787914276123, + "learning_rate": 4.185382565664256e-06, + "loss": 1.4655, + "step": 736 + }, + { + "epoch": 0.7570621468926554, + "grad_norm": 2.266413927078247, + "learning_rate": 4.183479253901789e-06, + "loss": 1.4402, + "step": 737 + }, + { + "epoch": 0.7580893682588598, + "grad_norm": 2.2785308361053467, + "learning_rate": 4.181575942139323e-06, + "loss": 1.4874, + "step": 738 + }, + { + "epoch": 0.7591165896250642, + "grad_norm": 2.2710752487182617, + "learning_rate": 4.179672630376856e-06, + "loss": 1.4644, + "step": 739 + }, + { + "epoch": 0.7601438109912686, + "grad_norm": 2.2460498809814453, + "learning_rate": 4.177769318614389e-06, + "loss": 1.5267, + "step": 740 + }, + { + "epoch": 0.7601438109912686, + "eval_loss": 1.8187947273254395, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 740 + }, + { + "epoch": 0.7611710323574731, + "grad_norm": 2.3589370250701904, + "learning_rate": 4.175866006851923e-06, + "loss": 1.5115, + "step": 741 + }, + { + "epoch": 0.7621982537236774, + "grad_norm": 2.2818758487701416, + "learning_rate": 4.173962695089456e-06, + "loss": 1.4945, + "step": 742 + }, + { + "epoch": 0.7632254750898819, + "grad_norm": 2.338085889816284, + "learning_rate": 4.1720593833269894e-06, + "loss": 1.5076, + "step": 743 + }, + { + "epoch": 0.7642526964560863, + "grad_norm": 2.3874809741973877, + "learning_rate": 4.170156071564523e-06, + "loss": 1.5665, + "step": 744 + }, + { + "epoch": 0.7652799178222907, + "grad_norm": 2.2418739795684814, + "learning_rate": 4.168252759802056e-06, + "loss": 1.4827, + "step": 745 + }, + { + "epoch": 0.7663071391884951, + "grad_norm": 2.441315174102783, + "learning_rate": 4.1663494480395895e-06, + "loss": 1.5171, + "step": 746 + }, + { + "epoch": 0.7673343605546995, + "grad_norm": 2.3201963901519775, + "learning_rate": 4.164446136277123e-06, + "loss": 1.5153, + "step": 747 + }, + { + "epoch": 0.768361581920904, + "grad_norm": 2.3395135402679443, + "learning_rate": 4.162542824514656e-06, + "loss": 1.4911, + "step": 748 + }, + { + "epoch": 0.7693888032871083, + "grad_norm": 2.3455941677093506, + "learning_rate": 4.16063951275219e-06, + "loss": 1.4493, + "step": 749 + }, + { + "epoch": 0.7704160246533128, + "grad_norm": 2.3737382888793945, + "learning_rate": 4.158736200989722e-06, + "loss": 1.4893, + "step": 750 + }, + { + "epoch": 0.7704160246533128, + "eval_loss": 1.8153141736984253, + "eval_runtime": 37.9216, + "eval_samples_per_second": 26.37, + "eval_steps_per_second": 1.108, + "step": 750 + }, + { + "epoch": 0.7714432460195172, + "grad_norm": 2.503803014755249, + "learning_rate": 4.1568328892272555e-06, + "loss": 1.499, + "step": 751 + }, + { + "epoch": 0.7724704673857217, + "grad_norm": 2.469083547592163, + "learning_rate": 4.154929577464789e-06, + "loss": 1.4984, + "step": 752 + }, + { + "epoch": 0.773497688751926, + "grad_norm": 2.387991428375244, + "learning_rate": 4.153026265702322e-06, + "loss": 1.4283, + "step": 753 + }, + { + "epoch": 0.7745249101181304, + "grad_norm": 2.3355393409729004, + "learning_rate": 4.1511229539398556e-06, + "loss": 1.4869, + "step": 754 + }, + { + "epoch": 0.7755521314843349, + "grad_norm": 2.629345417022705, + "learning_rate": 4.149219642177389e-06, + "loss": 1.4885, + "step": 755 + }, + { + "epoch": 0.7765793528505393, + "grad_norm": 2.458944797515869, + "learning_rate": 4.147316330414922e-06, + "loss": 1.415, + "step": 756 + }, + { + "epoch": 0.7776065742167437, + "grad_norm": 2.40051531791687, + "learning_rate": 4.145413018652456e-06, + "loss": 1.5129, + "step": 757 + }, + { + "epoch": 0.7786337955829481, + "grad_norm": 2.5249011516571045, + "learning_rate": 4.143509706889989e-06, + "loss": 1.5002, + "step": 758 + }, + { + "epoch": 0.7796610169491526, + "grad_norm": 2.413422107696533, + "learning_rate": 4.141606395127522e-06, + "loss": 1.4837, + "step": 759 + }, + { + "epoch": 0.7806882383153569, + "grad_norm": 2.3438029289245605, + "learning_rate": 4.139703083365056e-06, + "loss": 1.5213, + "step": 760 + }, + { + "epoch": 0.7806882383153569, + "eval_loss": 1.8037596940994263, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 760 + }, + { + "epoch": 0.7817154596815614, + "grad_norm": 2.326828718185425, + "learning_rate": 4.137799771602588e-06, + "loss": 1.5031, + "step": 761 + }, + { + "epoch": 0.7827426810477658, + "grad_norm": 2.4902493953704834, + "learning_rate": 4.135896459840122e-06, + "loss": 1.4602, + "step": 762 + }, + { + "epoch": 0.7837699024139703, + "grad_norm": 2.488004684448242, + "learning_rate": 4.133993148077656e-06, + "loss": 1.4764, + "step": 763 + }, + { + "epoch": 0.7847971237801746, + "grad_norm": 2.5515761375427246, + "learning_rate": 4.132089836315189e-06, + "loss": 1.4946, + "step": 764 + }, + { + "epoch": 0.785824345146379, + "grad_norm": 2.556278705596924, + "learning_rate": 4.1301865245527226e-06, + "loss": 1.4338, + "step": 765 + }, + { + "epoch": 0.7868515665125835, + "grad_norm": 2.3017969131469727, + "learning_rate": 4.128283212790256e-06, + "loss": 1.4053, + "step": 766 + }, + { + "epoch": 0.7878787878787878, + "grad_norm": 2.4371304512023926, + "learning_rate": 4.1263799010277884e-06, + "loss": 1.5238, + "step": 767 + }, + { + "epoch": 0.7889060092449923, + "grad_norm": 2.6256046295166016, + "learning_rate": 4.124476589265322e-06, + "loss": 1.4672, + "step": 768 + }, + { + "epoch": 0.7899332306111967, + "grad_norm": 2.3725669384002686, + "learning_rate": 4.122573277502855e-06, + "loss": 1.4677, + "step": 769 + }, + { + "epoch": 0.7909604519774012, + "grad_norm": 2.2928454875946045, + "learning_rate": 4.1206699657403885e-06, + "loss": 1.4927, + "step": 770 + }, + { + "epoch": 0.7909604519774012, + "eval_loss": 1.792726755142212, + "eval_runtime": 37.9273, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 770 + }, + { + "epoch": 0.7919876733436055, + "grad_norm": 2.315674066543579, + "learning_rate": 4.118766653977922e-06, + "loss": 1.4632, + "step": 771 + }, + { + "epoch": 0.79301489470981, + "grad_norm": 2.418849468231201, + "learning_rate": 4.116863342215455e-06, + "loss": 1.446, + "step": 772 + }, + { + "epoch": 0.7940421160760144, + "grad_norm": 2.3209879398345947, + "learning_rate": 4.114960030452989e-06, + "loss": 1.4755, + "step": 773 + }, + { + "epoch": 0.7950693374422187, + "grad_norm": 2.2322115898132324, + "learning_rate": 4.113056718690522e-06, + "loss": 1.4149, + "step": 774 + }, + { + "epoch": 0.7960965588084232, + "grad_norm": 2.6517558097839355, + "learning_rate": 4.111153406928055e-06, + "loss": 1.5323, + "step": 775 + }, + { + "epoch": 0.7971237801746276, + "grad_norm": 2.4602737426757812, + "learning_rate": 4.109250095165589e-06, + "loss": 1.5179, + "step": 776 + }, + { + "epoch": 0.7981510015408321, + "grad_norm": 2.4015536308288574, + "learning_rate": 4.107346783403122e-06, + "loss": 1.4711, + "step": 777 + }, + { + "epoch": 0.7991782229070364, + "grad_norm": 2.4367337226867676, + "learning_rate": 4.105443471640655e-06, + "loss": 1.4425, + "step": 778 + }, + { + "epoch": 0.8002054442732409, + "grad_norm": 2.5161020755767822, + "learning_rate": 4.103540159878188e-06, + "loss": 1.4832, + "step": 779 + }, + { + "epoch": 0.8012326656394453, + "grad_norm": 2.483154773712158, + "learning_rate": 4.101636848115721e-06, + "loss": 1.4508, + "step": 780 + }, + { + "epoch": 0.8012326656394453, + "eval_loss": 1.792397379875183, + "eval_runtime": 37.8937, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 780 + }, + { + "epoch": 0.8022598870056498, + "grad_norm": 2.4649806022644043, + "learning_rate": 4.099733536353255e-06, + "loss": 1.4919, + "step": 781 + }, + { + "epoch": 0.8032871083718541, + "grad_norm": 2.578634262084961, + "learning_rate": 4.097830224590788e-06, + "loss": 1.565, + "step": 782 + }, + { + "epoch": 0.8043143297380585, + "grad_norm": 2.444427967071533, + "learning_rate": 4.095926912828322e-06, + "loss": 1.5309, + "step": 783 + }, + { + "epoch": 0.805341551104263, + "grad_norm": 2.32051420211792, + "learning_rate": 4.094023601065855e-06, + "loss": 1.4777, + "step": 784 + }, + { + "epoch": 0.8063687724704673, + "grad_norm": 2.3207180500030518, + "learning_rate": 4.092120289303388e-06, + "loss": 1.4863, + "step": 785 + }, + { + "epoch": 0.8073959938366718, + "grad_norm": 2.1404120922088623, + "learning_rate": 4.0902169775409215e-06, + "loss": 1.437, + "step": 786 + }, + { + "epoch": 0.8084232152028762, + "grad_norm": 2.2980856895446777, + "learning_rate": 4.088313665778455e-06, + "loss": 1.4571, + "step": 787 + }, + { + "epoch": 0.8094504365690807, + "grad_norm": 2.2555243968963623, + "learning_rate": 4.086410354015988e-06, + "loss": 1.4451, + "step": 788 + }, + { + "epoch": 0.810477657935285, + "grad_norm": 2.3837077617645264, + "learning_rate": 4.0845070422535216e-06, + "loss": 1.4641, + "step": 789 + }, + { + "epoch": 0.8115048793014895, + "grad_norm": 2.5199060440063477, + "learning_rate": 4.082603730491055e-06, + "loss": 1.5048, + "step": 790 + }, + { + "epoch": 0.8115048793014895, + "eval_loss": 1.8019492626190186, + "eval_runtime": 37.8805, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 790 + }, + { + "epoch": 0.8125321006676939, + "grad_norm": 2.4010138511657715, + "learning_rate": 4.080700418728588e-06, + "loss": 1.4881, + "step": 791 + }, + { + "epoch": 0.8135593220338984, + "grad_norm": 2.342094898223877, + "learning_rate": 4.078797106966122e-06, + "loss": 1.3792, + "step": 792 + }, + { + "epoch": 0.8145865434001027, + "grad_norm": 2.276772975921631, + "learning_rate": 4.076893795203655e-06, + "loss": 1.4539, + "step": 793 + }, + { + "epoch": 0.8156137647663071, + "grad_norm": 2.569275140762329, + "learning_rate": 4.074990483441188e-06, + "loss": 1.5059, + "step": 794 + }, + { + "epoch": 0.8166409861325116, + "grad_norm": 2.345761299133301, + "learning_rate": 4.073087171678721e-06, + "loss": 1.494, + "step": 795 + }, + { + "epoch": 0.8176682074987159, + "grad_norm": 2.4425504207611084, + "learning_rate": 4.071183859916254e-06, + "loss": 1.471, + "step": 796 + }, + { + "epoch": 0.8186954288649204, + "grad_norm": 2.5758635997772217, + "learning_rate": 4.069280548153788e-06, + "loss": 1.4378, + "step": 797 + }, + { + "epoch": 0.8197226502311248, + "grad_norm": 2.4320778846740723, + "learning_rate": 4.067377236391321e-06, + "loss": 1.474, + "step": 798 + }, + { + "epoch": 0.8207498715973293, + "grad_norm": 2.466978073120117, + "learning_rate": 4.065473924628854e-06, + "loss": 1.4997, + "step": 799 + }, + { + "epoch": 0.8217770929635336, + "grad_norm": 2.462489128112793, + "learning_rate": 4.063570612866388e-06, + "loss": 1.5, + "step": 800 + }, + { + "epoch": 0.8217770929635336, + "eval_loss": 1.783312201499939, + "eval_runtime": 37.8832, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 800 + }, + { + "epoch": 0.8228043143297381, + "grad_norm": 2.367455005645752, + "learning_rate": 4.061667301103921e-06, + "loss": 1.4658, + "step": 801 + }, + { + "epoch": 0.8238315356959425, + "grad_norm": 2.292250394821167, + "learning_rate": 4.0597639893414544e-06, + "loss": 1.476, + "step": 802 + }, + { + "epoch": 0.8248587570621468, + "grad_norm": 2.3460025787353516, + "learning_rate": 4.057860677578988e-06, + "loss": 1.4932, + "step": 803 + }, + { + "epoch": 0.8258859784283513, + "grad_norm": 2.3708906173706055, + "learning_rate": 4.055957365816521e-06, + "loss": 1.4396, + "step": 804 + }, + { + "epoch": 0.8269131997945557, + "grad_norm": 2.353513479232788, + "learning_rate": 4.0540540540540545e-06, + "loss": 1.4285, + "step": 805 + }, + { + "epoch": 0.8279404211607602, + "grad_norm": 2.319577932357788, + "learning_rate": 4.052150742291588e-06, + "loss": 1.4326, + "step": 806 + }, + { + "epoch": 0.8289676425269645, + "grad_norm": 2.450364112854004, + "learning_rate": 4.050247430529121e-06, + "loss": 1.5109, + "step": 807 + }, + { + "epoch": 0.829994863893169, + "grad_norm": 2.393907070159912, + "learning_rate": 4.048344118766655e-06, + "loss": 1.4948, + "step": 808 + }, + { + "epoch": 0.8310220852593734, + "grad_norm": 2.336002826690674, + "learning_rate": 4.046440807004188e-06, + "loss": 1.4851, + "step": 809 + }, + { + "epoch": 0.8320493066255779, + "grad_norm": 2.281139850616455, + "learning_rate": 4.044537495241721e-06, + "loss": 1.4445, + "step": 810 + }, + { + "epoch": 0.8320493066255779, + "eval_loss": 1.784231185913086, + "eval_runtime": 37.6614, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 810 + }, + { + "epoch": 0.8330765279917822, + "grad_norm": 2.4848175048828125, + "learning_rate": 4.042634183479255e-06, + "loss": 1.4982, + "step": 811 + }, + { + "epoch": 0.8341037493579867, + "grad_norm": 2.263347864151001, + "learning_rate": 4.040730871716787e-06, + "loss": 1.4795, + "step": 812 + }, + { + "epoch": 0.8351309707241911, + "grad_norm": 2.394005060195923, + "learning_rate": 4.038827559954321e-06, + "loss": 1.5231, + "step": 813 + }, + { + "epoch": 0.8361581920903954, + "grad_norm": 2.5051357746124268, + "learning_rate": 4.036924248191854e-06, + "loss": 1.4886, + "step": 814 + }, + { + "epoch": 0.8371854134565999, + "grad_norm": 2.3580374717712402, + "learning_rate": 4.035020936429387e-06, + "loss": 1.4713, + "step": 815 + }, + { + "epoch": 0.8382126348228043, + "grad_norm": 2.3179264068603516, + "learning_rate": 4.033117624666921e-06, + "loss": 1.4369, + "step": 816 + }, + { + "epoch": 0.8392398561890088, + "grad_norm": 2.269911289215088, + "learning_rate": 4.031214312904454e-06, + "loss": 1.4326, + "step": 817 + }, + { + "epoch": 0.8402670775552131, + "grad_norm": 2.3177807331085205, + "learning_rate": 4.029311001141987e-06, + "loss": 1.4096, + "step": 818 + }, + { + "epoch": 0.8412942989214176, + "grad_norm": 2.394324541091919, + "learning_rate": 4.027407689379521e-06, + "loss": 1.4647, + "step": 819 + }, + { + "epoch": 0.842321520287622, + "grad_norm": 2.288536548614502, + "learning_rate": 4.025504377617054e-06, + "loss": 1.4472, + "step": 820 + }, + { + "epoch": 0.842321520287622, + "eval_loss": 1.7811495065689087, + "eval_runtime": 37.8132, + "eval_samples_per_second": 26.446, + "eval_steps_per_second": 1.111, + "step": 820 + }, + { + "epoch": 0.8433487416538265, + "grad_norm": 2.376574754714966, + "learning_rate": 4.0236010658545875e-06, + "loss": 1.4918, + "step": 821 + }, + { + "epoch": 0.8443759630200308, + "grad_norm": 2.5850327014923096, + "learning_rate": 4.02169775409212e-06, + "loss": 1.5342, + "step": 822 + }, + { + "epoch": 0.8454031843862352, + "grad_norm": 2.247206211090088, + "learning_rate": 4.019794442329653e-06, + "loss": 1.4177, + "step": 823 + }, + { + "epoch": 0.8464304057524397, + "grad_norm": 2.4411795139312744, + "learning_rate": 4.0178911305671876e-06, + "loss": 1.4992, + "step": 824 + }, + { + "epoch": 0.847457627118644, + "grad_norm": 2.3377397060394287, + "learning_rate": 4.015987818804721e-06, + "loss": 1.4359, + "step": 825 + }, + { + "epoch": 0.8484848484848485, + "grad_norm": 2.329857349395752, + "learning_rate": 4.014084507042254e-06, + "loss": 1.4096, + "step": 826 + }, + { + "epoch": 0.8495120698510529, + "grad_norm": 2.563685178756714, + "learning_rate": 4.012181195279788e-06, + "loss": 1.5174, + "step": 827 + }, + { + "epoch": 0.8505392912172574, + "grad_norm": 2.4521079063415527, + "learning_rate": 4.010277883517321e-06, + "loss": 1.5028, + "step": 828 + }, + { + "epoch": 0.8515665125834617, + "grad_norm": 2.4091477394104004, + "learning_rate": 4.0083745717548535e-06, + "loss": 1.4615, + "step": 829 + }, + { + "epoch": 0.8525937339496662, + "grad_norm": 2.4010677337646484, + "learning_rate": 4.006471259992387e-06, + "loss": 1.4834, + "step": 830 + }, + { + "epoch": 0.8525937339496662, + "eval_loss": 1.7797678709030151, + "eval_runtime": 37.9378, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 830 + }, + { + "epoch": 0.8536209553158706, + "grad_norm": 2.3712635040283203, + "learning_rate": 4.00456794822992e-06, + "loss": 1.5196, + "step": 831 + }, + { + "epoch": 0.8546481766820749, + "grad_norm": 2.4096601009368896, + "learning_rate": 4.002664636467454e-06, + "loss": 1.4769, + "step": 832 + }, + { + "epoch": 0.8556753980482794, + "grad_norm": 2.408466100692749, + "learning_rate": 4.000761324704987e-06, + "loss": 1.4268, + "step": 833 + }, + { + "epoch": 0.8567026194144838, + "grad_norm": 2.6857047080993652, + "learning_rate": 3.99885801294252e-06, + "loss": 1.492, + "step": 834 + }, + { + "epoch": 0.8577298407806883, + "grad_norm": 2.297550916671753, + "learning_rate": 3.996954701180054e-06, + "loss": 1.4184, + "step": 835 + }, + { + "epoch": 0.8587570621468926, + "grad_norm": 2.423224925994873, + "learning_rate": 3.995051389417587e-06, + "loss": 1.4337, + "step": 836 + }, + { + "epoch": 0.8597842835130971, + "grad_norm": 2.260493040084839, + "learning_rate": 3.9931480776551204e-06, + "loss": 1.5267, + "step": 837 + }, + { + "epoch": 0.8608115048793015, + "grad_norm": 2.3803012371063232, + "learning_rate": 3.991244765892654e-06, + "loss": 1.4426, + "step": 838 + }, + { + "epoch": 0.861838726245506, + "grad_norm": 2.1425721645355225, + "learning_rate": 3.989341454130186e-06, + "loss": 1.4148, + "step": 839 + }, + { + "epoch": 0.8628659476117103, + "grad_norm": 2.316624641418457, + "learning_rate": 3.98743814236772e-06, + "loss": 1.4544, + "step": 840 + }, + { + "epoch": 0.8628659476117103, + "eval_loss": 1.7824597358703613, + "eval_runtime": 37.9676, + "eval_samples_per_second": 26.338, + "eval_steps_per_second": 1.106, + "step": 840 + }, + { + "epoch": 0.8638931689779148, + "grad_norm": 2.3723156452178955, + "learning_rate": 3.985534830605253e-06, + "loss": 1.473, + "step": 841 + }, + { + "epoch": 0.8649203903441192, + "grad_norm": 2.5167760848999023, + "learning_rate": 3.983631518842786e-06, + "loss": 1.4919, + "step": 842 + }, + { + "epoch": 0.8659476117103235, + "grad_norm": 2.3018670082092285, + "learning_rate": 3.98172820708032e-06, + "loss": 1.447, + "step": 843 + }, + { + "epoch": 0.866974833076528, + "grad_norm": 2.420483112335205, + "learning_rate": 3.979824895317854e-06, + "loss": 1.4159, + "step": 844 + }, + { + "epoch": 0.8680020544427324, + "grad_norm": 2.39443039894104, + "learning_rate": 3.977921583555387e-06, + "loss": 1.4984, + "step": 845 + }, + { + "epoch": 0.8690292758089369, + "grad_norm": 2.2953920364379883, + "learning_rate": 3.97601827179292e-06, + "loss": 1.4297, + "step": 846 + }, + { + "epoch": 0.8700564971751412, + "grad_norm": 2.2919528484344482, + "learning_rate": 3.974114960030453e-06, + "loss": 1.4728, + "step": 847 + }, + { + "epoch": 0.8710837185413457, + "grad_norm": 2.327914237976074, + "learning_rate": 3.9722116482679866e-06, + "loss": 1.4584, + "step": 848 + }, + { + "epoch": 0.8721109399075501, + "grad_norm": 2.305720567703247, + "learning_rate": 3.97030833650552e-06, + "loss": 1.4688, + "step": 849 + }, + { + "epoch": 0.8731381612737545, + "grad_norm": 2.2969236373901367, + "learning_rate": 3.968405024743053e-06, + "loss": 1.3972, + "step": 850 + }, + { + "epoch": 0.8731381612737545, + "eval_loss": 1.775634765625, + "eval_runtime": 37.9916, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 850 + }, + { + "epoch": 0.8741653826399589, + "grad_norm": 2.314387083053589, + "learning_rate": 3.966501712980587e-06, + "loss": 1.4476, + "step": 851 + }, + { + "epoch": 0.8751926040061633, + "grad_norm": 2.3271584510803223, + "learning_rate": 3.96459840121812e-06, + "loss": 1.4615, + "step": 852 + }, + { + "epoch": 0.8762198253723678, + "grad_norm": 2.494795322418213, + "learning_rate": 3.962695089455653e-06, + "loss": 1.5062, + "step": 853 + }, + { + "epoch": 0.8772470467385721, + "grad_norm": 2.4132444858551025, + "learning_rate": 3.960791777693187e-06, + "loss": 1.4861, + "step": 854 + }, + { + "epoch": 0.8782742681047766, + "grad_norm": 2.3869900703430176, + "learning_rate": 3.95888846593072e-06, + "loss": 1.4923, + "step": 855 + }, + { + "epoch": 0.879301489470981, + "grad_norm": 2.5167367458343506, + "learning_rate": 3.956985154168253e-06, + "loss": 1.5137, + "step": 856 + }, + { + "epoch": 0.8803287108371854, + "grad_norm": 2.510361671447754, + "learning_rate": 3.955081842405786e-06, + "loss": 1.5414, + "step": 857 + }, + { + "epoch": 0.8813559322033898, + "grad_norm": 2.554427146911621, + "learning_rate": 3.953178530643319e-06, + "loss": 1.5569, + "step": 858 + }, + { + "epoch": 0.8823831535695943, + "grad_norm": 2.306509256362915, + "learning_rate": 3.951275218880853e-06, + "loss": 1.3982, + "step": 859 + }, + { + "epoch": 0.8834103749357987, + "grad_norm": 2.3805363178253174, + "learning_rate": 3.949371907118386e-06, + "loss": 1.4977, + "step": 860 + }, + { + "epoch": 0.8834103749357987, + "eval_loss": 1.7882694005966187, + "eval_runtime": 38.0072, + "eval_samples_per_second": 26.311, + "eval_steps_per_second": 1.105, + "step": 860 + }, + { + "epoch": 0.884437596302003, + "grad_norm": 2.5413448810577393, + "learning_rate": 3.9474685953559194e-06, + "loss": 1.472, + "step": 861 + }, + { + "epoch": 0.8854648176682075, + "grad_norm": 2.484834909439087, + "learning_rate": 3.945565283593453e-06, + "loss": 1.4597, + "step": 862 + }, + { + "epoch": 0.8864920390344119, + "grad_norm": 2.6060585975646973, + "learning_rate": 3.943661971830986e-06, + "loss": 1.509, + "step": 863 + }, + { + "epoch": 0.8875192604006163, + "grad_norm": 2.6536927223205566, + "learning_rate": 3.9417586600685195e-06, + "loss": 1.4836, + "step": 864 + }, + { + "epoch": 0.8885464817668207, + "grad_norm": 2.5198960304260254, + "learning_rate": 3.939855348306053e-06, + "loss": 1.486, + "step": 865 + }, + { + "epoch": 0.8895737031330252, + "grad_norm": 2.335204839706421, + "learning_rate": 3.937952036543586e-06, + "loss": 1.4701, + "step": 866 + }, + { + "epoch": 0.8906009244992296, + "grad_norm": 2.7058098316192627, + "learning_rate": 3.93604872478112e-06, + "loss": 1.4798, + "step": 867 + }, + { + "epoch": 0.891628145865434, + "grad_norm": 2.4798922538757324, + "learning_rate": 3.934145413018653e-06, + "loss": 1.4364, + "step": 868 + }, + { + "epoch": 0.8926553672316384, + "grad_norm": 2.3090760707855225, + "learning_rate": 3.932242101256186e-06, + "loss": 1.4405, + "step": 869 + }, + { + "epoch": 0.8936825885978429, + "grad_norm": 2.4251503944396973, + "learning_rate": 3.93033878949372e-06, + "loss": 1.4113, + "step": 870 + }, + { + "epoch": 0.8936825885978429, + "eval_loss": 1.794334888458252, + "eval_runtime": 38.0197, + "eval_samples_per_second": 26.302, + "eval_steps_per_second": 1.105, + "step": 870 + }, + { + "epoch": 0.8947098099640473, + "grad_norm": 2.594945192337036, + "learning_rate": 3.928435477731253e-06, + "loss": 1.502, + "step": 871 + }, + { + "epoch": 0.8957370313302516, + "grad_norm": 2.4224843978881836, + "learning_rate": 3.9265321659687864e-06, + "loss": 1.5005, + "step": 872 + }, + { + "epoch": 0.8967642526964561, + "grad_norm": 2.3968186378479004, + "learning_rate": 3.924628854206319e-06, + "loss": 1.481, + "step": 873 + }, + { + "epoch": 0.8977914740626605, + "grad_norm": 2.4923524856567383, + "learning_rate": 3.922725542443852e-06, + "loss": 1.4259, + "step": 874 + }, + { + "epoch": 0.8988186954288649, + "grad_norm": 2.5386104583740234, + "learning_rate": 3.920822230681386e-06, + "loss": 1.4497, + "step": 875 + }, + { + "epoch": 0.8998459167950693, + "grad_norm": 2.4014909267425537, + "learning_rate": 3.918918918918919e-06, + "loss": 1.4816, + "step": 876 + }, + { + "epoch": 0.9008731381612738, + "grad_norm": 2.4019317626953125, + "learning_rate": 3.917015607156452e-06, + "loss": 1.4669, + "step": 877 + }, + { + "epoch": 0.9019003595274782, + "grad_norm": 2.5408380031585693, + "learning_rate": 3.915112295393986e-06, + "loss": 1.5343, + "step": 878 + }, + { + "epoch": 0.9029275808936826, + "grad_norm": 2.364053964614868, + "learning_rate": 3.913208983631519e-06, + "loss": 1.5229, + "step": 879 + }, + { + "epoch": 0.903954802259887, + "grad_norm": 2.307412624359131, + "learning_rate": 3.9113056718690525e-06, + "loss": 1.4444, + "step": 880 + }, + { + "epoch": 0.903954802259887, + "eval_loss": 1.7910975217819214, + "eval_runtime": 38.0659, + "eval_samples_per_second": 26.27, + "eval_steps_per_second": 1.103, + "step": 880 + }, + { + "epoch": 0.9049820236260914, + "grad_norm": 2.496135950088501, + "learning_rate": 3.909402360106586e-06, + "loss": 1.4853, + "step": 881 + }, + { + "epoch": 0.9060092449922958, + "grad_norm": 2.4825589656829834, + "learning_rate": 3.907499048344119e-06, + "loss": 1.4886, + "step": 882 + }, + { + "epoch": 0.9070364663585002, + "grad_norm": 2.3458824157714844, + "learning_rate": 3.905595736581652e-06, + "loss": 1.4882, + "step": 883 + }, + { + "epoch": 0.9080636877247047, + "grad_norm": 2.2699403762817383, + "learning_rate": 3.903692424819186e-06, + "loss": 1.4417, + "step": 884 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 2.348593235015869, + "learning_rate": 3.901789113056719e-06, + "loss": 1.4991, + "step": 885 + }, + { + "epoch": 0.9101181304571135, + "grad_norm": 2.3292129039764404, + "learning_rate": 3.899885801294253e-06, + "loss": 1.4752, + "step": 886 + }, + { + "epoch": 0.9111453518233179, + "grad_norm": 2.487183094024658, + "learning_rate": 3.897982489531786e-06, + "loss": 1.5317, + "step": 887 + }, + { + "epoch": 0.9121725731895224, + "grad_norm": 2.3300845623016357, + "learning_rate": 3.896079177769319e-06, + "loss": 1.4964, + "step": 888 + }, + { + "epoch": 0.9131997945557268, + "grad_norm": 2.3292312622070312, + "learning_rate": 3.894175866006853e-06, + "loss": 1.5221, + "step": 889 + }, + { + "epoch": 0.9142270159219312, + "grad_norm": 2.3556439876556396, + "learning_rate": 3.892272554244385e-06, + "loss": 1.4471, + "step": 890 + }, + { + "epoch": 0.9142270159219312, + "eval_loss": 1.7807039022445679, + "eval_runtime": 38.0614, + "eval_samples_per_second": 26.273, + "eval_steps_per_second": 1.103, + "step": 890 + }, + { + "epoch": 0.9152542372881356, + "grad_norm": 2.3302760124206543, + "learning_rate": 3.890369242481919e-06, + "loss": 1.5179, + "step": 891 + }, + { + "epoch": 0.91628145865434, + "grad_norm": 2.327072858810425, + "learning_rate": 3.888465930719452e-06, + "loss": 1.4474, + "step": 892 + }, + { + "epoch": 0.9173086800205444, + "grad_norm": 2.2773101329803467, + "learning_rate": 3.886562618956985e-06, + "loss": 1.4783, + "step": 893 + }, + { + "epoch": 0.9183359013867488, + "grad_norm": 2.2717974185943604, + "learning_rate": 3.884659307194519e-06, + "loss": 1.4062, + "step": 894 + }, + { + "epoch": 0.9193631227529533, + "grad_norm": 2.3350367546081543, + "learning_rate": 3.882755995432052e-06, + "loss": 1.483, + "step": 895 + }, + { + "epoch": 0.9203903441191577, + "grad_norm": 2.302811622619629, + "learning_rate": 3.8808526836695854e-06, + "loss": 1.4407, + "step": 896 + }, + { + "epoch": 0.9214175654853621, + "grad_norm": 2.4031271934509277, + "learning_rate": 3.878949371907119e-06, + "loss": 1.4331, + "step": 897 + }, + { + "epoch": 0.9224447868515665, + "grad_norm": 2.4806716442108154, + "learning_rate": 3.877046060144652e-06, + "loss": 1.4341, + "step": 898 + }, + { + "epoch": 0.923472008217771, + "grad_norm": 2.4243357181549072, + "learning_rate": 3.8751427483821855e-06, + "loss": 1.4701, + "step": 899 + }, + { + "epoch": 0.9244992295839753, + "grad_norm": 2.4953269958496094, + "learning_rate": 3.873239436619718e-06, + "loss": 1.4287, + "step": 900 + }, + { + "epoch": 0.9244992295839753, + "eval_loss": 1.7872728109359741, + "eval_runtime": 38.0906, + "eval_samples_per_second": 26.253, + "eval_steps_per_second": 1.103, + "step": 900 + }, + { + "epoch": 0.9255264509501797, + "grad_norm": 2.307490587234497, + "learning_rate": 3.871336124857251e-06, + "loss": 1.4693, + "step": 901 + }, + { + "epoch": 0.9265536723163842, + "grad_norm": 2.4745755195617676, + "learning_rate": 3.869432813094785e-06, + "loss": 1.4779, + "step": 902 + }, + { + "epoch": 0.9275808936825886, + "grad_norm": 2.367715358734131, + "learning_rate": 3.867529501332318e-06, + "loss": 1.4771, + "step": 903 + }, + { + "epoch": 0.928608115048793, + "grad_norm": 2.388549566268921, + "learning_rate": 3.8656261895698515e-06, + "loss": 1.4648, + "step": 904 + }, + { + "epoch": 0.9296353364149974, + "grad_norm": 2.39432954788208, + "learning_rate": 3.863722877807386e-06, + "loss": 1.4932, + "step": 905 + }, + { + "epoch": 0.9306625577812019, + "grad_norm": 2.3921315670013428, + "learning_rate": 3.861819566044919e-06, + "loss": 1.5312, + "step": 906 + }, + { + "epoch": 0.9316897791474062, + "grad_norm": 2.400240659713745, + "learning_rate": 3.859916254282452e-06, + "loss": 1.4651, + "step": 907 + }, + { + "epoch": 0.9327170005136107, + "grad_norm": 2.394559860229492, + "learning_rate": 3.858012942519985e-06, + "loss": 1.4759, + "step": 908 + }, + { + "epoch": 0.9337442218798151, + "grad_norm": 2.3365273475646973, + "learning_rate": 3.856109630757518e-06, + "loss": 1.4502, + "step": 909 + }, + { + "epoch": 0.9347714432460195, + "grad_norm": 2.465691328048706, + "learning_rate": 3.854206318995052e-06, + "loss": 1.5036, + "step": 910 + }, + { + "epoch": 0.9347714432460195, + "eval_loss": 1.7682362794876099, + "eval_runtime": 37.6624, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 910 + }, + { + "epoch": 0.9357986646122239, + "grad_norm": 2.3376758098602295, + "learning_rate": 3.852303007232585e-06, + "loss": 1.4342, + "step": 911 + }, + { + "epoch": 0.9368258859784283, + "grad_norm": 2.345855236053467, + "learning_rate": 3.850399695470118e-06, + "loss": 1.4444, + "step": 912 + }, + { + "epoch": 0.9378531073446328, + "grad_norm": 2.4444501399993896, + "learning_rate": 3.848496383707652e-06, + "loss": 1.4315, + "step": 913 + }, + { + "epoch": 0.9388803287108372, + "grad_norm": 2.4896373748779297, + "learning_rate": 3.846593071945185e-06, + "loss": 1.4228, + "step": 914 + }, + { + "epoch": 0.9399075500770416, + "grad_norm": 2.3970463275909424, + "learning_rate": 3.8446897601827185e-06, + "loss": 1.5283, + "step": 915 + }, + { + "epoch": 0.940934771443246, + "grad_norm": 2.5194883346557617, + "learning_rate": 3.842786448420252e-06, + "loss": 1.5445, + "step": 916 + }, + { + "epoch": 0.9419619928094505, + "grad_norm": 2.446465253829956, + "learning_rate": 3.840883136657785e-06, + "loss": 1.5248, + "step": 917 + }, + { + "epoch": 0.9429892141756548, + "grad_norm": 2.348477840423584, + "learning_rate": 3.838979824895318e-06, + "loss": 1.402, + "step": 918 + }, + { + "epoch": 0.9440164355418593, + "grad_norm": 2.4511539936065674, + "learning_rate": 3.837076513132851e-06, + "loss": 1.4893, + "step": 919 + }, + { + "epoch": 0.9450436569080637, + "grad_norm": 2.3801777362823486, + "learning_rate": 3.8351732013703844e-06, + "loss": 1.4457, + "step": 920 + }, + { + "epoch": 0.9450436569080637, + "eval_loss": 1.7810813188552856, + "eval_runtime": 37.7647, + "eval_samples_per_second": 26.48, + "eval_steps_per_second": 1.112, + "step": 920 + }, + { + "epoch": 0.9460708782742681, + "grad_norm": 2.5260043144226074, + "learning_rate": 3.833269889607918e-06, + "loss": 1.4267, + "step": 921 + }, + { + "epoch": 0.9470980996404725, + "grad_norm": 2.270653009414673, + "learning_rate": 3.831366577845451e-06, + "loss": 1.4313, + "step": 922 + }, + { + "epoch": 0.9481253210066769, + "grad_norm": 2.5123250484466553, + "learning_rate": 3.8294632660829845e-06, + "loss": 1.4588, + "step": 923 + }, + { + "epoch": 0.9491525423728814, + "grad_norm": 2.4242913722991943, + "learning_rate": 3.827559954320518e-06, + "loss": 1.4659, + "step": 924 + }, + { + "epoch": 0.9501797637390857, + "grad_norm": 2.4337117671966553, + "learning_rate": 3.825656642558051e-06, + "loss": 1.4838, + "step": 925 + }, + { + "epoch": 0.9512069851052902, + "grad_norm": 2.3577871322631836, + "learning_rate": 3.823753330795585e-06, + "loss": 1.4674, + "step": 926 + }, + { + "epoch": 0.9522342064714946, + "grad_norm": 2.2574679851531982, + "learning_rate": 3.821850019033118e-06, + "loss": 1.4011, + "step": 927 + }, + { + "epoch": 0.9532614278376991, + "grad_norm": 2.38043475151062, + "learning_rate": 3.819946707270651e-06, + "loss": 1.459, + "step": 928 + }, + { + "epoch": 0.9542886492039034, + "grad_norm": 2.4249954223632812, + "learning_rate": 3.818043395508185e-06, + "loss": 1.4649, + "step": 929 + }, + { + "epoch": 0.9553158705701078, + "grad_norm": 2.4420788288116455, + "learning_rate": 3.816140083745718e-06, + "loss": 1.532, + "step": 930 + }, + { + "epoch": 0.9553158705701078, + "eval_loss": 1.7805145978927612, + "eval_runtime": 37.8606, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 930 + }, + { + "epoch": 0.9563430919363123, + "grad_norm": 2.391042470932007, + "learning_rate": 3.8142367719832514e-06, + "loss": 1.4365, + "step": 931 + }, + { + "epoch": 0.9573703133025167, + "grad_norm": 2.462303876876831, + "learning_rate": 3.8123334602207844e-06, + "loss": 1.4367, + "step": 932 + }, + { + "epoch": 0.9583975346687211, + "grad_norm": 2.330714702606201, + "learning_rate": 3.8104301484583177e-06, + "loss": 1.427, + "step": 933 + }, + { + "epoch": 0.9594247560349255, + "grad_norm": 2.3475441932678223, + "learning_rate": 3.808526836695851e-06, + "loss": 1.4836, + "step": 934 + }, + { + "epoch": 0.96045197740113, + "grad_norm": 2.4546470642089844, + "learning_rate": 3.8066235249333845e-06, + "loss": 1.4601, + "step": 935 + }, + { + "epoch": 0.9614791987673343, + "grad_norm": 2.4604640007019043, + "learning_rate": 3.804720213170918e-06, + "loss": 1.4616, + "step": 936 + }, + { + "epoch": 0.9625064201335388, + "grad_norm": 2.3087422847747803, + "learning_rate": 3.8028169014084508e-06, + "loss": 1.4816, + "step": 937 + }, + { + "epoch": 0.9635336414997432, + "grad_norm": 2.375377893447876, + "learning_rate": 3.800913589645984e-06, + "loss": 1.471, + "step": 938 + }, + { + "epoch": 0.9645608628659477, + "grad_norm": 2.358741283416748, + "learning_rate": 3.7990102778835175e-06, + "loss": 1.4754, + "step": 939 + }, + { + "epoch": 0.965588084232152, + "grad_norm": 2.3990139961242676, + "learning_rate": 3.797106966121051e-06, + "loss": 1.534, + "step": 940 + }, + { + "epoch": 0.965588084232152, + "eval_loss": 1.7745739221572876, + "eval_runtime": 37.8883, + "eval_samples_per_second": 26.393, + "eval_steps_per_second": 1.109, + "step": 940 + }, + { + "epoch": 0.9666153055983564, + "grad_norm": 2.417396068572998, + "learning_rate": 3.7952036543585842e-06, + "loss": 1.4994, + "step": 941 + }, + { + "epoch": 0.9676425269645609, + "grad_norm": 2.323125123977661, + "learning_rate": 3.793300342596117e-06, + "loss": 1.5045, + "step": 942 + }, + { + "epoch": 0.9686697483307652, + "grad_norm": 2.4424829483032227, + "learning_rate": 3.7913970308336505e-06, + "loss": 1.4936, + "step": 943 + }, + { + "epoch": 0.9696969696969697, + "grad_norm": 2.2769429683685303, + "learning_rate": 3.789493719071184e-06, + "loss": 1.4453, + "step": 944 + }, + { + "epoch": 0.9707241910631741, + "grad_norm": 2.183701276779175, + "learning_rate": 3.7875904073087177e-06, + "loss": 1.4005, + "step": 945 + }, + { + "epoch": 0.9717514124293786, + "grad_norm": 3.4538655281066895, + "learning_rate": 3.785687095546251e-06, + "loss": 1.4521, + "step": 946 + }, + { + "epoch": 0.9727786337955829, + "grad_norm": 2.3735523223876953, + "learning_rate": 3.7837837837837844e-06, + "loss": 1.4328, + "step": 947 + }, + { + "epoch": 0.9738058551617874, + "grad_norm": 2.6214165687561035, + "learning_rate": 3.7818804720213178e-06, + "loss": 1.4004, + "step": 948 + }, + { + "epoch": 0.9748330765279918, + "grad_norm": 2.365091562271118, + "learning_rate": 3.7799771602588507e-06, + "loss": 1.4006, + "step": 949 + }, + { + "epoch": 0.9758602978941961, + "grad_norm": 2.1926634311676025, + "learning_rate": 3.778073848496384e-06, + "loss": 1.4587, + "step": 950 + }, + { + "epoch": 0.9758602978941961, + "eval_loss": 1.767502784729004, + "eval_runtime": 37.9287, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 950 + }, + { + "epoch": 0.9768875192604006, + "grad_norm": 2.3047192096710205, + "learning_rate": 3.7761705367339174e-06, + "loss": 1.4826, + "step": 951 + }, + { + "epoch": 0.977914740626605, + "grad_norm": 2.3306028842926025, + "learning_rate": 3.7742672249714508e-06, + "loss": 1.4914, + "step": 952 + }, + { + "epoch": 0.9789419619928095, + "grad_norm": 2.3409693241119385, + "learning_rate": 3.772363913208984e-06, + "loss": 1.4556, + "step": 953 + }, + { + "epoch": 0.9799691833590138, + "grad_norm": 2.3337934017181396, + "learning_rate": 3.770460601446517e-06, + "loss": 1.4751, + "step": 954 + }, + { + "epoch": 0.9809964047252183, + "grad_norm": 2.3354830741882324, + "learning_rate": 3.7685572896840504e-06, + "loss": 1.4176, + "step": 955 + }, + { + "epoch": 0.9820236260914227, + "grad_norm": 2.4049782752990723, + "learning_rate": 3.766653977921584e-06, + "loss": 1.3715, + "step": 956 + }, + { + "epoch": 0.9830508474576272, + "grad_norm": 2.5127065181732178, + "learning_rate": 3.764750666159117e-06, + "loss": 1.468, + "step": 957 + }, + { + "epoch": 0.9840780688238315, + "grad_norm": 2.5576820373535156, + "learning_rate": 3.7628473543966505e-06, + "loss": 1.4237, + "step": 958 + }, + { + "epoch": 0.9851052901900359, + "grad_norm": 2.394098997116089, + "learning_rate": 3.7609440426341835e-06, + "loss": 1.5126, + "step": 959 + }, + { + "epoch": 0.9861325115562404, + "grad_norm": 2.3358702659606934, + "learning_rate": 3.759040730871717e-06, + "loss": 1.4109, + "step": 960 + }, + { + "epoch": 0.9861325115562404, + "eval_loss": 1.7665778398513794, + "eval_runtime": 37.9028, + "eval_samples_per_second": 26.383, + "eval_steps_per_second": 1.108, + "step": 960 + }, + { + "epoch": 0.9871597329224447, + "grad_norm": 2.4213271141052246, + "learning_rate": 3.75713741910925e-06, + "loss": 1.3995, + "step": 961 + }, + { + "epoch": 0.9881869542886492, + "grad_norm": 2.367835760116577, + "learning_rate": 3.7552341073467836e-06, + "loss": 1.4491, + "step": 962 + }, + { + "epoch": 0.9892141756548536, + "grad_norm": 2.433840751647949, + "learning_rate": 3.753330795584317e-06, + "loss": 1.5078, + "step": 963 + }, + { + "epoch": 0.9902413970210581, + "grad_norm": 2.3976447582244873, + "learning_rate": 3.75142748382185e-06, + "loss": 1.4107, + "step": 964 + }, + { + "epoch": 0.9912686183872624, + "grad_norm": 2.4369053840637207, + "learning_rate": 3.749524172059384e-06, + "loss": 1.4569, + "step": 965 + }, + { + "epoch": 0.9922958397534669, + "grad_norm": 2.3203182220458984, + "learning_rate": 3.747620860296917e-06, + "loss": 1.4976, + "step": 966 + }, + { + "epoch": 0.9933230611196713, + "grad_norm": 2.431885242462158, + "learning_rate": 3.7457175485344504e-06, + "loss": 1.4945, + "step": 967 + }, + { + "epoch": 0.9943502824858758, + "grad_norm": 2.2997021675109863, + "learning_rate": 3.7438142367719837e-06, + "loss": 1.48, + "step": 968 + }, + { + "epoch": 0.9953775038520801, + "grad_norm": 2.2502849102020264, + "learning_rate": 3.741910925009517e-06, + "loss": 1.4322, + "step": 969 + }, + { + "epoch": 0.9964047252182845, + "grad_norm": 2.311049222946167, + "learning_rate": 3.7400076132470505e-06, + "loss": 1.4498, + "step": 970 + }, + { + "epoch": 0.9964047252182845, + "eval_loss": 1.771975040435791, + "eval_runtime": 37.852, + "eval_samples_per_second": 26.419, + "eval_steps_per_second": 1.11, + "step": 970 + }, + { + "epoch": 0.997431946584489, + "grad_norm": 2.431821584701538, + "learning_rate": 3.7381043014845834e-06, + "loss": 1.4796, + "step": 971 + }, + { + "epoch": 0.9984591679506933, + "grad_norm": 2.319145441055298, + "learning_rate": 3.7362009897221168e-06, + "loss": 1.459, + "step": 972 + }, + { + "epoch": 0.9994863893168978, + "grad_norm": 2.4445269107818604, + "learning_rate": 3.73429767795965e-06, + "loss": 1.4764, + "step": 973 + }, + { + "epoch": 1.0005136106831023, + "grad_norm": 2.272233009338379, + "learning_rate": 3.7323943661971835e-06, + "loss": 1.3799, + "step": 974 + }, + { + "epoch": 1.0015408320493067, + "grad_norm": 2.5421202182769775, + "learning_rate": 3.730491054434717e-06, + "loss": 1.3257, + "step": 975 + }, + { + "epoch": 1.002568053415511, + "grad_norm": 2.3988611698150635, + "learning_rate": 3.72858774267225e-06, + "loss": 1.4244, + "step": 976 + }, + { + "epoch": 1.0035952747817154, + "grad_norm": 2.5080084800720215, + "learning_rate": 3.726684430909783e-06, + "loss": 1.4461, + "step": 977 + }, + { + "epoch": 1.00462249614792, + "grad_norm": 2.411081314086914, + "learning_rate": 3.7247811191473165e-06, + "loss": 1.45, + "step": 978 + }, + { + "epoch": 1.0056497175141244, + "grad_norm": 2.729546070098877, + "learning_rate": 3.72287780738485e-06, + "loss": 1.4296, + "step": 979 + }, + { + "epoch": 1.0066769388803287, + "grad_norm": 2.5932579040527344, + "learning_rate": 3.7209744956223832e-06, + "loss": 1.4312, + "step": 980 + }, + { + "epoch": 1.0066769388803287, + "eval_loss": 1.770633339881897, + "eval_runtime": 37.8977, + "eval_samples_per_second": 26.387, + "eval_steps_per_second": 1.108, + "step": 980 + }, + { + "epoch": 1.007704160246533, + "grad_norm": 2.4512336254119873, + "learning_rate": 3.719071183859916e-06, + "loss": 1.3637, + "step": 981 + }, + { + "epoch": 1.0087313816127375, + "grad_norm": 2.5416886806488037, + "learning_rate": 3.7171678720974495e-06, + "loss": 1.4591, + "step": 982 + }, + { + "epoch": 1.009758602978942, + "grad_norm": 2.464679002761841, + "learning_rate": 3.715264560334983e-06, + "loss": 1.4306, + "step": 983 + }, + { + "epoch": 1.0107858243451464, + "grad_norm": 2.3745839595794678, + "learning_rate": 3.7133612485725163e-06, + "loss": 1.4426, + "step": 984 + }, + { + "epoch": 1.0118130457113508, + "grad_norm": 2.3606340885162354, + "learning_rate": 3.71145793681005e-06, + "loss": 1.3618, + "step": 985 + }, + { + "epoch": 1.0128402670775551, + "grad_norm": 2.5836591720581055, + "learning_rate": 3.7095546250475834e-06, + "loss": 1.4776, + "step": 986 + }, + { + "epoch": 1.0138674884437597, + "grad_norm": 2.4108684062957764, + "learning_rate": 3.7076513132851168e-06, + "loss": 1.4047, + "step": 987 + }, + { + "epoch": 1.014894709809964, + "grad_norm": 2.4316983222961426, + "learning_rate": 3.7057480015226497e-06, + "loss": 1.3772, + "step": 988 + }, + { + "epoch": 1.0159219311761685, + "grad_norm": 2.5510315895080566, + "learning_rate": 3.703844689760183e-06, + "loss": 1.4536, + "step": 989 + }, + { + "epoch": 1.0169491525423728, + "grad_norm": 2.4947116374969482, + "learning_rate": 3.7019413779977164e-06, + "loss": 1.434, + "step": 990 + }, + { + "epoch": 1.0169491525423728, + "eval_loss": 1.7928223609924316, + "eval_runtime": 37.8838, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 990 + }, + { + "epoch": 1.0179763739085772, + "grad_norm": 2.5938122272491455, + "learning_rate": 3.70003806623525e-06, + "loss": 1.4083, + "step": 991 + }, + { + "epoch": 1.0190035952747818, + "grad_norm": 2.472865581512451, + "learning_rate": 3.698134754472783e-06, + "loss": 1.4696, + "step": 992 + }, + { + "epoch": 1.0200308166409862, + "grad_norm": 2.374344825744629, + "learning_rate": 3.696231442710316e-06, + "loss": 1.403, + "step": 993 + }, + { + "epoch": 1.0210580380071905, + "grad_norm": 2.6928529739379883, + "learning_rate": 3.6943281309478495e-06, + "loss": 1.4216, + "step": 994 + }, + { + "epoch": 1.022085259373395, + "grad_norm": 2.640890598297119, + "learning_rate": 3.692424819185383e-06, + "loss": 1.4362, + "step": 995 + }, + { + "epoch": 1.0231124807395995, + "grad_norm": 2.45357084274292, + "learning_rate": 3.690521507422916e-06, + "loss": 1.382, + "step": 996 + }, + { + "epoch": 1.0241397021058039, + "grad_norm": 2.712517023086548, + "learning_rate": 3.6886181956604496e-06, + "loss": 1.395, + "step": 997 + }, + { + "epoch": 1.0251669234720082, + "grad_norm": 2.7127771377563477, + "learning_rate": 3.686714883897983e-06, + "loss": 1.4589, + "step": 998 + }, + { + "epoch": 1.0261941448382126, + "grad_norm": 2.5379409790039062, + "learning_rate": 3.684811572135516e-06, + "loss": 1.4551, + "step": 999 + }, + { + "epoch": 1.027221366204417, + "grad_norm": 2.5015134811401367, + "learning_rate": 3.6829082603730492e-06, + "loss": 1.4112, + "step": 1000 + }, + { + "epoch": 1.027221366204417, + "eval_loss": 1.7806837558746338, + "eval_runtime": 37.9091, + "eval_samples_per_second": 26.379, + "eval_steps_per_second": 1.108, + "step": 1000 + }, + { + "epoch": 1.0282485875706215, + "grad_norm": 2.481306314468384, + "learning_rate": 3.6810049486105826e-06, + "loss": 1.4469, + "step": 1001 + }, + { + "epoch": 1.029275808936826, + "grad_norm": 2.481520891189575, + "learning_rate": 3.679101636848116e-06, + "loss": 1.4284, + "step": 1002 + }, + { + "epoch": 1.0303030303030303, + "grad_norm": 2.4291019439697266, + "learning_rate": 3.6771983250856493e-06, + "loss": 1.4067, + "step": 1003 + }, + { + "epoch": 1.0313302516692346, + "grad_norm": 2.5361173152923584, + "learning_rate": 3.6752950133231822e-06, + "loss": 1.4347, + "step": 1004 + }, + { + "epoch": 1.0323574730354392, + "grad_norm": 2.5588085651397705, + "learning_rate": 3.6733917015607156e-06, + "loss": 1.4455, + "step": 1005 + }, + { + "epoch": 1.0333846944016436, + "grad_norm": 2.555091619491577, + "learning_rate": 3.6714883897982494e-06, + "loss": 1.399, + "step": 1006 + }, + { + "epoch": 1.034411915767848, + "grad_norm": 2.5489652156829834, + "learning_rate": 3.6695850780357828e-06, + "loss": 1.3894, + "step": 1007 + }, + { + "epoch": 1.0354391371340523, + "grad_norm": 2.5258426666259766, + "learning_rate": 3.667681766273316e-06, + "loss": 1.3962, + "step": 1008 + }, + { + "epoch": 1.0364663585002567, + "grad_norm": 2.455916166305542, + "learning_rate": 3.6657784545108495e-06, + "loss": 1.4773, + "step": 1009 + }, + { + "epoch": 1.0374935798664613, + "grad_norm": 2.5466208457946777, + "learning_rate": 3.6638751427483824e-06, + "loss": 1.4337, + "step": 1010 + }, + { + "epoch": 1.0374935798664613, + "eval_loss": 1.7840261459350586, + "eval_runtime": 37.6685, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 1010 + }, + { + "epoch": 1.0385208012326657, + "grad_norm": 2.5446081161499023, + "learning_rate": 3.6619718309859158e-06, + "loss": 1.4777, + "step": 1011 + }, + { + "epoch": 1.03954802259887, + "grad_norm": 2.509493350982666, + "learning_rate": 3.660068519223449e-06, + "loss": 1.4295, + "step": 1012 + }, + { + "epoch": 1.0405752439650744, + "grad_norm": 2.417632818222046, + "learning_rate": 3.6581652074609825e-06, + "loss": 1.4044, + "step": 1013 + }, + { + "epoch": 1.041602465331279, + "grad_norm": 2.457902193069458, + "learning_rate": 3.656261895698516e-06, + "loss": 1.3948, + "step": 1014 + }, + { + "epoch": 1.0426296866974833, + "grad_norm": 2.4869253635406494, + "learning_rate": 3.6543585839360492e-06, + "loss": 1.397, + "step": 1015 + }, + { + "epoch": 1.0436569080636877, + "grad_norm": 2.495419979095459, + "learning_rate": 3.652455272173582e-06, + "loss": 1.4037, + "step": 1016 + }, + { + "epoch": 1.044684129429892, + "grad_norm": 2.5231642723083496, + "learning_rate": 3.6505519604111155e-06, + "loss": 1.4156, + "step": 1017 + }, + { + "epoch": 1.0457113507960965, + "grad_norm": 2.8302087783813477, + "learning_rate": 3.648648648648649e-06, + "loss": 1.4529, + "step": 1018 + }, + { + "epoch": 1.046738572162301, + "grad_norm": 2.467430353164673, + "learning_rate": 3.6467453368861823e-06, + "loss": 1.4078, + "step": 1019 + }, + { + "epoch": 1.0477657935285054, + "grad_norm": 2.469128131866455, + "learning_rate": 3.6448420251237156e-06, + "loss": 1.4344, + "step": 1020 + }, + { + "epoch": 1.0477657935285054, + "eval_loss": 1.7800015211105347, + "eval_runtime": 37.854, + "eval_samples_per_second": 26.417, + "eval_steps_per_second": 1.11, + "step": 1020 + }, + { + "epoch": 1.0487930148947098, + "grad_norm": 2.59389328956604, + "learning_rate": 3.6429387133612486e-06, + "loss": 1.451, + "step": 1021 + }, + { + "epoch": 1.0498202362609141, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.641035401598782e-06, + "loss": 1.4026, + "step": 1022 + }, + { + "epoch": 1.0508474576271187, + "grad_norm": 2.601926565170288, + "learning_rate": 3.6391320898363153e-06, + "loss": 1.4382, + "step": 1023 + }, + { + "epoch": 1.051874678993323, + "grad_norm": 2.5336716175079346, + "learning_rate": 3.6372287780738487e-06, + "loss": 1.4184, + "step": 1024 + }, + { + "epoch": 1.0529019003595275, + "grad_norm": 2.503507614135742, + "learning_rate": 3.635325466311382e-06, + "loss": 1.4426, + "step": 1025 + }, + { + "epoch": 1.0539291217257318, + "grad_norm": 2.483013153076172, + "learning_rate": 3.633422154548916e-06, + "loss": 1.4563, + "step": 1026 + }, + { + "epoch": 1.0549563430919364, + "grad_norm": 2.4803354740142822, + "learning_rate": 3.6315188427864487e-06, + "loss": 1.4284, + "step": 1027 + }, + { + "epoch": 1.0559835644581408, + "grad_norm": 2.3735713958740234, + "learning_rate": 3.629615531023982e-06, + "loss": 1.3772, + "step": 1028 + }, + { + "epoch": 1.0570107858243452, + "grad_norm": 2.471226215362549, + "learning_rate": 3.6277122192615155e-06, + "loss": 1.4399, + "step": 1029 + }, + { + "epoch": 1.0580380071905495, + "grad_norm": 2.3827438354492188, + "learning_rate": 3.625808907499049e-06, + "loss": 1.3951, + "step": 1030 + }, + { + "epoch": 1.0580380071905495, + "eval_loss": 1.7787058353424072, + "eval_runtime": 37.9484, + "eval_samples_per_second": 26.352, + "eval_steps_per_second": 1.107, + "step": 1030 + }, + { + "epoch": 1.059065228556754, + "grad_norm": 2.403416633605957, + "learning_rate": 3.623905595736582e-06, + "loss": 1.3946, + "step": 1031 + }, + { + "epoch": 1.0600924499229585, + "grad_norm": 2.455740213394165, + "learning_rate": 3.6220022839741156e-06, + "loss": 1.3859, + "step": 1032 + }, + { + "epoch": 1.0611196712891628, + "grad_norm": 2.6349902153015137, + "learning_rate": 3.6200989722116485e-06, + "loss": 1.4092, + "step": 1033 + }, + { + "epoch": 1.0621468926553672, + "grad_norm": 2.6292974948883057, + "learning_rate": 3.618195660449182e-06, + "loss": 1.419, + "step": 1034 + }, + { + "epoch": 1.0631741140215716, + "grad_norm": 2.4338958263397217, + "learning_rate": 3.6162923486867152e-06, + "loss": 1.3546, + "step": 1035 + }, + { + "epoch": 1.064201335387776, + "grad_norm": 2.376420259475708, + "learning_rate": 3.6143890369242486e-06, + "loss": 1.3899, + "step": 1036 + }, + { + "epoch": 1.0652285567539805, + "grad_norm": 2.428738832473755, + "learning_rate": 3.612485725161782e-06, + "loss": 1.3955, + "step": 1037 + }, + { + "epoch": 1.066255778120185, + "grad_norm": 2.4316651821136475, + "learning_rate": 3.610582413399315e-06, + "loss": 1.3875, + "step": 1038 + }, + { + "epoch": 1.0672829994863893, + "grad_norm": 2.3969266414642334, + "learning_rate": 3.6086791016368482e-06, + "loss": 1.404, + "step": 1039 + }, + { + "epoch": 1.0683102208525936, + "grad_norm": 2.662283420562744, + "learning_rate": 3.6067757898743816e-06, + "loss": 1.4639, + "step": 1040 + }, + { + "epoch": 1.0683102208525936, + "eval_loss": 1.7603929042816162, + "eval_runtime": 38.0098, + "eval_samples_per_second": 26.309, + "eval_steps_per_second": 1.105, + "step": 1040 + }, + { + "epoch": 1.0693374422187982, + "grad_norm": 2.601452112197876, + "learning_rate": 3.604872478111915e-06, + "loss": 1.47, + "step": 1041 + }, + { + "epoch": 1.0703646635850026, + "grad_norm": 2.4221863746643066, + "learning_rate": 3.6029691663494483e-06, + "loss": 1.4162, + "step": 1042 + }, + { + "epoch": 1.071391884951207, + "grad_norm": 2.4996650218963623, + "learning_rate": 3.6010658545869813e-06, + "loss": 1.443, + "step": 1043 + }, + { + "epoch": 1.0724191063174113, + "grad_norm": 2.547239303588867, + "learning_rate": 3.5991625428245146e-06, + "loss": 1.4464, + "step": 1044 + }, + { + "epoch": 1.073446327683616, + "grad_norm": 2.481187343597412, + "learning_rate": 3.597259231062048e-06, + "loss": 1.3749, + "step": 1045 + }, + { + "epoch": 1.0744735490498203, + "grad_norm": 2.5504934787750244, + "learning_rate": 3.5953559192995818e-06, + "loss": 1.4231, + "step": 1046 + }, + { + "epoch": 1.0755007704160247, + "grad_norm": 2.465390205383301, + "learning_rate": 3.593452607537115e-06, + "loss": 1.4805, + "step": 1047 + }, + { + "epoch": 1.076527991782229, + "grad_norm": 2.4399969577789307, + "learning_rate": 3.5915492957746485e-06, + "loss": 1.4091, + "step": 1048 + }, + { + "epoch": 1.0775552131484334, + "grad_norm": 2.4819061756134033, + "learning_rate": 3.589645984012182e-06, + "loss": 1.4367, + "step": 1049 + }, + { + "epoch": 1.078582434514638, + "grad_norm": 2.459517002105713, + "learning_rate": 3.587742672249715e-06, + "loss": 1.3882, + "step": 1050 + }, + { + "epoch": 1.078582434514638, + "eval_loss": 1.7557458877563477, + "eval_runtime": 38.0271, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1050 + }, + { + "epoch": 1.0796096558808423, + "grad_norm": 2.5434813499450684, + "learning_rate": 3.585839360487248e-06, + "loss": 1.3926, + "step": 1051 + }, + { + "epoch": 1.0806368772470467, + "grad_norm": 2.6173205375671387, + "learning_rate": 3.5839360487247815e-06, + "loss": 1.4044, + "step": 1052 + }, + { + "epoch": 1.081664098613251, + "grad_norm": 2.677584171295166, + "learning_rate": 3.582032736962315e-06, + "loss": 1.4695, + "step": 1053 + }, + { + "epoch": 1.0826913199794557, + "grad_norm": 2.4151623249053955, + "learning_rate": 3.5801294251998483e-06, + "loss": 1.3699, + "step": 1054 + }, + { + "epoch": 1.08371854134566, + "grad_norm": 2.4683942794799805, + "learning_rate": 3.578226113437381e-06, + "loss": 1.4132, + "step": 1055 + }, + { + "epoch": 1.0847457627118644, + "grad_norm": 2.6076581478118896, + "learning_rate": 3.5763228016749146e-06, + "loss": 1.4729, + "step": 1056 + }, + { + "epoch": 1.0857729840780688, + "grad_norm": 2.4766876697540283, + "learning_rate": 3.574419489912448e-06, + "loss": 1.4196, + "step": 1057 + }, + { + "epoch": 1.0868002054442731, + "grad_norm": 2.466458559036255, + "learning_rate": 3.5725161781499813e-06, + "loss": 1.4049, + "step": 1058 + }, + { + "epoch": 1.0878274268104777, + "grad_norm": 2.4220409393310547, + "learning_rate": 3.5706128663875147e-06, + "loss": 1.4246, + "step": 1059 + }, + { + "epoch": 1.088854648176682, + "grad_norm": 2.5571770668029785, + "learning_rate": 3.5687095546250476e-06, + "loss": 1.4397, + "step": 1060 + }, + { + "epoch": 1.088854648176682, + "eval_loss": 1.7701127529144287, + "eval_runtime": 38.0754, + "eval_samples_per_second": 26.264, + "eval_steps_per_second": 1.103, + "step": 1060 + }, + { + "epoch": 1.0898818695428865, + "grad_norm": 2.648768186569214, + "learning_rate": 3.566806242862581e-06, + "loss": 1.423, + "step": 1061 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 2.403651714324951, + "learning_rate": 3.5649029311001143e-06, + "loss": 1.3995, + "step": 1062 + }, + { + "epoch": 1.0919363122752954, + "grad_norm": 2.465538263320923, + "learning_rate": 3.5629996193376477e-06, + "loss": 1.4139, + "step": 1063 + }, + { + "epoch": 1.0929635336414998, + "grad_norm": 2.464662551879883, + "learning_rate": 3.561096307575181e-06, + "loss": 1.4499, + "step": 1064 + }, + { + "epoch": 1.0939907550077042, + "grad_norm": 2.4943597316741943, + "learning_rate": 3.559192995812714e-06, + "loss": 1.4555, + "step": 1065 + }, + { + "epoch": 1.0950179763739085, + "grad_norm": 2.6332199573516846, + "learning_rate": 3.557289684050248e-06, + "loss": 1.4484, + "step": 1066 + }, + { + "epoch": 1.0960451977401129, + "grad_norm": 2.4370667934417725, + "learning_rate": 3.555386372287781e-06, + "loss": 1.4588, + "step": 1067 + }, + { + "epoch": 1.0970724191063175, + "grad_norm": 2.568368673324585, + "learning_rate": 3.5534830605253145e-06, + "loss": 1.4269, + "step": 1068 + }, + { + "epoch": 1.0980996404725218, + "grad_norm": 2.4618077278137207, + "learning_rate": 3.551579748762848e-06, + "loss": 1.4165, + "step": 1069 + }, + { + "epoch": 1.0991268618387262, + "grad_norm": 2.551320791244507, + "learning_rate": 3.5496764370003812e-06, + "loss": 1.3991, + "step": 1070 + }, + { + "epoch": 1.0991268618387262, + "eval_loss": 1.7648736238479614, + "eval_runtime": 38.0273, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1070 + }, + { + "epoch": 1.1001540832049306, + "grad_norm": 2.350905418395996, + "learning_rate": 3.5477731252379146e-06, + "loss": 1.3556, + "step": 1071 + }, + { + "epoch": 1.1011813045711352, + "grad_norm": 2.4294588565826416, + "learning_rate": 3.5458698134754475e-06, + "loss": 1.3801, + "step": 1072 + }, + { + "epoch": 1.1022085259373395, + "grad_norm": 2.576774835586548, + "learning_rate": 3.543966501712981e-06, + "loss": 1.3876, + "step": 1073 + }, + { + "epoch": 1.103235747303544, + "grad_norm": 2.716935873031616, + "learning_rate": 3.5420631899505142e-06, + "loss": 1.4907, + "step": 1074 + }, + { + "epoch": 1.1042629686697483, + "grad_norm": 2.425305128097534, + "learning_rate": 3.5401598781880476e-06, + "loss": 1.3161, + "step": 1075 + }, + { + "epoch": 1.1052901900359529, + "grad_norm": 2.54738712310791, + "learning_rate": 3.538256566425581e-06, + "loss": 1.4462, + "step": 1076 + }, + { + "epoch": 1.1063174114021572, + "grad_norm": 2.5109341144561768, + "learning_rate": 3.536353254663114e-06, + "loss": 1.4022, + "step": 1077 + }, + { + "epoch": 1.1073446327683616, + "grad_norm": 2.468749523162842, + "learning_rate": 3.5344499429006473e-06, + "loss": 1.3857, + "step": 1078 + }, + { + "epoch": 1.108371854134566, + "grad_norm": 2.5285134315490723, + "learning_rate": 3.5325466311381806e-06, + "loss": 1.4309, + "step": 1079 + }, + { + "epoch": 1.1093990755007703, + "grad_norm": 2.6391055583953857, + "learning_rate": 3.530643319375714e-06, + "loss": 1.3644, + "step": 1080 + }, + { + "epoch": 1.1093990755007703, + "eval_loss": 1.7846598625183105, + "eval_runtime": 38.0604, + "eval_samples_per_second": 26.274, + "eval_steps_per_second": 1.104, + "step": 1080 + }, + { + "epoch": 1.110426296866975, + "grad_norm": 2.660356283187866, + "learning_rate": 3.5287400076132474e-06, + "loss": 1.3922, + "step": 1081 + }, + { + "epoch": 1.1114535182331793, + "grad_norm": 2.5648741722106934, + "learning_rate": 3.5268366958507803e-06, + "loss": 1.4282, + "step": 1082 + }, + { + "epoch": 1.1124807395993837, + "grad_norm": 2.6862876415252686, + "learning_rate": 3.5249333840883137e-06, + "loss": 1.4208, + "step": 1083 + }, + { + "epoch": 1.113507960965588, + "grad_norm": 2.6200191974639893, + "learning_rate": 3.523030072325847e-06, + "loss": 1.3883, + "step": 1084 + }, + { + "epoch": 1.1145351823317924, + "grad_norm": 2.5802111625671387, + "learning_rate": 3.5211267605633804e-06, + "loss": 1.4033, + "step": 1085 + }, + { + "epoch": 1.115562403697997, + "grad_norm": 2.4148080348968506, + "learning_rate": 3.519223448800914e-06, + "loss": 1.4085, + "step": 1086 + }, + { + "epoch": 1.1165896250642013, + "grad_norm": 2.7452127933502197, + "learning_rate": 3.5173201370384475e-06, + "loss": 1.4021, + "step": 1087 + }, + { + "epoch": 1.1176168464304057, + "grad_norm": 2.6437106132507324, + "learning_rate": 3.515416825275981e-06, + "loss": 1.4161, + "step": 1088 + }, + { + "epoch": 1.11864406779661, + "grad_norm": 2.4709842205047607, + "learning_rate": 3.513513513513514e-06, + "loss": 1.4322, + "step": 1089 + }, + { + "epoch": 1.1196712891628147, + "grad_norm": 2.5138099193573, + "learning_rate": 3.511610201751047e-06, + "loss": 1.3435, + "step": 1090 + }, + { + "epoch": 1.1196712891628147, + "eval_loss": 1.7687023878097534, + "eval_runtime": 38.0137, + "eval_samples_per_second": 26.306, + "eval_steps_per_second": 1.105, + "step": 1090 + }, + { + "epoch": 1.120698510529019, + "grad_norm": 2.4194881916046143, + "learning_rate": 3.5097068899885806e-06, + "loss": 1.4133, + "step": 1091 + }, + { + "epoch": 1.1217257318952234, + "grad_norm": 2.6564371585845947, + "learning_rate": 3.507803578226114e-06, + "loss": 1.4759, + "step": 1092 + }, + { + "epoch": 1.1227529532614278, + "grad_norm": 2.4480180740356445, + "learning_rate": 3.5059002664636473e-06, + "loss": 1.3577, + "step": 1093 + }, + { + "epoch": 1.1237801746276324, + "grad_norm": 2.5538978576660156, + "learning_rate": 3.5039969547011802e-06, + "loss": 1.3984, + "step": 1094 + }, + { + "epoch": 1.1248073959938367, + "grad_norm": 2.6467089653015137, + "learning_rate": 3.5020936429387136e-06, + "loss": 1.4049, + "step": 1095 + }, + { + "epoch": 1.125834617360041, + "grad_norm": 2.6502673625946045, + "learning_rate": 3.500190331176247e-06, + "loss": 1.462, + "step": 1096 + }, + { + "epoch": 1.1268618387262455, + "grad_norm": 2.5696403980255127, + "learning_rate": 3.4982870194137803e-06, + "loss": 1.3926, + "step": 1097 + }, + { + "epoch": 1.1278890600924498, + "grad_norm": 2.5951733589172363, + "learning_rate": 3.4963837076513137e-06, + "loss": 1.4031, + "step": 1098 + }, + { + "epoch": 1.1289162814586544, + "grad_norm": 2.4910218715667725, + "learning_rate": 3.4944803958888466e-06, + "loss": 1.3348, + "step": 1099 + }, + { + "epoch": 1.1299435028248588, + "grad_norm": 2.6325888633728027, + "learning_rate": 3.49257708412638e-06, + "loss": 1.4641, + "step": 1100 + }, + { + "epoch": 1.1299435028248588, + "eval_loss": 1.7720683813095093, + "eval_runtime": 37.9585, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 1100 + }, + { + "epoch": 1.1309707241910631, + "grad_norm": 2.6712775230407715, + "learning_rate": 3.4906737723639133e-06, + "loss": 1.4726, + "step": 1101 + }, + { + "epoch": 1.1319979455572675, + "grad_norm": 2.580404043197632, + "learning_rate": 3.4887704606014467e-06, + "loss": 1.4158, + "step": 1102 + }, + { + "epoch": 1.133025166923472, + "grad_norm": 2.467453718185425, + "learning_rate": 3.48686714883898e-06, + "loss": 1.4636, + "step": 1103 + }, + { + "epoch": 1.1340523882896765, + "grad_norm": 2.627136468887329, + "learning_rate": 3.484963837076513e-06, + "loss": 1.4082, + "step": 1104 + }, + { + "epoch": 1.1350796096558808, + "grad_norm": 2.6198604106903076, + "learning_rate": 3.4830605253140464e-06, + "loss": 1.3628, + "step": 1105 + }, + { + "epoch": 1.1361068310220852, + "grad_norm": 2.439512014389038, + "learning_rate": 3.4811572135515797e-06, + "loss": 1.3885, + "step": 1106 + }, + { + "epoch": 1.1371340523882898, + "grad_norm": 2.5543010234832764, + "learning_rate": 3.4792539017891135e-06, + "loss": 1.3674, + "step": 1107 + }, + { + "epoch": 1.1381612737544942, + "grad_norm": 2.7263364791870117, + "learning_rate": 3.477350590026647e-06, + "loss": 1.4387, + "step": 1108 + }, + { + "epoch": 1.1391884951206985, + "grad_norm": 2.5030503273010254, + "learning_rate": 3.4754472782641802e-06, + "loss": 1.4222, + "step": 1109 + }, + { + "epoch": 1.140215716486903, + "grad_norm": 2.5090320110321045, + "learning_rate": 3.4735439665017136e-06, + "loss": 1.3981, + "step": 1110 + }, + { + "epoch": 1.140215716486903, + "eval_loss": 1.769408941268921, + "eval_runtime": 37.608, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 1110 + }, + { + "epoch": 1.1412429378531073, + "grad_norm": 2.575934648513794, + "learning_rate": 3.4716406547392465e-06, + "loss": 1.4074, + "step": 1111 + }, + { + "epoch": 1.1422701592193119, + "grad_norm": 2.6381359100341797, + "learning_rate": 3.46973734297678e-06, + "loss": 1.4162, + "step": 1112 + }, + { + "epoch": 1.1432973805855162, + "grad_norm": 2.7605674266815186, + "learning_rate": 3.4678340312143133e-06, + "loss": 1.4616, + "step": 1113 + }, + { + "epoch": 1.1443246019517206, + "grad_norm": 2.5548479557037354, + "learning_rate": 3.4659307194518466e-06, + "loss": 1.4511, + "step": 1114 + }, + { + "epoch": 1.145351823317925, + "grad_norm": 2.471168279647827, + "learning_rate": 3.46402740768938e-06, + "loss": 1.3777, + "step": 1115 + }, + { + "epoch": 1.1463790446841293, + "grad_norm": 2.4416346549987793, + "learning_rate": 3.462124095926913e-06, + "loss": 1.3805, + "step": 1116 + }, + { + "epoch": 1.147406266050334, + "grad_norm": 2.4972171783447266, + "learning_rate": 3.4602207841644463e-06, + "loss": 1.4282, + "step": 1117 + }, + { + "epoch": 1.1484334874165383, + "grad_norm": 2.4379162788391113, + "learning_rate": 3.4583174724019797e-06, + "loss": 1.4744, + "step": 1118 + }, + { + "epoch": 1.1494607087827426, + "grad_norm": 2.508047580718994, + "learning_rate": 3.456414160639513e-06, + "loss": 1.4303, + "step": 1119 + }, + { + "epoch": 1.150487930148947, + "grad_norm": 2.4299631118774414, + "learning_rate": 3.4545108488770464e-06, + "loss": 1.4179, + "step": 1120 + }, + { + "epoch": 1.150487930148947, + "eval_loss": 1.7768158912658691, + "eval_runtime": 37.736, + "eval_samples_per_second": 26.5, + "eval_steps_per_second": 1.113, + "step": 1120 + }, + { + "epoch": 1.1515151515151516, + "grad_norm": 2.41538143157959, + "learning_rate": 3.4526075371145793e-06, + "loss": 1.4235, + "step": 1121 + }, + { + "epoch": 1.152542372881356, + "grad_norm": 2.5614655017852783, + "learning_rate": 3.4507042253521127e-06, + "loss": 1.4879, + "step": 1122 + }, + { + "epoch": 1.1535695942475603, + "grad_norm": 2.4826605319976807, + "learning_rate": 3.448800913589646e-06, + "loss": 1.3968, + "step": 1123 + }, + { + "epoch": 1.1545968156137647, + "grad_norm": 2.5341391563415527, + "learning_rate": 3.4468976018271794e-06, + "loss": 1.4903, + "step": 1124 + }, + { + "epoch": 1.1556240369799693, + "grad_norm": 2.493818998336792, + "learning_rate": 3.4449942900647128e-06, + "loss": 1.4466, + "step": 1125 + }, + { + "epoch": 1.1566512583461737, + "grad_norm": 2.6057965755462646, + "learning_rate": 3.443090978302246e-06, + "loss": 1.4425, + "step": 1126 + }, + { + "epoch": 1.157678479712378, + "grad_norm": 2.5481390953063965, + "learning_rate": 3.44118766653978e-06, + "loss": 1.4296, + "step": 1127 + }, + { + "epoch": 1.1587057010785824, + "grad_norm": 2.5753650665283203, + "learning_rate": 3.439284354777313e-06, + "loss": 1.4408, + "step": 1128 + }, + { + "epoch": 1.1597329224447868, + "grad_norm": 2.5423803329467773, + "learning_rate": 3.4373810430148462e-06, + "loss": 1.373, + "step": 1129 + }, + { + "epoch": 1.1607601438109914, + "grad_norm": 2.593487024307251, + "learning_rate": 3.4354777312523796e-06, + "loss": 1.3992, + "step": 1130 + }, + { + "epoch": 1.1607601438109914, + "eval_loss": 1.7784411907196045, + "eval_runtime": 37.8687, + "eval_samples_per_second": 26.407, + "eval_steps_per_second": 1.109, + "step": 1130 + }, + { + "epoch": 1.1617873651771957, + "grad_norm": 2.4693503379821777, + "learning_rate": 3.433574419489913e-06, + "loss": 1.4339, + "step": 1131 + }, + { + "epoch": 1.1628145865434, + "grad_norm": 2.5477423667907715, + "learning_rate": 3.4316711077274463e-06, + "loss": 1.4035, + "step": 1132 + }, + { + "epoch": 1.1638418079096045, + "grad_norm": 2.5650672912597656, + "learning_rate": 3.4297677959649792e-06, + "loss": 1.4517, + "step": 1133 + }, + { + "epoch": 1.1648690292758088, + "grad_norm": 2.3217973709106445, + "learning_rate": 3.4278644842025126e-06, + "loss": 1.3983, + "step": 1134 + }, + { + "epoch": 1.1658962506420134, + "grad_norm": 2.674342155456543, + "learning_rate": 3.425961172440046e-06, + "loss": 1.4799, + "step": 1135 + }, + { + "epoch": 1.1669234720082178, + "grad_norm": 2.4861397743225098, + "learning_rate": 3.4240578606775793e-06, + "loss": 1.414, + "step": 1136 + }, + { + "epoch": 1.1679506933744221, + "grad_norm": 2.4306106567382812, + "learning_rate": 3.4221545489151127e-06, + "loss": 1.3918, + "step": 1137 + }, + { + "epoch": 1.1689779147406265, + "grad_norm": 2.6873955726623535, + "learning_rate": 3.4202512371526456e-06, + "loss": 1.4979, + "step": 1138 + }, + { + "epoch": 1.170005136106831, + "grad_norm": 2.559042453765869, + "learning_rate": 3.418347925390179e-06, + "loss": 1.4424, + "step": 1139 + }, + { + "epoch": 1.1710323574730355, + "grad_norm": 2.397489547729492, + "learning_rate": 3.4164446136277124e-06, + "loss": 1.418, + "step": 1140 + }, + { + "epoch": 1.1710323574730355, + "eval_loss": 1.7702736854553223, + "eval_runtime": 37.8963, + "eval_samples_per_second": 26.388, + "eval_steps_per_second": 1.108, + "step": 1140 + }, + { + "epoch": 1.1720595788392398, + "grad_norm": 2.4295787811279297, + "learning_rate": 3.4145413018652457e-06, + "loss": 1.4436, + "step": 1141 + }, + { + "epoch": 1.1730868002054442, + "grad_norm": 2.503202199935913, + "learning_rate": 3.412637990102779e-06, + "loss": 1.4598, + "step": 1142 + }, + { + "epoch": 1.1741140215716488, + "grad_norm": 2.540195941925049, + "learning_rate": 3.4107346783403125e-06, + "loss": 1.3704, + "step": 1143 + }, + { + "epoch": 1.1751412429378532, + "grad_norm": 2.4832522869110107, + "learning_rate": 3.4088313665778454e-06, + "loss": 1.4081, + "step": 1144 + }, + { + "epoch": 1.1761684643040575, + "grad_norm": 2.5639121532440186, + "learning_rate": 3.4069280548153788e-06, + "loss": 1.3866, + "step": 1145 + }, + { + "epoch": 1.177195685670262, + "grad_norm": 2.526610851287842, + "learning_rate": 3.405024743052912e-06, + "loss": 1.4181, + "step": 1146 + }, + { + "epoch": 1.1782229070364663, + "grad_norm": 2.6945254802703857, + "learning_rate": 3.403121431290446e-06, + "loss": 1.3871, + "step": 1147 + }, + { + "epoch": 1.1792501284026708, + "grad_norm": 2.6561484336853027, + "learning_rate": 3.4012181195279793e-06, + "loss": 1.3987, + "step": 1148 + }, + { + "epoch": 1.1802773497688752, + "grad_norm": 2.6104650497436523, + "learning_rate": 3.3993148077655126e-06, + "loss": 1.4057, + "step": 1149 + }, + { + "epoch": 1.1813045711350796, + "grad_norm": 2.571707248687744, + "learning_rate": 3.3974114960030456e-06, + "loss": 1.3606, + "step": 1150 + }, + { + "epoch": 1.1813045711350796, + "eval_loss": 1.7780840396881104, + "eval_runtime": 37.8616, + "eval_samples_per_second": 26.412, + "eval_steps_per_second": 1.109, + "step": 1150 + }, + { + "epoch": 1.182331792501284, + "grad_norm": 2.5804402828216553, + "learning_rate": 3.395508184240579e-06, + "loss": 1.3876, + "step": 1151 + }, + { + "epoch": 1.1833590138674885, + "grad_norm": 2.832446336746216, + "learning_rate": 3.3936048724781123e-06, + "loss": 1.4674, + "step": 1152 + }, + { + "epoch": 1.184386235233693, + "grad_norm": 2.6478688716888428, + "learning_rate": 3.3917015607156457e-06, + "loss": 1.3643, + "step": 1153 + }, + { + "epoch": 1.1854134565998973, + "grad_norm": 2.579416513442993, + "learning_rate": 3.389798248953179e-06, + "loss": 1.3874, + "step": 1154 + }, + { + "epoch": 1.1864406779661016, + "grad_norm": 2.487861156463623, + "learning_rate": 3.387894937190712e-06, + "loss": 1.4323, + "step": 1155 + }, + { + "epoch": 1.1874678993323062, + "grad_norm": 2.5020716190338135, + "learning_rate": 3.3859916254282453e-06, + "loss": 1.423, + "step": 1156 + }, + { + "epoch": 1.1884951206985106, + "grad_norm": 2.6588852405548096, + "learning_rate": 3.3840883136657787e-06, + "loss": 1.4808, + "step": 1157 + }, + { + "epoch": 1.189522342064715, + "grad_norm": 2.5702178478240967, + "learning_rate": 3.382185001903312e-06, + "loss": 1.4007, + "step": 1158 + }, + { + "epoch": 1.1905495634309193, + "grad_norm": 2.5082061290740967, + "learning_rate": 3.3802816901408454e-06, + "loss": 1.4278, + "step": 1159 + }, + { + "epoch": 1.1915767847971237, + "grad_norm": 2.4165053367614746, + "learning_rate": 3.3783783783783788e-06, + "loss": 1.4088, + "step": 1160 + }, + { + "epoch": 1.1915767847971237, + "eval_loss": 1.7722209692001343, + "eval_runtime": 37.8843, + "eval_samples_per_second": 26.396, + "eval_steps_per_second": 1.109, + "step": 1160 + }, + { + "epoch": 1.1926040061633283, + "grad_norm": 2.4510512351989746, + "learning_rate": 3.3764750666159117e-06, + "loss": 1.4089, + "step": 1161 + }, + { + "epoch": 1.1936312275295327, + "grad_norm": 2.5006556510925293, + "learning_rate": 3.374571754853445e-06, + "loss": 1.3988, + "step": 1162 + }, + { + "epoch": 1.194658448895737, + "grad_norm": 2.416264057159424, + "learning_rate": 3.3726684430909784e-06, + "loss": 1.3576, + "step": 1163 + }, + { + "epoch": 1.1956856702619414, + "grad_norm": 2.465613842010498, + "learning_rate": 3.370765131328512e-06, + "loss": 1.3796, + "step": 1164 + }, + { + "epoch": 1.1967128916281458, + "grad_norm": 2.510542154312134, + "learning_rate": 3.368861819566045e-06, + "loss": 1.391, + "step": 1165 + }, + { + "epoch": 1.1977401129943503, + "grad_norm": 2.729495048522949, + "learning_rate": 3.366958507803578e-06, + "loss": 1.431, + "step": 1166 + }, + { + "epoch": 1.1987673343605547, + "grad_norm": 2.4473495483398438, + "learning_rate": 3.365055196041112e-06, + "loss": 1.376, + "step": 1167 + }, + { + "epoch": 1.199794555726759, + "grad_norm": 2.3869211673736572, + "learning_rate": 3.3631518842786452e-06, + "loss": 1.4498, + "step": 1168 + }, + { + "epoch": 1.2008217770929634, + "grad_norm": 2.4202256202697754, + "learning_rate": 3.3612485725161786e-06, + "loss": 1.3754, + "step": 1169 + }, + { + "epoch": 1.201848998459168, + "grad_norm": 2.452707529067993, + "learning_rate": 3.359345260753712e-06, + "loss": 1.4647, + "step": 1170 + }, + { + "epoch": 1.201848998459168, + "eval_loss": 1.769252896308899, + "eval_runtime": 37.9385, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1170 + }, + { + "epoch": 1.2028762198253724, + "grad_norm": 2.56616473197937, + "learning_rate": 3.3574419489912453e-06, + "loss": 1.443, + "step": 1171 + }, + { + "epoch": 1.2039034411915768, + "grad_norm": 2.6483540534973145, + "learning_rate": 3.3555386372287783e-06, + "loss": 1.4755, + "step": 1172 + }, + { + "epoch": 1.2049306625577811, + "grad_norm": 2.5579090118408203, + "learning_rate": 3.3536353254663116e-06, + "loss": 1.3732, + "step": 1173 + }, + { + "epoch": 1.2059578839239857, + "grad_norm": 2.5310683250427246, + "learning_rate": 3.351732013703845e-06, + "loss": 1.4077, + "step": 1174 + }, + { + "epoch": 1.20698510529019, + "grad_norm": 2.616957187652588, + "learning_rate": 3.3498287019413784e-06, + "loss": 1.4237, + "step": 1175 + }, + { + "epoch": 1.2080123266563945, + "grad_norm": 2.6008999347686768, + "learning_rate": 3.3479253901789117e-06, + "loss": 1.4381, + "step": 1176 + }, + { + "epoch": 1.2090395480225988, + "grad_norm": 2.5235824584960938, + "learning_rate": 3.346022078416445e-06, + "loss": 1.3471, + "step": 1177 + }, + { + "epoch": 1.2100667693888032, + "grad_norm": 2.3749704360961914, + "learning_rate": 3.344118766653978e-06, + "loss": 1.417, + "step": 1178 + }, + { + "epoch": 1.2110939907550078, + "grad_norm": 2.493380069732666, + "learning_rate": 3.3422154548915114e-06, + "loss": 1.3924, + "step": 1179 + }, + { + "epoch": 1.2121212121212122, + "grad_norm": 2.4233086109161377, + "learning_rate": 3.3403121431290448e-06, + "loss": 1.4257, + "step": 1180 + }, + { + "epoch": 1.2121212121212122, + "eval_loss": 1.7660146951675415, + "eval_runtime": 37.9268, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 1180 + }, + { + "epoch": 1.2131484334874165, + "grad_norm": 2.535982608795166, + "learning_rate": 3.338408831366578e-06, + "loss": 1.3587, + "step": 1181 + }, + { + "epoch": 1.214175654853621, + "grad_norm": 2.5744845867156982, + "learning_rate": 3.3365055196041115e-06, + "loss": 1.436, + "step": 1182 + }, + { + "epoch": 1.2152028762198253, + "grad_norm": 2.5544042587280273, + "learning_rate": 3.3346022078416444e-06, + "loss": 1.464, + "step": 1183 + }, + { + "epoch": 1.2162300975860298, + "grad_norm": 2.4980483055114746, + "learning_rate": 3.3326988960791778e-06, + "loss": 1.3915, + "step": 1184 + }, + { + "epoch": 1.2172573189522342, + "grad_norm": 2.403468608856201, + "learning_rate": 3.330795584316711e-06, + "loss": 1.3239, + "step": 1185 + }, + { + "epoch": 1.2182845403184386, + "grad_norm": 2.6861846446990967, + "learning_rate": 3.3288922725542445e-06, + "loss": 1.4013, + "step": 1186 + }, + { + "epoch": 1.219311761684643, + "grad_norm": 2.525033712387085, + "learning_rate": 3.326988960791778e-06, + "loss": 1.4135, + "step": 1187 + }, + { + "epoch": 1.2203389830508475, + "grad_norm": 2.6705894470214844, + "learning_rate": 3.3250856490293117e-06, + "loss": 1.4587, + "step": 1188 + }, + { + "epoch": 1.221366204417052, + "grad_norm": 2.6108217239379883, + "learning_rate": 3.323182337266845e-06, + "loss": 1.4482, + "step": 1189 + }, + { + "epoch": 1.2223934257832563, + "grad_norm": 2.5227413177490234, + "learning_rate": 3.321279025504378e-06, + "loss": 1.432, + "step": 1190 + }, + { + "epoch": 1.2223934257832563, + "eval_loss": 1.756821870803833, + "eval_runtime": 37.9929, + "eval_samples_per_second": 26.321, + "eval_steps_per_second": 1.105, + "step": 1190 + }, + { + "epoch": 1.2234206471494606, + "grad_norm": 2.469766139984131, + "learning_rate": 3.3193757137419113e-06, + "loss": 1.4215, + "step": 1191 + }, + { + "epoch": 1.2244478685156652, + "grad_norm": 2.5257861614227295, + "learning_rate": 3.3174724019794447e-06, + "loss": 1.4624, + "step": 1192 + }, + { + "epoch": 1.2254750898818696, + "grad_norm": 2.5281760692596436, + "learning_rate": 3.315569090216978e-06, + "loss": 1.3799, + "step": 1193 + }, + { + "epoch": 1.226502311248074, + "grad_norm": 2.701521396636963, + "learning_rate": 3.3136657784545114e-06, + "loss": 1.4631, + "step": 1194 + }, + { + "epoch": 1.2275295326142783, + "grad_norm": 2.4239444732666016, + "learning_rate": 3.3117624666920443e-06, + "loss": 1.3746, + "step": 1195 + }, + { + "epoch": 1.2285567539804827, + "grad_norm": 2.430377721786499, + "learning_rate": 3.3098591549295777e-06, + "loss": 1.3584, + "step": 1196 + }, + { + "epoch": 1.2295839753466873, + "grad_norm": 2.7177658081054688, + "learning_rate": 3.307955843167111e-06, + "loss": 1.4301, + "step": 1197 + }, + { + "epoch": 1.2306111967128917, + "grad_norm": 2.512943983078003, + "learning_rate": 3.3060525314046444e-06, + "loss": 1.4111, + "step": 1198 + }, + { + "epoch": 1.231638418079096, + "grad_norm": 2.5269784927368164, + "learning_rate": 3.304149219642178e-06, + "loss": 1.4096, + "step": 1199 + }, + { + "epoch": 1.2326656394453004, + "grad_norm": 2.476409673690796, + "learning_rate": 3.3022459078797107e-06, + "loss": 1.4322, + "step": 1200 + }, + { + "epoch": 1.2326656394453004, + "eval_loss": 1.7645230293273926, + "eval_runtime": 37.9898, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1200 + }, + { + "epoch": 1.2336928608115048, + "grad_norm": 2.5339248180389404, + "learning_rate": 3.300342596117244e-06, + "loss": 1.4093, + "step": 1201 + }, + { + "epoch": 1.2347200821777093, + "grad_norm": 2.3505656719207764, + "learning_rate": 3.2984392843547775e-06, + "loss": 1.3623, + "step": 1202 + }, + { + "epoch": 1.2357473035439137, + "grad_norm": 2.400259256362915, + "learning_rate": 3.296535972592311e-06, + "loss": 1.3512, + "step": 1203 + }, + { + "epoch": 1.236774524910118, + "grad_norm": 2.561558723449707, + "learning_rate": 3.294632660829844e-06, + "loss": 1.3946, + "step": 1204 + }, + { + "epoch": 1.2378017462763227, + "grad_norm": 2.623100519180298, + "learning_rate": 3.292729349067377e-06, + "loss": 1.3972, + "step": 1205 + }, + { + "epoch": 1.238828967642527, + "grad_norm": 2.4889469146728516, + "learning_rate": 3.2908260373049105e-06, + "loss": 1.381, + "step": 1206 + }, + { + "epoch": 1.2398561890087314, + "grad_norm": 2.457334518432617, + "learning_rate": 3.288922725542444e-06, + "loss": 1.4345, + "step": 1207 + }, + { + "epoch": 1.2408834103749358, + "grad_norm": 2.440988540649414, + "learning_rate": 3.2870194137799776e-06, + "loss": 1.3901, + "step": 1208 + }, + { + "epoch": 1.2419106317411401, + "grad_norm": 2.4992308616638184, + "learning_rate": 3.285116102017511e-06, + "loss": 1.3694, + "step": 1209 + }, + { + "epoch": 1.2429378531073447, + "grad_norm": 2.611128091812134, + "learning_rate": 3.2832127902550444e-06, + "loss": 1.4344, + "step": 1210 + }, + { + "epoch": 1.2429378531073447, + "eval_loss": 1.763407826423645, + "eval_runtime": 37.7142, + "eval_samples_per_second": 26.515, + "eval_steps_per_second": 1.114, + "step": 1210 + }, + { + "epoch": 1.243965074473549, + "grad_norm": 2.6573522090911865, + "learning_rate": 3.2813094784925777e-06, + "loss": 1.4204, + "step": 1211 + }, + { + "epoch": 1.2449922958397535, + "grad_norm": 2.4785778522491455, + "learning_rate": 3.2794061667301107e-06, + "loss": 1.3617, + "step": 1212 + }, + { + "epoch": 1.2460195172059578, + "grad_norm": 2.4831974506378174, + "learning_rate": 3.277502854967644e-06, + "loss": 1.4004, + "step": 1213 + }, + { + "epoch": 1.2470467385721622, + "grad_norm": 2.6142327785491943, + "learning_rate": 3.2755995432051774e-06, + "loss": 1.3881, + "step": 1214 + }, + { + "epoch": 1.2480739599383668, + "grad_norm": 2.447962999343872, + "learning_rate": 3.2736962314427107e-06, + "loss": 1.3984, + "step": 1215 + }, + { + "epoch": 1.2491011813045712, + "grad_norm": 2.3895885944366455, + "learning_rate": 3.271792919680244e-06, + "loss": 1.4099, + "step": 1216 + }, + { + "epoch": 1.2501284026707755, + "grad_norm": 2.599364995956421, + "learning_rate": 3.269889607917777e-06, + "loss": 1.4575, + "step": 1217 + }, + { + "epoch": 1.25115562403698, + "grad_norm": 2.6294758319854736, + "learning_rate": 3.2679862961553104e-06, + "loss": 1.3752, + "step": 1218 + }, + { + "epoch": 1.2521828454031843, + "grad_norm": 2.50923228263855, + "learning_rate": 3.2660829843928438e-06, + "loss": 1.4017, + "step": 1219 + }, + { + "epoch": 1.2532100667693888, + "grad_norm": 2.626434564590454, + "learning_rate": 3.264179672630377e-06, + "loss": 1.4505, + "step": 1220 + }, + { + "epoch": 1.2532100667693888, + "eval_loss": 1.7672324180603027, + "eval_runtime": 37.8605, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 1220 + }, + { + "epoch": 1.2542372881355932, + "grad_norm": 2.541663885116577, + "learning_rate": 3.2622763608679105e-06, + "loss": 1.4129, + "step": 1221 + }, + { + "epoch": 1.2552645095017976, + "grad_norm": 2.559877395629883, + "learning_rate": 3.2603730491054434e-06, + "loss": 1.3793, + "step": 1222 + }, + { + "epoch": 1.2562917308680022, + "grad_norm": 2.4085795879364014, + "learning_rate": 3.258469737342977e-06, + "loss": 1.3895, + "step": 1223 + }, + { + "epoch": 1.2573189522342065, + "grad_norm": 2.2942678928375244, + "learning_rate": 3.25656642558051e-06, + "loss": 1.3713, + "step": 1224 + }, + { + "epoch": 1.258346173600411, + "grad_norm": 2.561086893081665, + "learning_rate": 3.2546631138180435e-06, + "loss": 1.3835, + "step": 1225 + }, + { + "epoch": 1.2593733949666153, + "grad_norm": 2.4947667121887207, + "learning_rate": 3.252759802055577e-06, + "loss": 1.3793, + "step": 1226 + }, + { + "epoch": 1.2604006163328196, + "grad_norm": 2.6293253898620605, + "learning_rate": 3.25085649029311e-06, + "loss": 1.3935, + "step": 1227 + }, + { + "epoch": 1.2614278376990242, + "grad_norm": 2.5461983680725098, + "learning_rate": 3.248953178530644e-06, + "loss": 1.4201, + "step": 1228 + }, + { + "epoch": 1.2624550590652286, + "grad_norm": 2.381537675857544, + "learning_rate": 3.247049866768177e-06, + "loss": 1.3932, + "step": 1229 + }, + { + "epoch": 1.263482280431433, + "grad_norm": 2.381558418273926, + "learning_rate": 3.2451465550057103e-06, + "loss": 1.3425, + "step": 1230 + }, + { + "epoch": 1.263482280431433, + "eval_loss": 1.7713288068771362, + "eval_runtime": 37.9629, + "eval_samples_per_second": 26.342, + "eval_steps_per_second": 1.106, + "step": 1230 + }, + { + "epoch": 1.2645095017976373, + "grad_norm": 2.5877468585968018, + "learning_rate": 3.2432432432432437e-06, + "loss": 1.3917, + "step": 1231 + }, + { + "epoch": 1.2655367231638417, + "grad_norm": 2.6057848930358887, + "learning_rate": 3.241339931480777e-06, + "loss": 1.4217, + "step": 1232 + }, + { + "epoch": 1.2665639445300463, + "grad_norm": 2.5157413482666016, + "learning_rate": 3.2394366197183104e-06, + "loss": 1.427, + "step": 1233 + }, + { + "epoch": 1.2675911658962506, + "grad_norm": 2.4153692722320557, + "learning_rate": 3.2375333079558434e-06, + "loss": 1.3822, + "step": 1234 + }, + { + "epoch": 1.268618387262455, + "grad_norm": 2.5459582805633545, + "learning_rate": 3.2356299961933767e-06, + "loss": 1.4152, + "step": 1235 + }, + { + "epoch": 1.2696456086286596, + "grad_norm": 2.606468915939331, + "learning_rate": 3.23372668443091e-06, + "loss": 1.4278, + "step": 1236 + }, + { + "epoch": 1.270672829994864, + "grad_norm": 2.5061633586883545, + "learning_rate": 3.2318233726684435e-06, + "loss": 1.4512, + "step": 1237 + }, + { + "epoch": 1.2717000513610683, + "grad_norm": 2.3693087100982666, + "learning_rate": 3.229920060905977e-06, + "loss": 1.3791, + "step": 1238 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 2.479583978652954, + "learning_rate": 3.2280167491435098e-06, + "loss": 1.3814, + "step": 1239 + }, + { + "epoch": 1.273754494093477, + "grad_norm": 2.3258883953094482, + "learning_rate": 3.226113437381043e-06, + "loss": 1.3904, + "step": 1240 + }, + { + "epoch": 1.273754494093477, + "eval_loss": 1.7602728605270386, + "eval_runtime": 37.9604, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 1240 + }, + { + "epoch": 1.2747817154596817, + "grad_norm": 2.4170308113098145, + "learning_rate": 3.2242101256185765e-06, + "loss": 1.4045, + "step": 1241 + }, + { + "epoch": 1.275808936825886, + "grad_norm": 2.578066825866699, + "learning_rate": 3.22230681385611e-06, + "loss": 1.4556, + "step": 1242 + }, + { + "epoch": 1.2768361581920904, + "grad_norm": 2.558187246322632, + "learning_rate": 3.220403502093643e-06, + "loss": 1.4036, + "step": 1243 + }, + { + "epoch": 1.2778633795582948, + "grad_norm": 2.598916530609131, + "learning_rate": 3.218500190331176e-06, + "loss": 1.4462, + "step": 1244 + }, + { + "epoch": 1.2788906009244991, + "grad_norm": 2.4528257846832275, + "learning_rate": 3.2165968785687095e-06, + "loss": 1.3855, + "step": 1245 + }, + { + "epoch": 1.2799178222907037, + "grad_norm": 2.5937459468841553, + "learning_rate": 3.214693566806243e-06, + "loss": 1.4191, + "step": 1246 + }, + { + "epoch": 1.280945043656908, + "grad_norm": 2.5324676036834717, + "learning_rate": 3.2127902550437762e-06, + "loss": 1.4162, + "step": 1247 + }, + { + "epoch": 1.2819722650231125, + "grad_norm": 2.4849791526794434, + "learning_rate": 3.21088694328131e-06, + "loss": 1.4037, + "step": 1248 + }, + { + "epoch": 1.2829994863893168, + "grad_norm": 2.389582633972168, + "learning_rate": 3.2089836315188434e-06, + "loss": 1.3966, + "step": 1249 + }, + { + "epoch": 1.2840267077555212, + "grad_norm": 2.387909173965454, + "learning_rate": 3.2070803197563767e-06, + "loss": 1.396, + "step": 1250 + }, + { + "epoch": 1.2840267077555212, + "eval_loss": 1.7642556428909302, + "eval_runtime": 37.9446, + "eval_samples_per_second": 26.354, + "eval_steps_per_second": 1.107, + "step": 1250 + }, + { + "epoch": 1.2850539291217258, + "grad_norm": 2.5014612674713135, + "learning_rate": 3.2051770079939097e-06, + "loss": 1.4128, + "step": 1251 + }, + { + "epoch": 1.2860811504879301, + "grad_norm": 2.497950553894043, + "learning_rate": 3.203273696231443e-06, + "loss": 1.4412, + "step": 1252 + }, + { + "epoch": 1.2871083718541345, + "grad_norm": 2.4371297359466553, + "learning_rate": 3.2013703844689764e-06, + "loss": 1.3754, + "step": 1253 + }, + { + "epoch": 1.288135593220339, + "grad_norm": 2.4336745738983154, + "learning_rate": 3.1994670727065098e-06, + "loss": 1.4258, + "step": 1254 + }, + { + "epoch": 1.2891628145865435, + "grad_norm": 2.5165536403656006, + "learning_rate": 3.197563760944043e-06, + "loss": 1.4414, + "step": 1255 + }, + { + "epoch": 1.2901900359527478, + "grad_norm": 2.5060458183288574, + "learning_rate": 3.195660449181576e-06, + "loss": 1.4336, + "step": 1256 + }, + { + "epoch": 1.2912172573189522, + "grad_norm": 2.584284543991089, + "learning_rate": 3.1937571374191094e-06, + "loss": 1.3768, + "step": 1257 + }, + { + "epoch": 1.2922444786851566, + "grad_norm": 2.615629196166992, + "learning_rate": 3.191853825656643e-06, + "loss": 1.401, + "step": 1258 + }, + { + "epoch": 1.2932717000513612, + "grad_norm": 2.5989596843719482, + "learning_rate": 3.189950513894176e-06, + "loss": 1.3915, + "step": 1259 + }, + { + "epoch": 1.2942989214175655, + "grad_norm": 2.5507843494415283, + "learning_rate": 3.1880472021317095e-06, + "loss": 1.3979, + "step": 1260 + }, + { + "epoch": 1.2942989214175655, + "eval_loss": 1.7715612649917603, + "eval_runtime": 37.9859, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1260 + }, + { + "epoch": 1.29532614278377, + "grad_norm": 2.512140989303589, + "learning_rate": 3.1861438903692425e-06, + "loss": 1.3523, + "step": 1261 + }, + { + "epoch": 1.2963533641499743, + "grad_norm": 2.4311885833740234, + "learning_rate": 3.184240578606776e-06, + "loss": 1.3828, + "step": 1262 + }, + { + "epoch": 1.2973805855161786, + "grad_norm": 2.439265489578247, + "learning_rate": 3.182337266844309e-06, + "loss": 1.402, + "step": 1263 + }, + { + "epoch": 1.2984078068823832, + "grad_norm": 2.5773651599884033, + "learning_rate": 3.1804339550818426e-06, + "loss": 1.4362, + "step": 1264 + }, + { + "epoch": 1.2994350282485876, + "grad_norm": 2.456021547317505, + "learning_rate": 3.178530643319376e-06, + "loss": 1.3918, + "step": 1265 + }, + { + "epoch": 1.300462249614792, + "grad_norm": 2.44404673576355, + "learning_rate": 3.176627331556909e-06, + "loss": 1.3662, + "step": 1266 + }, + { + "epoch": 1.3014894709809965, + "grad_norm": 2.778027057647705, + "learning_rate": 3.1747240197944422e-06, + "loss": 1.4112, + "step": 1267 + }, + { + "epoch": 1.3025166923472007, + "grad_norm": 2.476496934890747, + "learning_rate": 3.172820708031976e-06, + "loss": 1.3654, + "step": 1268 + }, + { + "epoch": 1.3035439137134053, + "grad_norm": 2.6415998935699463, + "learning_rate": 3.1709173962695094e-06, + "loss": 1.4441, + "step": 1269 + }, + { + "epoch": 1.3045711350796096, + "grad_norm": 2.5046234130859375, + "learning_rate": 3.1690140845070427e-06, + "loss": 1.4316, + "step": 1270 + }, + { + "epoch": 1.3045711350796096, + "eval_loss": 1.7736135721206665, + "eval_runtime": 37.9392, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1270 + }, + { + "epoch": 1.305598356445814, + "grad_norm": 2.6234474182128906, + "learning_rate": 3.167110772744576e-06, + "loss": 1.4, + "step": 1271 + }, + { + "epoch": 1.3066255778120186, + "grad_norm": 2.488297700881958, + "learning_rate": 3.1652074609821095e-06, + "loss": 1.3908, + "step": 1272 + }, + { + "epoch": 1.307652799178223, + "grad_norm": 2.3759512901306152, + "learning_rate": 3.1633041492196424e-06, + "loss": 1.3908, + "step": 1273 + }, + { + "epoch": 1.3086800205444273, + "grad_norm": 2.487281084060669, + "learning_rate": 3.1614008374571758e-06, + "loss": 1.4217, + "step": 1274 + }, + { + "epoch": 1.3097072419106317, + "grad_norm": 2.5366780757904053, + "learning_rate": 3.159497525694709e-06, + "loss": 1.4229, + "step": 1275 + }, + { + "epoch": 1.310734463276836, + "grad_norm": 2.416141986846924, + "learning_rate": 3.1575942139322425e-06, + "loss": 1.3865, + "step": 1276 + }, + { + "epoch": 1.3117616846430407, + "grad_norm": 2.537964344024658, + "learning_rate": 3.155690902169776e-06, + "loss": 1.4141, + "step": 1277 + }, + { + "epoch": 1.312788906009245, + "grad_norm": 2.481692314147949, + "learning_rate": 3.1537875904073088e-06, + "loss": 1.383, + "step": 1278 + }, + { + "epoch": 1.3138161273754494, + "grad_norm": 2.4898593425750732, + "learning_rate": 3.151884278644842e-06, + "loss": 1.4078, + "step": 1279 + }, + { + "epoch": 1.3148433487416538, + "grad_norm": 2.4692323207855225, + "learning_rate": 3.1499809668823755e-06, + "loss": 1.3818, + "step": 1280 + }, + { + "epoch": 1.3148433487416538, + "eval_loss": 1.7660051584243774, + "eval_runtime": 37.8922, + "eval_samples_per_second": 26.391, + "eval_steps_per_second": 1.108, + "step": 1280 + }, + { + "epoch": 1.3158705701078581, + "grad_norm": 2.475862503051758, + "learning_rate": 3.148077655119909e-06, + "loss": 1.4384, + "step": 1281 + }, + { + "epoch": 1.3168977914740627, + "grad_norm": 2.2839648723602295, + "learning_rate": 3.1461743433574422e-06, + "loss": 1.3722, + "step": 1282 + }, + { + "epoch": 1.317925012840267, + "grad_norm": 2.3965585231781006, + "learning_rate": 3.144271031594975e-06, + "loss": 1.4262, + "step": 1283 + }, + { + "epoch": 1.3189522342064715, + "grad_norm": 2.4107236862182617, + "learning_rate": 3.1423677198325085e-06, + "loss": 1.4216, + "step": 1284 + }, + { + "epoch": 1.319979455572676, + "grad_norm": 2.438899040222168, + "learning_rate": 3.140464408070042e-06, + "loss": 1.383, + "step": 1285 + }, + { + "epoch": 1.3210066769388804, + "grad_norm": 2.4146060943603516, + "learning_rate": 3.1385610963075753e-06, + "loss": 1.3827, + "step": 1286 + }, + { + "epoch": 1.3220338983050848, + "grad_norm": 2.4193496704101562, + "learning_rate": 3.1366577845451086e-06, + "loss": 1.3707, + "step": 1287 + }, + { + "epoch": 1.3230611196712891, + "grad_norm": 2.5284323692321777, + "learning_rate": 3.134754472782642e-06, + "loss": 1.3623, + "step": 1288 + }, + { + "epoch": 1.3240883410374935, + "grad_norm": 2.421509265899658, + "learning_rate": 3.1328511610201758e-06, + "loss": 1.35, + "step": 1289 + }, + { + "epoch": 1.325115562403698, + "grad_norm": 2.3964431285858154, + "learning_rate": 3.1309478492577087e-06, + "loss": 1.3486, + "step": 1290 + }, + { + "epoch": 1.325115562403698, + "eval_loss": 1.7699168920516968, + "eval_runtime": 37.9003, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 1290 + }, + { + "epoch": 1.3261427837699025, + "grad_norm": 2.4371817111968994, + "learning_rate": 3.129044537495242e-06, + "loss": 1.3958, + "step": 1291 + }, + { + "epoch": 1.3271700051361068, + "grad_norm": 2.45233154296875, + "learning_rate": 3.1271412257327754e-06, + "loss": 1.4175, + "step": 1292 + }, + { + "epoch": 1.3281972265023112, + "grad_norm": 2.314384937286377, + "learning_rate": 3.125237913970309e-06, + "loss": 1.3701, + "step": 1293 + }, + { + "epoch": 1.3292244478685156, + "grad_norm": 2.5603694915771484, + "learning_rate": 3.123334602207842e-06, + "loss": 1.3489, + "step": 1294 + }, + { + "epoch": 1.3302516692347202, + "grad_norm": 2.5919623374938965, + "learning_rate": 3.121431290445375e-06, + "loss": 1.4472, + "step": 1295 + }, + { + "epoch": 1.3312788906009245, + "grad_norm": 2.6912758350372314, + "learning_rate": 3.1195279786829085e-06, + "loss": 1.4917, + "step": 1296 + }, + { + "epoch": 1.332306111967129, + "grad_norm": 2.439089298248291, + "learning_rate": 3.117624666920442e-06, + "loss": 1.3776, + "step": 1297 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 2.5207953453063965, + "learning_rate": 3.115721355157975e-06, + "loss": 1.3784, + "step": 1298 + }, + { + "epoch": 1.3343605546995376, + "grad_norm": 2.507791519165039, + "learning_rate": 3.1138180433955085e-06, + "loss": 1.3346, + "step": 1299 + }, + { + "epoch": 1.3353877760657422, + "grad_norm": 2.450531482696533, + "learning_rate": 3.1119147316330415e-06, + "loss": 1.4151, + "step": 1300 + }, + { + "epoch": 1.3353877760657422, + "eval_loss": 1.7761608362197876, + "eval_runtime": 37.9041, + "eval_samples_per_second": 26.382, + "eval_steps_per_second": 1.108, + "step": 1300 + }, + { + "epoch": 1.3364149974319466, + "grad_norm": 2.488255262374878, + "learning_rate": 3.110011419870575e-06, + "loss": 1.4475, + "step": 1301 + }, + { + "epoch": 1.337442218798151, + "grad_norm": 2.5059807300567627, + "learning_rate": 3.1081081081081082e-06, + "loss": 1.3812, + "step": 1302 + }, + { + "epoch": 1.3384694401643555, + "grad_norm": 2.5579488277435303, + "learning_rate": 3.1062047963456416e-06, + "loss": 1.4134, + "step": 1303 + }, + { + "epoch": 1.33949666153056, + "grad_norm": 2.6170148849487305, + "learning_rate": 3.104301484583175e-06, + "loss": 1.459, + "step": 1304 + }, + { + "epoch": 1.3405238828967643, + "grad_norm": 2.5572125911712646, + "learning_rate": 3.1023981728207083e-06, + "loss": 1.4354, + "step": 1305 + }, + { + "epoch": 1.3415511042629686, + "grad_norm": 2.396883487701416, + "learning_rate": 3.1004948610582412e-06, + "loss": 1.4016, + "step": 1306 + }, + { + "epoch": 1.342578325629173, + "grad_norm": 2.5869390964508057, + "learning_rate": 3.0985915492957746e-06, + "loss": 1.3945, + "step": 1307 + }, + { + "epoch": 1.3436055469953776, + "grad_norm": 2.4271085262298584, + "learning_rate": 3.096688237533308e-06, + "loss": 1.3506, + "step": 1308 + }, + { + "epoch": 1.344632768361582, + "grad_norm": 2.398407220840454, + "learning_rate": 3.0947849257708418e-06, + "loss": 1.4475, + "step": 1309 + }, + { + "epoch": 1.3456599897277863, + "grad_norm": 2.441934823989868, + "learning_rate": 3.092881614008375e-06, + "loss": 1.3972, + "step": 1310 + }, + { + "epoch": 1.3456599897277863, + "eval_loss": 1.774910569190979, + "eval_runtime": 37.6105, + "eval_samples_per_second": 26.588, + "eval_steps_per_second": 1.117, + "step": 1310 + }, + { + "epoch": 1.3466872110939907, + "grad_norm": 2.5015370845794678, + "learning_rate": 3.0909783022459085e-06, + "loss": 1.4418, + "step": 1311 + }, + { + "epoch": 1.347714432460195, + "grad_norm": 2.4616997241973877, + "learning_rate": 3.0890749904834414e-06, + "loss": 1.3833, + "step": 1312 + }, + { + "epoch": 1.3487416538263997, + "grad_norm": 2.5993902683258057, + "learning_rate": 3.0871716787209748e-06, + "loss": 1.4385, + "step": 1313 + }, + { + "epoch": 1.349768875192604, + "grad_norm": 2.518195867538452, + "learning_rate": 3.085268366958508e-06, + "loss": 1.4127, + "step": 1314 + }, + { + "epoch": 1.3507960965588084, + "grad_norm": 2.545189380645752, + "learning_rate": 3.0833650551960415e-06, + "loss": 1.42, + "step": 1315 + }, + { + "epoch": 1.3518233179250128, + "grad_norm": 2.461818218231201, + "learning_rate": 3.081461743433575e-06, + "loss": 1.3796, + "step": 1316 + }, + { + "epoch": 1.3528505392912171, + "grad_norm": 2.4081408977508545, + "learning_rate": 3.079558431671108e-06, + "loss": 1.3939, + "step": 1317 + }, + { + "epoch": 1.3538777606574217, + "grad_norm": 2.2885539531707764, + "learning_rate": 3.077655119908641e-06, + "loss": 1.3645, + "step": 1318 + }, + { + "epoch": 1.354904982023626, + "grad_norm": 2.453791618347168, + "learning_rate": 3.0757518081461745e-06, + "loss": 1.3906, + "step": 1319 + }, + { + "epoch": 1.3559322033898304, + "grad_norm": 2.4538352489471436, + "learning_rate": 3.073848496383708e-06, + "loss": 1.4127, + "step": 1320 + }, + { + "epoch": 1.3559322033898304, + "eval_loss": 1.7821755409240723, + "eval_runtime": 37.7585, + "eval_samples_per_second": 26.484, + "eval_steps_per_second": 1.112, + "step": 1320 + }, + { + "epoch": 1.356959424756035, + "grad_norm": 2.438425064086914, + "learning_rate": 3.0719451846212413e-06, + "loss": 1.3763, + "step": 1321 + }, + { + "epoch": 1.3579866461222394, + "grad_norm": 2.4398279190063477, + "learning_rate": 3.0700418728587746e-06, + "loss": 1.3963, + "step": 1322 + }, + { + "epoch": 1.3590138674884438, + "grad_norm": 2.6763861179351807, + "learning_rate": 3.0681385610963076e-06, + "loss": 1.3837, + "step": 1323 + }, + { + "epoch": 1.3600410888546481, + "grad_norm": 2.685056209564209, + "learning_rate": 3.066235249333841e-06, + "loss": 1.4158, + "step": 1324 + }, + { + "epoch": 1.3610683102208525, + "grad_norm": 2.6663355827331543, + "learning_rate": 3.0643319375713743e-06, + "loss": 1.4533, + "step": 1325 + }, + { + "epoch": 1.362095531587057, + "grad_norm": 2.4613428115844727, + "learning_rate": 3.0624286258089076e-06, + "loss": 1.3927, + "step": 1326 + }, + { + "epoch": 1.3631227529532615, + "grad_norm": 2.348334789276123, + "learning_rate": 3.060525314046441e-06, + "loss": 1.3782, + "step": 1327 + }, + { + "epoch": 1.3641499743194658, + "grad_norm": 2.5011332035064697, + "learning_rate": 3.058622002283974e-06, + "loss": 1.4175, + "step": 1328 + }, + { + "epoch": 1.3651771956856702, + "grad_norm": 2.4692256450653076, + "learning_rate": 3.0567186905215077e-06, + "loss": 1.4319, + "step": 1329 + }, + { + "epoch": 1.3662044170518746, + "grad_norm": 2.616990089416504, + "learning_rate": 3.054815378759041e-06, + "loss": 1.4264, + "step": 1330 + }, + { + "epoch": 1.3662044170518746, + "eval_loss": 1.781583309173584, + "eval_runtime": 37.9291, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 1330 + }, + { + "epoch": 1.3672316384180792, + "grad_norm": 2.5416364669799805, + "learning_rate": 3.0529120669965745e-06, + "loss": 1.4249, + "step": 1331 + }, + { + "epoch": 1.3682588597842835, + "grad_norm": 2.5346689224243164, + "learning_rate": 3.051008755234108e-06, + "loss": 1.3647, + "step": 1332 + }, + { + "epoch": 1.3692860811504879, + "grad_norm": 2.585714340209961, + "learning_rate": 3.049105443471641e-06, + "loss": 1.3714, + "step": 1333 + }, + { + "epoch": 1.3703133025166925, + "grad_norm": 2.5958940982818604, + "learning_rate": 3.0472021317091745e-06, + "loss": 1.5041, + "step": 1334 + }, + { + "epoch": 1.3713405238828968, + "grad_norm": 2.4795925617218018, + "learning_rate": 3.0452988199467075e-06, + "loss": 1.4317, + "step": 1335 + }, + { + "epoch": 1.3723677452491012, + "grad_norm": 2.4608285427093506, + "learning_rate": 3.043395508184241e-06, + "loss": 1.4259, + "step": 1336 + }, + { + "epoch": 1.3733949666153056, + "grad_norm": 2.5970380306243896, + "learning_rate": 3.041492196421774e-06, + "loss": 1.3859, + "step": 1337 + }, + { + "epoch": 1.37442218798151, + "grad_norm": 2.517047166824341, + "learning_rate": 3.0395888846593076e-06, + "loss": 1.4086, + "step": 1338 + }, + { + "epoch": 1.3754494093477145, + "grad_norm": 2.4541358947753906, + "learning_rate": 3.037685572896841e-06, + "loss": 1.3959, + "step": 1339 + }, + { + "epoch": 1.376476630713919, + "grad_norm": 2.6143250465393066, + "learning_rate": 3.035782261134374e-06, + "loss": 1.4088, + "step": 1340 + }, + { + "epoch": 1.376476630713919, + "eval_loss": 1.7629026174545288, + "eval_runtime": 38.0156, + "eval_samples_per_second": 26.305, + "eval_steps_per_second": 1.105, + "step": 1340 + }, + { + "epoch": 1.3775038520801233, + "grad_norm": 2.763033151626587, + "learning_rate": 3.0338789493719072e-06, + "loss": 1.4104, + "step": 1341 + }, + { + "epoch": 1.3785310734463276, + "grad_norm": 2.523888349533081, + "learning_rate": 3.0319756376094406e-06, + "loss": 1.4073, + "step": 1342 + }, + { + "epoch": 1.379558294812532, + "grad_norm": 2.5393309593200684, + "learning_rate": 3.030072325846974e-06, + "loss": 1.4145, + "step": 1343 + }, + { + "epoch": 1.3805855161787366, + "grad_norm": 2.455733060836792, + "learning_rate": 3.0281690140845073e-06, + "loss": 1.3782, + "step": 1344 + }, + { + "epoch": 1.381612737544941, + "grad_norm": 2.600311517715454, + "learning_rate": 3.0262657023220403e-06, + "loss": 1.4158, + "step": 1345 + }, + { + "epoch": 1.3826399589111453, + "grad_norm": 2.453249931335449, + "learning_rate": 3.0243623905595736e-06, + "loss": 1.414, + "step": 1346 + }, + { + "epoch": 1.3836671802773497, + "grad_norm": 2.603515625, + "learning_rate": 3.022459078797107e-06, + "loss": 1.4102, + "step": 1347 + }, + { + "epoch": 1.384694401643554, + "grad_norm": 2.6515872478485107, + "learning_rate": 3.0205557670346404e-06, + "loss": 1.4063, + "step": 1348 + }, + { + "epoch": 1.3857216230097587, + "grad_norm": 2.412846565246582, + "learning_rate": 3.018652455272174e-06, + "loss": 1.3739, + "step": 1349 + }, + { + "epoch": 1.386748844375963, + "grad_norm": 2.748563051223755, + "learning_rate": 3.0167491435097075e-06, + "loss": 1.4909, + "step": 1350 + }, + { + "epoch": 1.386748844375963, + "eval_loss": 1.7653924226760864, + "eval_runtime": 37.9855, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1350 + }, + { + "epoch": 1.3877760657421674, + "grad_norm": 2.7861363887786865, + "learning_rate": 3.014845831747241e-06, + "loss": 1.4028, + "step": 1351 + }, + { + "epoch": 1.388803287108372, + "grad_norm": 2.5898189544677734, + "learning_rate": 3.012942519984774e-06, + "loss": 1.3793, + "step": 1352 + }, + { + "epoch": 1.3898305084745763, + "grad_norm": 2.4240355491638184, + "learning_rate": 3.011039208222307e-06, + "loss": 1.3889, + "step": 1353 + }, + { + "epoch": 1.3908577298407807, + "grad_norm": 2.4492504596710205, + "learning_rate": 3.0091358964598405e-06, + "loss": 1.3464, + "step": 1354 + }, + { + "epoch": 1.391884951206985, + "grad_norm": 2.4425718784332275, + "learning_rate": 3.007232584697374e-06, + "loss": 1.4073, + "step": 1355 + }, + { + "epoch": 1.3929121725731894, + "grad_norm": 2.5057594776153564, + "learning_rate": 3.0053292729349073e-06, + "loss": 1.4053, + "step": 1356 + }, + { + "epoch": 1.393939393939394, + "grad_norm": 2.5386273860931396, + "learning_rate": 3.00342596117244e-06, + "loss": 1.3991, + "step": 1357 + }, + { + "epoch": 1.3949666153055984, + "grad_norm": 2.6289730072021484, + "learning_rate": 3.0015226494099736e-06, + "loss": 1.3971, + "step": 1358 + }, + { + "epoch": 1.3959938366718028, + "grad_norm": 2.5091774463653564, + "learning_rate": 2.999619337647507e-06, + "loss": 1.3815, + "step": 1359 + }, + { + "epoch": 1.3970210580380071, + "grad_norm": 2.567659616470337, + "learning_rate": 2.9977160258850403e-06, + "loss": 1.3786, + "step": 1360 + }, + { + "epoch": 1.3970210580380071, + "eval_loss": 1.7706127166748047, + "eval_runtime": 37.9897, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1360 + }, + { + "epoch": 1.3980482794042115, + "grad_norm": 2.5508553981781006, + "learning_rate": 2.9958127141225736e-06, + "loss": 1.3629, + "step": 1361 + }, + { + "epoch": 1.399075500770416, + "grad_norm": 2.624800205230713, + "learning_rate": 2.9939094023601066e-06, + "loss": 1.4355, + "step": 1362 + }, + { + "epoch": 1.4001027221366205, + "grad_norm": 2.432114839553833, + "learning_rate": 2.99200609059764e-06, + "loss": 1.3995, + "step": 1363 + }, + { + "epoch": 1.4011299435028248, + "grad_norm": 2.4633607864379883, + "learning_rate": 2.9901027788351733e-06, + "loss": 1.3499, + "step": 1364 + }, + { + "epoch": 1.4021571648690292, + "grad_norm": 2.557415246963501, + "learning_rate": 2.9881994670727067e-06, + "loss": 1.3614, + "step": 1365 + }, + { + "epoch": 1.4031843862352336, + "grad_norm": 2.411295175552368, + "learning_rate": 2.98629615531024e-06, + "loss": 1.3857, + "step": 1366 + }, + { + "epoch": 1.4042116076014381, + "grad_norm": 2.429797410964966, + "learning_rate": 2.984392843547773e-06, + "loss": 1.3701, + "step": 1367 + }, + { + "epoch": 1.4052388289676425, + "grad_norm": 2.6080522537231445, + "learning_rate": 2.9824895317853063e-06, + "loss": 1.3996, + "step": 1368 + }, + { + "epoch": 1.4062660503338469, + "grad_norm": 2.5128798484802246, + "learning_rate": 2.98058622002284e-06, + "loss": 1.3973, + "step": 1369 + }, + { + "epoch": 1.4072932717000515, + "grad_norm": 2.5906126499176025, + "learning_rate": 2.9786829082603735e-06, + "loss": 1.4412, + "step": 1370 + }, + { + "epoch": 1.4072932717000515, + "eval_loss": 1.7700779438018799, + "eval_runtime": 38.017, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 1370 + }, + { + "epoch": 1.4083204930662558, + "grad_norm": 2.4890248775482178, + "learning_rate": 2.976779596497907e-06, + "loss": 1.4251, + "step": 1371 + }, + { + "epoch": 1.4093477144324602, + "grad_norm": 2.3803439140319824, + "learning_rate": 2.97487628473544e-06, + "loss": 1.4363, + "step": 1372 + }, + { + "epoch": 1.4103749357986646, + "grad_norm": 2.461958646774292, + "learning_rate": 2.9729729729729736e-06, + "loss": 1.3948, + "step": 1373 + }, + { + "epoch": 1.411402157164869, + "grad_norm": 2.419793128967285, + "learning_rate": 2.9710696612105065e-06, + "loss": 1.4367, + "step": 1374 + }, + { + "epoch": 1.4124293785310735, + "grad_norm": 2.5748846530914307, + "learning_rate": 2.96916634944804e-06, + "loss": 1.3873, + "step": 1375 + }, + { + "epoch": 1.413456599897278, + "grad_norm": 2.564558506011963, + "learning_rate": 2.9672630376855732e-06, + "loss": 1.4395, + "step": 1376 + }, + { + "epoch": 1.4144838212634823, + "grad_norm": 2.596085548400879, + "learning_rate": 2.9653597259231066e-06, + "loss": 1.4091, + "step": 1377 + }, + { + "epoch": 1.4155110426296866, + "grad_norm": 2.49029278755188, + "learning_rate": 2.96345641416064e-06, + "loss": 1.3639, + "step": 1378 + }, + { + "epoch": 1.416538263995891, + "grad_norm": 2.506887912750244, + "learning_rate": 2.961553102398173e-06, + "loss": 1.3065, + "step": 1379 + }, + { + "epoch": 1.4175654853620956, + "grad_norm": 2.517609119415283, + "learning_rate": 2.9596497906357063e-06, + "loss": 1.4365, + "step": 1380 + }, + { + "epoch": 1.4175654853620956, + "eval_loss": 1.7673696279525757, + "eval_runtime": 37.9941, + "eval_samples_per_second": 26.32, + "eval_steps_per_second": 1.105, + "step": 1380 + }, + { + "epoch": 1.4185927067283, + "grad_norm": 2.6870028972625732, + "learning_rate": 2.9577464788732396e-06, + "loss": 1.4386, + "step": 1381 + }, + { + "epoch": 1.4196199280945043, + "grad_norm": 2.3195550441741943, + "learning_rate": 2.955843167110773e-06, + "loss": 1.3771, + "step": 1382 + }, + { + "epoch": 1.420647149460709, + "grad_norm": 2.4306302070617676, + "learning_rate": 2.9539398553483064e-06, + "loss": 1.3963, + "step": 1383 + }, + { + "epoch": 1.4216743708269133, + "grad_norm": 2.4227781295776367, + "learning_rate": 2.9520365435858393e-06, + "loss": 1.394, + "step": 1384 + }, + { + "epoch": 1.4227015921931176, + "grad_norm": 2.521425485610962, + "learning_rate": 2.9501332318233727e-06, + "loss": 1.4664, + "step": 1385 + }, + { + "epoch": 1.423728813559322, + "grad_norm": 2.533363103866577, + "learning_rate": 2.948229920060906e-06, + "loss": 1.4072, + "step": 1386 + }, + { + "epoch": 1.4247560349255264, + "grad_norm": 2.378007173538208, + "learning_rate": 2.9463266082984394e-06, + "loss": 1.38, + "step": 1387 + }, + { + "epoch": 1.425783256291731, + "grad_norm": 2.476451873779297, + "learning_rate": 2.9444232965359727e-06, + "loss": 1.3864, + "step": 1388 + }, + { + "epoch": 1.4268104776579353, + "grad_norm": 2.428365707397461, + "learning_rate": 2.9425199847735057e-06, + "loss": 1.4081, + "step": 1389 + }, + { + "epoch": 1.4278376990241397, + "grad_norm": 2.4382991790771484, + "learning_rate": 2.94061667301104e-06, + "loss": 1.3806, + "step": 1390 + }, + { + "epoch": 1.4278376990241397, + "eval_loss": 1.7905043363571167, + "eval_runtime": 38.0264, + "eval_samples_per_second": 26.298, + "eval_steps_per_second": 1.104, + "step": 1390 + }, + { + "epoch": 1.428864920390344, + "grad_norm": 2.414433240890503, + "learning_rate": 2.938713361248573e-06, + "loss": 1.3113, + "step": 1391 + }, + { + "epoch": 1.4298921417565484, + "grad_norm": 2.450570821762085, + "learning_rate": 2.936810049486106e-06, + "loss": 1.4023, + "step": 1392 + }, + { + "epoch": 1.430919363122753, + "grad_norm": 2.4372148513793945, + "learning_rate": 2.9349067377236396e-06, + "loss": 1.4057, + "step": 1393 + }, + { + "epoch": 1.4319465844889574, + "grad_norm": 2.4278311729431152, + "learning_rate": 2.933003425961173e-06, + "loss": 1.3574, + "step": 1394 + }, + { + "epoch": 1.4329738058551618, + "grad_norm": 2.448517322540283, + "learning_rate": 2.9311001141987063e-06, + "loss": 1.3722, + "step": 1395 + }, + { + "epoch": 1.4340010272213661, + "grad_norm": 2.354310989379883, + "learning_rate": 2.9291968024362392e-06, + "loss": 1.3682, + "step": 1396 + }, + { + "epoch": 1.4350282485875705, + "grad_norm": 2.3318262100219727, + "learning_rate": 2.9272934906737726e-06, + "loss": 1.3604, + "step": 1397 + }, + { + "epoch": 1.436055469953775, + "grad_norm": 2.377561092376709, + "learning_rate": 2.925390178911306e-06, + "loss": 1.372, + "step": 1398 + }, + { + "epoch": 1.4370826913199795, + "grad_norm": 2.5543324947357178, + "learning_rate": 2.9234868671488393e-06, + "loss": 1.3846, + "step": 1399 + }, + { + "epoch": 1.4381099126861838, + "grad_norm": 2.6063973903656006, + "learning_rate": 2.9215835553863727e-06, + "loss": 1.3989, + "step": 1400 + }, + { + "epoch": 1.4381099126861838, + "eval_loss": 1.7730437517166138, + "eval_runtime": 38.0407, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 1400 + }, + { + "epoch": 1.4391371340523884, + "grad_norm": 2.4241244792938232, + "learning_rate": 2.9196802436239056e-06, + "loss": 1.3945, + "step": 1401 + }, + { + "epoch": 1.4401643554185928, + "grad_norm": 2.306528329849243, + "learning_rate": 2.917776931861439e-06, + "loss": 1.3686, + "step": 1402 + }, + { + "epoch": 1.4411915767847971, + "grad_norm": 2.4980850219726562, + "learning_rate": 2.9158736200989723e-06, + "loss": 1.3714, + "step": 1403 + }, + { + "epoch": 1.4422187981510015, + "grad_norm": 2.385195732116699, + "learning_rate": 2.9139703083365057e-06, + "loss": 1.3706, + "step": 1404 + }, + { + "epoch": 1.4432460195172059, + "grad_norm": 2.4087095260620117, + "learning_rate": 2.912066996574039e-06, + "loss": 1.3967, + "step": 1405 + }, + { + "epoch": 1.4442732408834105, + "grad_norm": 2.470900535583496, + "learning_rate": 2.910163684811572e-06, + "loss": 1.3351, + "step": 1406 + }, + { + "epoch": 1.4453004622496148, + "grad_norm": 2.8330743312835693, + "learning_rate": 2.9082603730491054e-06, + "loss": 1.4413, + "step": 1407 + }, + { + "epoch": 1.4463276836158192, + "grad_norm": 2.5553359985351562, + "learning_rate": 2.9063570612866387e-06, + "loss": 1.4207, + "step": 1408 + }, + { + "epoch": 1.4473549049820236, + "grad_norm": 2.509373903274536, + "learning_rate": 2.904453749524172e-06, + "loss": 1.3387, + "step": 1409 + }, + { + "epoch": 1.448382126348228, + "grad_norm": 2.5485401153564453, + "learning_rate": 2.902550437761706e-06, + "loss": 1.3765, + "step": 1410 + }, + { + "epoch": 1.448382126348228, + "eval_loss": 1.7703789472579956, + "eval_runtime": 37.7078, + "eval_samples_per_second": 26.52, + "eval_steps_per_second": 1.114, + "step": 1410 + }, + { + "epoch": 1.4494093477144325, + "grad_norm": 2.502002716064453, + "learning_rate": 2.9006471259992392e-06, + "loss": 1.3842, + "step": 1411 + }, + { + "epoch": 1.450436569080637, + "grad_norm": 2.4621520042419434, + "learning_rate": 2.8987438142367726e-06, + "loss": 1.3809, + "step": 1412 + }, + { + "epoch": 1.4514637904468413, + "grad_norm": 2.4255263805389404, + "learning_rate": 2.8968405024743055e-06, + "loss": 1.4411, + "step": 1413 + }, + { + "epoch": 1.4524910118130456, + "grad_norm": 2.2450802326202393, + "learning_rate": 2.894937190711839e-06, + "loss": 1.3389, + "step": 1414 + }, + { + "epoch": 1.45351823317925, + "grad_norm": 2.502903938293457, + "learning_rate": 2.8930338789493723e-06, + "loss": 1.4066, + "step": 1415 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 2.3685200214385986, + "learning_rate": 2.8911305671869056e-06, + "loss": 1.3995, + "step": 1416 + }, + { + "epoch": 1.455572675911659, + "grad_norm": 2.355095624923706, + "learning_rate": 2.889227255424439e-06, + "loss": 1.3352, + "step": 1417 + }, + { + "epoch": 1.4565998972778633, + "grad_norm": 2.363035202026367, + "learning_rate": 2.887323943661972e-06, + "loss": 1.3792, + "step": 1418 + }, + { + "epoch": 1.457627118644068, + "grad_norm": 2.451078176498413, + "learning_rate": 2.8854206318995053e-06, + "loss": 1.358, + "step": 1419 + }, + { + "epoch": 1.4586543400102723, + "grad_norm": 2.554501533508301, + "learning_rate": 2.8835173201370386e-06, + "loss": 1.3811, + "step": 1420 + }, + { + "epoch": 1.4586543400102723, + "eval_loss": 1.773111343383789, + "eval_runtime": 37.7693, + "eval_samples_per_second": 26.477, + "eval_steps_per_second": 1.112, + "step": 1420 + }, + { + "epoch": 1.4596815613764766, + "grad_norm": 2.556537389755249, + "learning_rate": 2.881614008374572e-06, + "loss": 1.4397, + "step": 1421 + }, + { + "epoch": 1.460708782742681, + "grad_norm": 2.5945091247558594, + "learning_rate": 2.8797106966121054e-06, + "loss": 1.3976, + "step": 1422 + }, + { + "epoch": 1.4617360041088854, + "grad_norm": 2.5158495903015137, + "learning_rate": 2.8778073848496383e-06, + "loss": 1.381, + "step": 1423 + }, + { + "epoch": 1.46276322547509, + "grad_norm": 2.359591007232666, + "learning_rate": 2.8759040730871717e-06, + "loss": 1.4156, + "step": 1424 + }, + { + "epoch": 1.4637904468412943, + "grad_norm": 2.3388748168945312, + "learning_rate": 2.874000761324705e-06, + "loss": 1.3808, + "step": 1425 + }, + { + "epoch": 1.4648176682074987, + "grad_norm": 2.438465118408203, + "learning_rate": 2.8720974495622384e-06, + "loss": 1.4158, + "step": 1426 + }, + { + "epoch": 1.465844889573703, + "grad_norm": 2.5624589920043945, + "learning_rate": 2.8701941377997718e-06, + "loss": 1.3917, + "step": 1427 + }, + { + "epoch": 1.4668721109399074, + "grad_norm": 2.467921495437622, + "learning_rate": 2.8682908260373047e-06, + "loss": 1.4316, + "step": 1428 + }, + { + "epoch": 1.467899332306112, + "grad_norm": 2.6034083366394043, + "learning_rate": 2.866387514274838e-06, + "loss": 1.4058, + "step": 1429 + }, + { + "epoch": 1.4689265536723164, + "grad_norm": 2.560844898223877, + "learning_rate": 2.864484202512372e-06, + "loss": 1.4114, + "step": 1430 + }, + { + "epoch": 1.4689265536723164, + "eval_loss": 1.7799324989318848, + "eval_runtime": 37.8814, + "eval_samples_per_second": 26.398, + "eval_steps_per_second": 1.109, + "step": 1430 + }, + { + "epoch": 1.4699537750385208, + "grad_norm": 2.469799280166626, + "learning_rate": 2.8625808907499052e-06, + "loss": 1.3767, + "step": 1431 + }, + { + "epoch": 1.4709809964047253, + "grad_norm": 2.408665895462036, + "learning_rate": 2.8606775789874386e-06, + "loss": 1.4064, + "step": 1432 + }, + { + "epoch": 1.4720082177709297, + "grad_norm": 2.497950553894043, + "learning_rate": 2.858774267224972e-06, + "loss": 1.3747, + "step": 1433 + }, + { + "epoch": 1.473035439137134, + "grad_norm": 2.6274573802948, + "learning_rate": 2.8568709554625053e-06, + "loss": 1.3994, + "step": 1434 + }, + { + "epoch": 1.4740626605033385, + "grad_norm": 2.545614719390869, + "learning_rate": 2.8549676437000382e-06, + "loss": 1.4354, + "step": 1435 + }, + { + "epoch": 1.4750898818695428, + "grad_norm": 2.796013355255127, + "learning_rate": 2.8530643319375716e-06, + "loss": 1.4081, + "step": 1436 + }, + { + "epoch": 1.4761171032357474, + "grad_norm": 2.68571400642395, + "learning_rate": 2.851161020175105e-06, + "loss": 1.3833, + "step": 1437 + }, + { + "epoch": 1.4771443246019518, + "grad_norm": 2.455244541168213, + "learning_rate": 2.8492577084126383e-06, + "loss": 1.3587, + "step": 1438 + }, + { + "epoch": 1.4781715459681561, + "grad_norm": 2.552747964859009, + "learning_rate": 2.8473543966501717e-06, + "loss": 1.4004, + "step": 1439 + }, + { + "epoch": 1.4791987673343605, + "grad_norm": 2.9428722858428955, + "learning_rate": 2.8454510848877046e-06, + "loss": 1.437, + "step": 1440 + }, + { + "epoch": 1.4791987673343605, + "eval_loss": 1.7853318452835083, + "eval_runtime": 37.8848, + "eval_samples_per_second": 26.396, + "eval_steps_per_second": 1.109, + "step": 1440 + }, + { + "epoch": 1.4802259887005649, + "grad_norm": 2.686199188232422, + "learning_rate": 2.843547773125238e-06, + "loss": 1.3901, + "step": 1441 + }, + { + "epoch": 1.4812532100667695, + "grad_norm": 2.47476863861084, + "learning_rate": 2.8416444613627714e-06, + "loss": 1.3887, + "step": 1442 + }, + { + "epoch": 1.4822804314329738, + "grad_norm": 2.425445795059204, + "learning_rate": 2.8397411496003047e-06, + "loss": 1.3502, + "step": 1443 + }, + { + "epoch": 1.4833076527991782, + "grad_norm": 2.529341459274292, + "learning_rate": 2.837837837837838e-06, + "loss": 1.3973, + "step": 1444 + }, + { + "epoch": 1.4843348741653826, + "grad_norm": 2.3861210346221924, + "learning_rate": 2.835934526075371e-06, + "loss": 1.4072, + "step": 1445 + }, + { + "epoch": 1.485362095531587, + "grad_norm": 2.359403371810913, + "learning_rate": 2.8340312143129044e-06, + "loss": 1.3553, + "step": 1446 + }, + { + "epoch": 1.4863893168977915, + "grad_norm": 2.417792558670044, + "learning_rate": 2.8321279025504377e-06, + "loss": 1.3539, + "step": 1447 + }, + { + "epoch": 1.487416538263996, + "grad_norm": 2.7177648544311523, + "learning_rate": 2.830224590787971e-06, + "loss": 1.4575, + "step": 1448 + }, + { + "epoch": 1.4884437596302003, + "grad_norm": 2.514181613922119, + "learning_rate": 2.8283212790255045e-06, + "loss": 1.388, + "step": 1449 + }, + { + "epoch": 1.4894709809964048, + "grad_norm": 2.4293761253356934, + "learning_rate": 2.8264179672630383e-06, + "loss": 1.4143, + "step": 1450 + }, + { + "epoch": 1.4894709809964048, + "eval_loss": 1.781496286392212, + "eval_runtime": 37.8942, + "eval_samples_per_second": 26.389, + "eval_steps_per_second": 1.108, + "step": 1450 + }, + { + "epoch": 1.4904982023626092, + "grad_norm": 2.495591878890991, + "learning_rate": 2.8245146555005716e-06, + "loss": 1.3809, + "step": 1451 + }, + { + "epoch": 1.4915254237288136, + "grad_norm": 2.520158290863037, + "learning_rate": 2.8226113437381046e-06, + "loss": 1.3297, + "step": 1452 + }, + { + "epoch": 1.492552645095018, + "grad_norm": 2.403061866760254, + "learning_rate": 2.820708031975638e-06, + "loss": 1.3985, + "step": 1453 + }, + { + "epoch": 1.4935798664612223, + "grad_norm": 2.606884479522705, + "learning_rate": 2.8188047202131713e-06, + "loss": 1.4179, + "step": 1454 + }, + { + "epoch": 1.494607087827427, + "grad_norm": 2.4543404579162598, + "learning_rate": 2.8169014084507046e-06, + "loss": 1.4325, + "step": 1455 + }, + { + "epoch": 1.4956343091936313, + "grad_norm": 2.3947877883911133, + "learning_rate": 2.814998096688238e-06, + "loss": 1.3876, + "step": 1456 + }, + { + "epoch": 1.4966615305598356, + "grad_norm": 2.4632408618927, + "learning_rate": 2.813094784925771e-06, + "loss": 1.3392, + "step": 1457 + }, + { + "epoch": 1.49768875192604, + "grad_norm": 2.507312059402466, + "learning_rate": 2.8111914731633043e-06, + "loss": 1.4351, + "step": 1458 + }, + { + "epoch": 1.4987159732922444, + "grad_norm": 2.6338255405426025, + "learning_rate": 2.8092881614008377e-06, + "loss": 1.389, + "step": 1459 + }, + { + "epoch": 1.499743194658449, + "grad_norm": 2.3906662464141846, + "learning_rate": 2.807384849638371e-06, + "loss": 1.3655, + "step": 1460 + }, + { + "epoch": 1.499743194658449, + "eval_loss": 1.7716007232666016, + "eval_runtime": 37.8931, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 1460 + }, + { + "epoch": 1.5007704160246533, + "grad_norm": 2.4841926097869873, + "learning_rate": 2.8054815378759044e-06, + "loss": 1.4046, + "step": 1461 + }, + { + "epoch": 1.5017976373908577, + "grad_norm": 2.4934608936309814, + "learning_rate": 2.8035782261134378e-06, + "loss": 1.3781, + "step": 1462 + }, + { + "epoch": 1.5028248587570623, + "grad_norm": 2.538504123687744, + "learning_rate": 2.8016749143509707e-06, + "loss": 1.383, + "step": 1463 + }, + { + "epoch": 1.5038520801232664, + "grad_norm": 2.461595058441162, + "learning_rate": 2.799771602588504e-06, + "loss": 1.3892, + "step": 1464 + }, + { + "epoch": 1.504879301489471, + "grad_norm": 2.3686397075653076, + "learning_rate": 2.7978682908260374e-06, + "loss": 1.4037, + "step": 1465 + }, + { + "epoch": 1.5059065228556754, + "grad_norm": 2.47755765914917, + "learning_rate": 2.795964979063571e-06, + "loss": 1.4173, + "step": 1466 + }, + { + "epoch": 1.5069337442218798, + "grad_norm": 2.504168748855591, + "learning_rate": 2.794061667301104e-06, + "loss": 1.3955, + "step": 1467 + }, + { + "epoch": 1.5079609655880843, + "grad_norm": 2.4311411380767822, + "learning_rate": 2.792158355538637e-06, + "loss": 1.4079, + "step": 1468 + }, + { + "epoch": 1.5089881869542885, + "grad_norm": 2.523200273513794, + "learning_rate": 2.7902550437761705e-06, + "loss": 1.4019, + "step": 1469 + }, + { + "epoch": 1.510015408320493, + "grad_norm": 2.3743972778320312, + "learning_rate": 2.7883517320137042e-06, + "loss": 1.3677, + "step": 1470 + }, + { + "epoch": 1.510015408320493, + "eval_loss": 1.7622252702713013, + "eval_runtime": 37.8933, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 1470 + }, + { + "epoch": 1.5110426296866974, + "grad_norm": 2.4073667526245117, + "learning_rate": 2.7864484202512376e-06, + "loss": 1.3623, + "step": 1471 + }, + { + "epoch": 1.5120698510529018, + "grad_norm": 2.5012941360473633, + "learning_rate": 2.784545108488771e-06, + "loss": 1.382, + "step": 1472 + }, + { + "epoch": 1.5130970724191064, + "grad_norm": 2.425480365753174, + "learning_rate": 2.7826417967263043e-06, + "loss": 1.3896, + "step": 1473 + }, + { + "epoch": 1.5141242937853108, + "grad_norm": 2.4600207805633545, + "learning_rate": 2.7807384849638373e-06, + "loss": 1.4364, + "step": 1474 + }, + { + "epoch": 1.5151515151515151, + "grad_norm": 2.5489892959594727, + "learning_rate": 2.7788351732013706e-06, + "loss": 1.3832, + "step": 1475 + }, + { + "epoch": 1.5161787365177197, + "grad_norm": 2.4228315353393555, + "learning_rate": 2.776931861438904e-06, + "loss": 1.3755, + "step": 1476 + }, + { + "epoch": 1.5172059578839239, + "grad_norm": 2.519340991973877, + "learning_rate": 2.7750285496764374e-06, + "loss": 1.3915, + "step": 1477 + }, + { + "epoch": 1.5182331792501285, + "grad_norm": 2.3544020652770996, + "learning_rate": 2.7731252379139707e-06, + "loss": 1.394, + "step": 1478 + }, + { + "epoch": 1.5192604006163328, + "grad_norm": 2.4037892818450928, + "learning_rate": 2.771221926151504e-06, + "loss": 1.3396, + "step": 1479 + }, + { + "epoch": 1.5202876219825372, + "grad_norm": 2.409222364425659, + "learning_rate": 2.769318614389037e-06, + "loss": 1.4052, + "step": 1480 + }, + { + "epoch": 1.5202876219825372, + "eval_loss": 1.7622238397598267, + "eval_runtime": 37.9064, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 1480 + }, + { + "epoch": 1.5213148433487418, + "grad_norm": 2.3649749755859375, + "learning_rate": 2.7674153026265704e-06, + "loss": 1.401, + "step": 1481 + }, + { + "epoch": 1.522342064714946, + "grad_norm": 2.4484262466430664, + "learning_rate": 2.7655119908641037e-06, + "loss": 1.4026, + "step": 1482 + }, + { + "epoch": 1.5233692860811505, + "grad_norm": 2.492107629776001, + "learning_rate": 2.763608679101637e-06, + "loss": 1.3985, + "step": 1483 + }, + { + "epoch": 1.5243965074473549, + "grad_norm": 2.5192503929138184, + "learning_rate": 2.7617053673391705e-06, + "loss": 1.4282, + "step": 1484 + }, + { + "epoch": 1.5254237288135593, + "grad_norm": 2.471250534057617, + "learning_rate": 2.7598020555767034e-06, + "loss": 1.4217, + "step": 1485 + }, + { + "epoch": 1.5264509501797638, + "grad_norm": 2.2985496520996094, + "learning_rate": 2.7578987438142368e-06, + "loss": 1.3599, + "step": 1486 + }, + { + "epoch": 1.5274781715459682, + "grad_norm": 2.492047071456909, + "learning_rate": 2.75599543205177e-06, + "loss": 1.4036, + "step": 1487 + }, + { + "epoch": 1.5285053929121726, + "grad_norm": 2.5131945610046387, + "learning_rate": 2.7540921202893035e-06, + "loss": 1.4206, + "step": 1488 + }, + { + "epoch": 1.529532614278377, + "grad_norm": 2.619874954223633, + "learning_rate": 2.752188808526837e-06, + "loss": 1.4055, + "step": 1489 + }, + { + "epoch": 1.5305598356445813, + "grad_norm": 2.4526877403259277, + "learning_rate": 2.75028549676437e-06, + "loss": 1.3955, + "step": 1490 + }, + { + "epoch": 1.5305598356445813, + "eval_loss": 1.758982539176941, + "eval_runtime": 37.8756, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1490 + }, + { + "epoch": 1.531587057010786, + "grad_norm": 2.4925529956817627, + "learning_rate": 2.7483821850019036e-06, + "loss": 1.3899, + "step": 1491 + }, + { + "epoch": 1.5326142783769903, + "grad_norm": 2.3789961338043213, + "learning_rate": 2.746478873239437e-06, + "loss": 1.377, + "step": 1492 + }, + { + "epoch": 1.5336414997431946, + "grad_norm": 2.4063172340393066, + "learning_rate": 2.7445755614769703e-06, + "loss": 1.384, + "step": 1493 + }, + { + "epoch": 1.5346687211093992, + "grad_norm": 2.3189985752105713, + "learning_rate": 2.7426722497145037e-06, + "loss": 1.3635, + "step": 1494 + }, + { + "epoch": 1.5356959424756034, + "grad_norm": 2.4792535305023193, + "learning_rate": 2.740768937952037e-06, + "loss": 1.3676, + "step": 1495 + }, + { + "epoch": 1.536723163841808, + "grad_norm": 2.342355489730835, + "learning_rate": 2.7388656261895704e-06, + "loss": 1.3812, + "step": 1496 + }, + { + "epoch": 1.5377503852080123, + "grad_norm": 2.556312084197998, + "learning_rate": 2.7369623144271033e-06, + "loss": 1.4529, + "step": 1497 + }, + { + "epoch": 1.5387776065742167, + "grad_norm": 2.555600643157959, + "learning_rate": 2.7350590026646367e-06, + "loss": 1.3819, + "step": 1498 + }, + { + "epoch": 1.5398048279404213, + "grad_norm": 2.4179797172546387, + "learning_rate": 2.73315569090217e-06, + "loss": 1.3383, + "step": 1499 + }, + { + "epoch": 1.5408320493066254, + "grad_norm": 2.5075631141662598, + "learning_rate": 2.7312523791397034e-06, + "loss": 1.4217, + "step": 1500 + }, + { + "epoch": 1.5408320493066254, + "eval_loss": 1.769559621810913, + "eval_runtime": 37.8701, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 1500 + }, + { + "epoch": 1.54185927067283, + "grad_norm": 2.5675160884857178, + "learning_rate": 2.7293490673772368e-06, + "loss": 1.4437, + "step": 1501 + }, + { + "epoch": 1.5428864920390344, + "grad_norm": 2.4768013954162598, + "learning_rate": 2.7274457556147697e-06, + "loss": 1.4022, + "step": 1502 + }, + { + "epoch": 1.5439137134052388, + "grad_norm": 2.4523043632507324, + "learning_rate": 2.725542443852303e-06, + "loss": 1.385, + "step": 1503 + }, + { + "epoch": 1.5449409347714433, + "grad_norm": 2.504263401031494, + "learning_rate": 2.7236391320898365e-06, + "loss": 1.398, + "step": 1504 + }, + { + "epoch": 1.5459681561376477, + "grad_norm": 2.6203269958496094, + "learning_rate": 2.72173582032737e-06, + "loss": 1.3651, + "step": 1505 + }, + { + "epoch": 1.546995377503852, + "grad_norm": 2.3805699348449707, + "learning_rate": 2.719832508564903e-06, + "loss": 1.3861, + "step": 1506 + }, + { + "epoch": 1.5480225988700564, + "grad_norm": 2.5000486373901367, + "learning_rate": 2.717929196802436e-06, + "loss": 1.3931, + "step": 1507 + }, + { + "epoch": 1.5490498202362608, + "grad_norm": 2.5567827224731445, + "learning_rate": 2.7160258850399695e-06, + "loss": 1.3608, + "step": 1508 + }, + { + "epoch": 1.5500770416024654, + "grad_norm": 2.531614303588867, + "learning_rate": 2.714122573277503e-06, + "loss": 1.3863, + "step": 1509 + }, + { + "epoch": 1.5511042629686698, + "grad_norm": 2.419722318649292, + "learning_rate": 2.712219261515036e-06, + "loss": 1.3963, + "step": 1510 + }, + { + "epoch": 1.5511042629686698, + "eval_loss": 1.7669111490249634, + "eval_runtime": 37.6622, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 1510 + }, + { + "epoch": 1.5521314843348741, + "grad_norm": 2.4597935676574707, + "learning_rate": 2.71031594975257e-06, + "loss": 1.4044, + "step": 1511 + }, + { + "epoch": 1.5531587057010787, + "grad_norm": 2.386899471282959, + "learning_rate": 2.7084126379901034e-06, + "loss": 1.4092, + "step": 1512 + }, + { + "epoch": 1.5541859270672829, + "grad_norm": 2.4398958683013916, + "learning_rate": 2.7065093262276367e-06, + "loss": 1.4185, + "step": 1513 + }, + { + "epoch": 1.5552131484334875, + "grad_norm": 2.319274663925171, + "learning_rate": 2.7046060144651697e-06, + "loss": 1.4316, + "step": 1514 + }, + { + "epoch": 1.5562403697996918, + "grad_norm": 2.4519565105438232, + "learning_rate": 2.702702702702703e-06, + "loss": 1.4387, + "step": 1515 + }, + { + "epoch": 1.5572675911658962, + "grad_norm": 2.2890849113464355, + "learning_rate": 2.7007993909402364e-06, + "loss": 1.3167, + "step": 1516 + }, + { + "epoch": 1.5582948125321008, + "grad_norm": 2.6062512397766113, + "learning_rate": 2.6988960791777697e-06, + "loss": 1.4332, + "step": 1517 + }, + { + "epoch": 1.559322033898305, + "grad_norm": 2.508957862854004, + "learning_rate": 2.696992767415303e-06, + "loss": 1.3995, + "step": 1518 + }, + { + "epoch": 1.5603492552645095, + "grad_norm": 2.4134879112243652, + "learning_rate": 2.695089455652836e-06, + "loss": 1.3793, + "step": 1519 + }, + { + "epoch": 1.5613764766307139, + "grad_norm": 2.498729705810547, + "learning_rate": 2.6931861438903694e-06, + "loss": 1.3564, + "step": 1520 + }, + { + "epoch": 1.5613764766307139, + "eval_loss": 1.76124107837677, + "eval_runtime": 37.831, + "eval_samples_per_second": 26.433, + "eval_steps_per_second": 1.11, + "step": 1520 + }, + { + "epoch": 1.5624036979969183, + "grad_norm": 2.455510139465332, + "learning_rate": 2.6912828321279028e-06, + "loss": 1.4109, + "step": 1521 + }, + { + "epoch": 1.5634309193631228, + "grad_norm": 2.422098159790039, + "learning_rate": 2.689379520365436e-06, + "loss": 1.3811, + "step": 1522 + }, + { + "epoch": 1.5644581407293272, + "grad_norm": 2.380890369415283, + "learning_rate": 2.6874762086029695e-06, + "loss": 1.4215, + "step": 1523 + }, + { + "epoch": 1.5654853620955316, + "grad_norm": 2.39363169670105, + "learning_rate": 2.6855728968405024e-06, + "loss": 1.3708, + "step": 1524 + }, + { + "epoch": 1.5665125834617362, + "grad_norm": 2.3624465465545654, + "learning_rate": 2.683669585078036e-06, + "loss": 1.4084, + "step": 1525 + }, + { + "epoch": 1.5675398048279403, + "grad_norm": 2.3589534759521484, + "learning_rate": 2.681766273315569e-06, + "loss": 1.3465, + "step": 1526 + }, + { + "epoch": 1.568567026194145, + "grad_norm": 2.4358372688293457, + "learning_rate": 2.6798629615531025e-06, + "loss": 1.3915, + "step": 1527 + }, + { + "epoch": 1.5695942475603493, + "grad_norm": 2.4162042140960693, + "learning_rate": 2.677959649790636e-06, + "loss": 1.3842, + "step": 1528 + }, + { + "epoch": 1.5706214689265536, + "grad_norm": 2.568127155303955, + "learning_rate": 2.676056338028169e-06, + "loss": 1.4197, + "step": 1529 + }, + { + "epoch": 1.5716486902927582, + "grad_norm": 2.5098628997802734, + "learning_rate": 2.674153026265702e-06, + "loss": 1.4132, + "step": 1530 + }, + { + "epoch": 1.5716486902927582, + "eval_loss": 1.7691000699996948, + "eval_runtime": 37.9709, + "eval_samples_per_second": 26.336, + "eval_steps_per_second": 1.106, + "step": 1530 + }, + { + "epoch": 1.5726759116589624, + "grad_norm": 2.5882396697998047, + "learning_rate": 2.672249714503236e-06, + "loss": 1.4223, + "step": 1531 + }, + { + "epoch": 1.573703133025167, + "grad_norm": 2.5202438831329346, + "learning_rate": 2.6703464027407693e-06, + "loss": 1.3804, + "step": 1532 + }, + { + "epoch": 1.5747303543913713, + "grad_norm": 2.4582319259643555, + "learning_rate": 2.6684430909783027e-06, + "loss": 1.3987, + "step": 1533 + }, + { + "epoch": 1.5757575757575757, + "grad_norm": 2.3516597747802734, + "learning_rate": 2.666539779215836e-06, + "loss": 1.3878, + "step": 1534 + }, + { + "epoch": 1.5767847971237803, + "grad_norm": 2.410268783569336, + "learning_rate": 2.6646364674533694e-06, + "loss": 1.3737, + "step": 1535 + }, + { + "epoch": 1.5778120184899846, + "grad_norm": 2.2420742511749268, + "learning_rate": 2.6627331556909024e-06, + "loss": 1.2908, + "step": 1536 + }, + { + "epoch": 1.578839239856189, + "grad_norm": 2.498166084289551, + "learning_rate": 2.6608298439284357e-06, + "loss": 1.383, + "step": 1537 + }, + { + "epoch": 1.5798664612223934, + "grad_norm": 2.389152765274048, + "learning_rate": 2.658926532165969e-06, + "loss": 1.3829, + "step": 1538 + }, + { + "epoch": 1.5808936825885977, + "grad_norm": 2.4197394847869873, + "learning_rate": 2.6570232204035024e-06, + "loss": 1.3438, + "step": 1539 + }, + { + "epoch": 1.5819209039548023, + "grad_norm": 2.321922540664673, + "learning_rate": 2.655119908641036e-06, + "loss": 1.3741, + "step": 1540 + }, + { + "epoch": 1.5819209039548023, + "eval_loss": 1.7643449306488037, + "eval_runtime": 38.0047, + "eval_samples_per_second": 26.313, + "eval_steps_per_second": 1.105, + "step": 1540 + }, + { + "epoch": 1.5829481253210067, + "grad_norm": 2.4056994915008545, + "learning_rate": 2.6532165968785687e-06, + "loss": 1.3857, + "step": 1541 + }, + { + "epoch": 1.583975346687211, + "grad_norm": 2.2752463817596436, + "learning_rate": 2.651313285116102e-06, + "loss": 1.3517, + "step": 1542 + }, + { + "epoch": 1.5850025680534157, + "grad_norm": 2.411092519760132, + "learning_rate": 2.6494099733536355e-06, + "loss": 1.3548, + "step": 1543 + }, + { + "epoch": 1.5860297894196198, + "grad_norm": 2.7468392848968506, + "learning_rate": 2.647506661591169e-06, + "loss": 1.5273, + "step": 1544 + }, + { + "epoch": 1.5870570107858244, + "grad_norm": 2.5081777572631836, + "learning_rate": 2.645603349828702e-06, + "loss": 1.4064, + "step": 1545 + }, + { + "epoch": 1.5880842321520288, + "grad_norm": 2.3444366455078125, + "learning_rate": 2.643700038066235e-06, + "loss": 1.3444, + "step": 1546 + }, + { + "epoch": 1.5891114535182331, + "grad_norm": 2.437304973602295, + "learning_rate": 2.6417967263037685e-06, + "loss": 1.4078, + "step": 1547 + }, + { + "epoch": 1.5901386748844377, + "grad_norm": 2.456725597381592, + "learning_rate": 2.639893414541302e-06, + "loss": 1.4212, + "step": 1548 + }, + { + "epoch": 1.5911658962506419, + "grad_norm": 2.4959378242492676, + "learning_rate": 2.6379901027788352e-06, + "loss": 1.4439, + "step": 1549 + }, + { + "epoch": 1.5921931176168465, + "grad_norm": 2.3856401443481445, + "learning_rate": 2.6360867910163686e-06, + "loss": 1.3748, + "step": 1550 + }, + { + "epoch": 1.5921931176168465, + "eval_loss": 1.7585556507110596, + "eval_runtime": 38.038, + "eval_samples_per_second": 26.289, + "eval_steps_per_second": 1.104, + "step": 1550 + }, + { + "epoch": 1.5932203389830508, + "grad_norm": 2.3285882472991943, + "learning_rate": 2.6341834792539024e-06, + "loss": 1.3912, + "step": 1551 + }, + { + "epoch": 1.5942475603492552, + "grad_norm": 2.365633249282837, + "learning_rate": 2.6322801674914357e-06, + "loss": 1.3922, + "step": 1552 + }, + { + "epoch": 1.5952747817154598, + "grad_norm": 2.460118293762207, + "learning_rate": 2.6303768557289687e-06, + "loss": 1.3899, + "step": 1553 + }, + { + "epoch": 1.5963020030816641, + "grad_norm": 2.244420051574707, + "learning_rate": 2.628473543966502e-06, + "loss": 1.347, + "step": 1554 + }, + { + "epoch": 1.5973292244478685, + "grad_norm": 2.404733180999756, + "learning_rate": 2.6265702322040354e-06, + "loss": 1.4163, + "step": 1555 + }, + { + "epoch": 1.5983564458140729, + "grad_norm": 2.3252789974212646, + "learning_rate": 2.6246669204415688e-06, + "loss": 1.3578, + "step": 1556 + }, + { + "epoch": 1.5993836671802772, + "grad_norm": 2.3664159774780273, + "learning_rate": 2.622763608679102e-06, + "loss": 1.3585, + "step": 1557 + }, + { + "epoch": 1.6004108885464818, + "grad_norm": 2.3188118934631348, + "learning_rate": 2.620860296916635e-06, + "loss": 1.3794, + "step": 1558 + }, + { + "epoch": 1.6014381099126862, + "grad_norm": 2.3163065910339355, + "learning_rate": 2.6189569851541684e-06, + "loss": 1.3157, + "step": 1559 + }, + { + "epoch": 1.6024653312788906, + "grad_norm": 2.4271090030670166, + "learning_rate": 2.617053673391702e-06, + "loss": 1.413, + "step": 1560 + }, + { + "epoch": 1.6024653312788906, + "eval_loss": 1.7588003873825073, + "eval_runtime": 38.0406, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 1560 + }, + { + "epoch": 1.6034925526450952, + "grad_norm": 2.5792505741119385, + "learning_rate": 2.615150361629235e-06, + "loss": 1.4431, + "step": 1561 + }, + { + "epoch": 1.6045197740112993, + "grad_norm": 2.4612159729003906, + "learning_rate": 2.6132470498667685e-06, + "loss": 1.3967, + "step": 1562 + }, + { + "epoch": 1.605546995377504, + "grad_norm": 2.4364020824432373, + "learning_rate": 2.6113437381043015e-06, + "loss": 1.3779, + "step": 1563 + }, + { + "epoch": 1.6065742167437083, + "grad_norm": 2.464442253112793, + "learning_rate": 2.609440426341835e-06, + "loss": 1.3685, + "step": 1564 + }, + { + "epoch": 1.6076014381099126, + "grad_norm": 2.4871442317962646, + "learning_rate": 2.607537114579368e-06, + "loss": 1.3979, + "step": 1565 + }, + { + "epoch": 1.6086286594761172, + "grad_norm": 2.375364303588867, + "learning_rate": 2.6056338028169015e-06, + "loss": 1.3836, + "step": 1566 + }, + { + "epoch": 1.6096558808423214, + "grad_norm": 2.364389657974243, + "learning_rate": 2.603730491054435e-06, + "loss": 1.3655, + "step": 1567 + }, + { + "epoch": 1.610683102208526, + "grad_norm": 2.421839714050293, + "learning_rate": 2.601827179291968e-06, + "loss": 1.398, + "step": 1568 + }, + { + "epoch": 1.6117103235747303, + "grad_norm": 2.3270041942596436, + "learning_rate": 2.599923867529501e-06, + "loss": 1.4061, + "step": 1569 + }, + { + "epoch": 1.6127375449409347, + "grad_norm": 2.4831383228302, + "learning_rate": 2.5980205557670346e-06, + "loss": 1.3808, + "step": 1570 + }, + { + "epoch": 1.6127375449409347, + "eval_loss": 1.773654818534851, + "eval_runtime": 38.046, + "eval_samples_per_second": 26.284, + "eval_steps_per_second": 1.104, + "step": 1570 + }, + { + "epoch": 1.6137647663071393, + "grad_norm": 2.499079942703247, + "learning_rate": 2.5961172440045684e-06, + "loss": 1.4167, + "step": 1571 + }, + { + "epoch": 1.6147919876733436, + "grad_norm": 2.3598642349243164, + "learning_rate": 2.5942139322421017e-06, + "loss": 1.4319, + "step": 1572 + }, + { + "epoch": 1.615819209039548, + "grad_norm": 2.3554258346557617, + "learning_rate": 2.592310620479635e-06, + "loss": 1.3723, + "step": 1573 + }, + { + "epoch": 1.6168464304057526, + "grad_norm": 2.2909231185913086, + "learning_rate": 2.5904073087171684e-06, + "loss": 1.339, + "step": 1574 + }, + { + "epoch": 1.6178736517719567, + "grad_norm": 2.3029019832611084, + "learning_rate": 2.5885039969547014e-06, + "loss": 1.4026, + "step": 1575 + }, + { + "epoch": 1.6189008731381613, + "grad_norm": 2.2370176315307617, + "learning_rate": 2.5866006851922347e-06, + "loss": 1.3531, + "step": 1576 + }, + { + "epoch": 1.6199280945043657, + "grad_norm": 2.2932159900665283, + "learning_rate": 2.584697373429768e-06, + "loss": 1.3585, + "step": 1577 + }, + { + "epoch": 1.62095531587057, + "grad_norm": 2.391493797302246, + "learning_rate": 2.5827940616673015e-06, + "loss": 1.3817, + "step": 1578 + }, + { + "epoch": 1.6219825372367747, + "grad_norm": 2.476572036743164, + "learning_rate": 2.580890749904835e-06, + "loss": 1.4241, + "step": 1579 + }, + { + "epoch": 1.6230097586029788, + "grad_norm": 2.3396525382995605, + "learning_rate": 2.5789874381423678e-06, + "loss": 1.3734, + "step": 1580 + }, + { + "epoch": 1.6230097586029788, + "eval_loss": 1.7641068696975708, + "eval_runtime": 38.0413, + "eval_samples_per_second": 26.287, + "eval_steps_per_second": 1.104, + "step": 1580 + }, + { + "epoch": 1.6240369799691834, + "grad_norm": 2.381436824798584, + "learning_rate": 2.577084126379901e-06, + "loss": 1.3764, + "step": 1581 + }, + { + "epoch": 1.6250642013353878, + "grad_norm": 2.391737937927246, + "learning_rate": 2.5751808146174345e-06, + "loss": 1.3419, + "step": 1582 + }, + { + "epoch": 1.6260914227015921, + "grad_norm": 2.453653573989868, + "learning_rate": 2.573277502854968e-06, + "loss": 1.398, + "step": 1583 + }, + { + "epoch": 1.6271186440677967, + "grad_norm": 2.3906450271606445, + "learning_rate": 2.5713741910925012e-06, + "loss": 1.3605, + "step": 1584 + }, + { + "epoch": 1.6281458654340009, + "grad_norm": 2.347041606903076, + "learning_rate": 2.569470879330034e-06, + "loss": 1.369, + "step": 1585 + }, + { + "epoch": 1.6291730868002054, + "grad_norm": 2.571047306060791, + "learning_rate": 2.5675675675675675e-06, + "loss": 1.421, + "step": 1586 + }, + { + "epoch": 1.6302003081664098, + "grad_norm": 2.3640646934509277, + "learning_rate": 2.565664255805101e-06, + "loss": 1.3736, + "step": 1587 + }, + { + "epoch": 1.6312275295326142, + "grad_norm": 2.419844388961792, + "learning_rate": 2.5637609440426343e-06, + "loss": 1.4008, + "step": 1588 + }, + { + "epoch": 1.6322547508988188, + "grad_norm": 2.3414361476898193, + "learning_rate": 2.5618576322801676e-06, + "loss": 1.3473, + "step": 1589 + }, + { + "epoch": 1.6332819722650231, + "grad_norm": 2.462259292602539, + "learning_rate": 2.5599543205177006e-06, + "loss": 1.4418, + "step": 1590 + }, + { + "epoch": 1.6332819722650231, + "eval_loss": 1.760501503944397, + "eval_runtime": 38.0428, + "eval_samples_per_second": 26.286, + "eval_steps_per_second": 1.104, + "step": 1590 + }, + { + "epoch": 1.6343091936312275, + "grad_norm": 2.49070405960083, + "learning_rate": 2.558051008755234e-06, + "loss": 1.3859, + "step": 1591 + }, + { + "epoch": 1.635336414997432, + "grad_norm": 2.500180244445801, + "learning_rate": 2.5561476969927677e-06, + "loss": 1.4379, + "step": 1592 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 2.4660637378692627, + "learning_rate": 2.554244385230301e-06, + "loss": 1.3656, + "step": 1593 + }, + { + "epoch": 1.6373908577298408, + "grad_norm": 2.5678634643554688, + "learning_rate": 2.5523410734678344e-06, + "loss": 1.3939, + "step": 1594 + }, + { + "epoch": 1.6384180790960452, + "grad_norm": 2.358930826187134, + "learning_rate": 2.550437761705368e-06, + "loss": 1.3281, + "step": 1595 + }, + { + "epoch": 1.6394453004622496, + "grad_norm": 2.462233543395996, + "learning_rate": 2.548534449942901e-06, + "loss": 1.3643, + "step": 1596 + }, + { + "epoch": 1.6404725218284542, + "grad_norm": 2.5294501781463623, + "learning_rate": 2.546631138180434e-06, + "loss": 1.4009, + "step": 1597 + }, + { + "epoch": 1.6414997431946583, + "grad_norm": 2.5325710773468018, + "learning_rate": 2.5447278264179675e-06, + "loss": 1.3797, + "step": 1598 + }, + { + "epoch": 1.642526964560863, + "grad_norm": 2.3293585777282715, + "learning_rate": 2.542824514655501e-06, + "loss": 1.3492, + "step": 1599 + }, + { + "epoch": 1.6435541859270673, + "grad_norm": 2.5357868671417236, + "learning_rate": 2.540921202893034e-06, + "loss": 1.4295, + "step": 1600 + }, + { + "epoch": 1.6435541859270673, + "eval_loss": 1.7499933242797852, + "eval_runtime": 38.0219, + "eval_samples_per_second": 26.301, + "eval_steps_per_second": 1.105, + "step": 1600 + }, + { + "epoch": 1.6445814072932716, + "grad_norm": 2.509333848953247, + "learning_rate": 2.5390178911305675e-06, + "loss": 1.4174, + "step": 1601 + }, + { + "epoch": 1.6456086286594762, + "grad_norm": 2.4453277587890625, + "learning_rate": 2.5371145793681005e-06, + "loss": 1.3898, + "step": 1602 + }, + { + "epoch": 1.6466358500256806, + "grad_norm": 2.506847381591797, + "learning_rate": 2.535211267605634e-06, + "loss": 1.3881, + "step": 1603 + }, + { + "epoch": 1.647663071391885, + "grad_norm": 2.2735178470611572, + "learning_rate": 2.533307955843167e-06, + "loss": 1.3268, + "step": 1604 + }, + { + "epoch": 1.6486902927580893, + "grad_norm": 2.479081630706787, + "learning_rate": 2.5314046440807006e-06, + "loss": 1.3897, + "step": 1605 + }, + { + "epoch": 1.6497175141242937, + "grad_norm": 2.3882839679718018, + "learning_rate": 2.529501332318234e-06, + "loss": 1.3734, + "step": 1606 + }, + { + "epoch": 1.6507447354904983, + "grad_norm": 2.521341323852539, + "learning_rate": 2.5275980205557673e-06, + "loss": 1.4298, + "step": 1607 + }, + { + "epoch": 1.6517719568567026, + "grad_norm": 2.55405855178833, + "learning_rate": 2.5256947087933002e-06, + "loss": 1.4497, + "step": 1608 + }, + { + "epoch": 1.652799178222907, + "grad_norm": 2.534478187561035, + "learning_rate": 2.5237913970308336e-06, + "loss": 1.438, + "step": 1609 + }, + { + "epoch": 1.6538263995891116, + "grad_norm": 2.453374147415161, + "learning_rate": 2.521888085268367e-06, + "loss": 1.3834, + "step": 1610 + }, + { + "epoch": 1.6538263995891116, + "eval_loss": 1.7550541162490845, + "eval_runtime": 37.6671, + "eval_samples_per_second": 26.548, + "eval_steps_per_second": 1.115, + "step": 1610 + }, + { + "epoch": 1.6548536209553157, + "grad_norm": 2.4754152297973633, + "learning_rate": 2.5199847735059003e-06, + "loss": 1.434, + "step": 1611 + }, + { + "epoch": 1.6558808423215203, + "grad_norm": 2.3785831928253174, + "learning_rate": 2.518081461743434e-06, + "loss": 1.3706, + "step": 1612 + }, + { + "epoch": 1.6569080636877247, + "grad_norm": 2.3936359882354736, + "learning_rate": 2.5161781499809675e-06, + "loss": 1.3955, + "step": 1613 + }, + { + "epoch": 1.657935285053929, + "grad_norm": 2.338012456893921, + "learning_rate": 2.5142748382185004e-06, + "loss": 1.3532, + "step": 1614 + }, + { + "epoch": 1.6589625064201337, + "grad_norm": 2.5027408599853516, + "learning_rate": 2.5123715264560338e-06, + "loss": 1.3522, + "step": 1615 + }, + { + "epoch": 1.6599897277863378, + "grad_norm": 2.4059648513793945, + "learning_rate": 2.510468214693567e-06, + "loss": 1.3767, + "step": 1616 + }, + { + "epoch": 1.6610169491525424, + "grad_norm": 2.324235200881958, + "learning_rate": 2.5085649029311005e-06, + "loss": 1.3248, + "step": 1617 + }, + { + "epoch": 1.6620441705187468, + "grad_norm": 2.336317300796509, + "learning_rate": 2.506661591168634e-06, + "loss": 1.288, + "step": 1618 + }, + { + "epoch": 1.6630713918849511, + "grad_norm": 2.3386120796203613, + "learning_rate": 2.504758279406167e-06, + "loss": 1.3715, + "step": 1619 + }, + { + "epoch": 1.6640986132511557, + "grad_norm": 2.5584142208099365, + "learning_rate": 2.5028549676437e-06, + "loss": 1.4157, + "step": 1620 + }, + { + "epoch": 1.6640986132511557, + "eval_loss": 1.7575956583023071, + "eval_runtime": 37.7897, + "eval_samples_per_second": 26.462, + "eval_steps_per_second": 1.111, + "step": 1620 + }, + { + "epoch": 1.66512583461736, + "grad_norm": 2.475881338119507, + "learning_rate": 2.5009516558812335e-06, + "loss": 1.357, + "step": 1621 + }, + { + "epoch": 1.6661530559835644, + "grad_norm": 2.4611690044403076, + "learning_rate": 2.499048344118767e-06, + "loss": 1.4077, + "step": 1622 + }, + { + "epoch": 1.667180277349769, + "grad_norm": 2.4230902194976807, + "learning_rate": 2.4971450323563002e-06, + "loss": 1.443, + "step": 1623 + }, + { + "epoch": 1.6682074987159732, + "grad_norm": 2.359095335006714, + "learning_rate": 2.4952417205938336e-06, + "loss": 1.3595, + "step": 1624 + }, + { + "epoch": 1.6692347200821778, + "grad_norm": 2.4128329753875732, + "learning_rate": 2.4933384088313666e-06, + "loss": 1.3636, + "step": 1625 + }, + { + "epoch": 1.6702619414483821, + "grad_norm": 2.3391597270965576, + "learning_rate": 2.4914350970689e-06, + "loss": 1.3536, + "step": 1626 + }, + { + "epoch": 1.6712891628145865, + "grad_norm": 2.478581190109253, + "learning_rate": 2.4895317853064337e-06, + "loss": 1.3642, + "step": 1627 + }, + { + "epoch": 1.672316384180791, + "grad_norm": 2.7438251972198486, + "learning_rate": 2.4876284735439666e-06, + "loss": 1.4192, + "step": 1628 + }, + { + "epoch": 1.6733436055469952, + "grad_norm": 2.508089542388916, + "learning_rate": 2.4857251617815e-06, + "loss": 1.3879, + "step": 1629 + }, + { + "epoch": 1.6743708269131998, + "grad_norm": 2.5251338481903076, + "learning_rate": 2.4838218500190334e-06, + "loss": 1.3936, + "step": 1630 + }, + { + "epoch": 1.6743708269131998, + "eval_loss": 1.7578516006469727, + "eval_runtime": 37.8753, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1630 + }, + { + "epoch": 1.6753980482794042, + "grad_norm": 2.3655474185943604, + "learning_rate": 2.4819185382565667e-06, + "loss": 1.3839, + "step": 1631 + }, + { + "epoch": 1.6764252696456086, + "grad_norm": 2.4703869819641113, + "learning_rate": 2.4800152264941e-06, + "loss": 1.3767, + "step": 1632 + }, + { + "epoch": 1.6774524910118132, + "grad_norm": 2.4739503860473633, + "learning_rate": 2.478111914731633e-06, + "loss": 1.3465, + "step": 1633 + }, + { + "epoch": 1.6784797123780173, + "grad_norm": 2.356006383895874, + "learning_rate": 2.4762086029691664e-06, + "loss": 1.3828, + "step": 1634 + }, + { + "epoch": 1.6795069337442219, + "grad_norm": 2.7019715309143066, + "learning_rate": 2.4743052912066998e-06, + "loss": 1.4875, + "step": 1635 + }, + { + "epoch": 1.6805341551104263, + "grad_norm": 2.3504486083984375, + "learning_rate": 2.472401979444233e-06, + "loss": 1.3494, + "step": 1636 + }, + { + "epoch": 1.6815613764766306, + "grad_norm": 2.6056807041168213, + "learning_rate": 2.4704986676817665e-06, + "loss": 1.4399, + "step": 1637 + }, + { + "epoch": 1.6825885978428352, + "grad_norm": 2.481422185897827, + "learning_rate": 2.4685953559193e-06, + "loss": 1.3803, + "step": 1638 + }, + { + "epoch": 1.6836158192090396, + "grad_norm": 2.483227491378784, + "learning_rate": 2.466692044156833e-06, + "loss": 1.4255, + "step": 1639 + }, + { + "epoch": 1.684643040575244, + "grad_norm": 2.4171221256256104, + "learning_rate": 2.4647887323943666e-06, + "loss": 1.3541, + "step": 1640 + }, + { + "epoch": 1.684643040575244, + "eval_loss": 1.7658556699752808, + "eval_runtime": 37.8673, + "eval_samples_per_second": 26.408, + "eval_steps_per_second": 1.109, + "step": 1640 + }, + { + "epoch": 1.6856702619414485, + "grad_norm": 2.2551865577697754, + "learning_rate": 2.4628854206319e-06, + "loss": 1.334, + "step": 1641 + }, + { + "epoch": 1.6866974833076527, + "grad_norm": 2.45120906829834, + "learning_rate": 2.460982108869433e-06, + "loss": 1.3416, + "step": 1642 + }, + { + "epoch": 1.6877247046738573, + "grad_norm": 2.4356141090393066, + "learning_rate": 2.4590787971069662e-06, + "loss": 1.4222, + "step": 1643 + }, + { + "epoch": 1.6887519260400616, + "grad_norm": 2.3313231468200684, + "learning_rate": 2.4571754853444996e-06, + "loss": 1.3654, + "step": 1644 + }, + { + "epoch": 1.689779147406266, + "grad_norm": 2.3518543243408203, + "learning_rate": 2.455272173582033e-06, + "loss": 1.3521, + "step": 1645 + }, + { + "epoch": 1.6908063687724706, + "grad_norm": 2.477158546447754, + "learning_rate": 2.4533688618195663e-06, + "loss": 1.3665, + "step": 1646 + }, + { + "epoch": 1.6918335901386747, + "grad_norm": 2.2717649936676025, + "learning_rate": 2.4514655500570997e-06, + "loss": 1.2992, + "step": 1647 + }, + { + "epoch": 1.6928608115048793, + "grad_norm": 2.2843265533447266, + "learning_rate": 2.449562238294633e-06, + "loss": 1.3458, + "step": 1648 + }, + { + "epoch": 1.6938880328710837, + "grad_norm": 2.4243931770324707, + "learning_rate": 2.4476589265321664e-06, + "loss": 1.3143, + "step": 1649 + }, + { + "epoch": 1.694915254237288, + "grad_norm": 2.4439947605133057, + "learning_rate": 2.4457556147696993e-06, + "loss": 1.41, + "step": 1650 + }, + { + "epoch": 1.694915254237288, + "eval_loss": 1.762351155281067, + "eval_runtime": 37.8973, + "eval_samples_per_second": 26.387, + "eval_steps_per_second": 1.108, + "step": 1650 + }, + { + "epoch": 1.6959424756034926, + "grad_norm": 2.3726727962493896, + "learning_rate": 2.4438523030072327e-06, + "loss": 1.388, + "step": 1651 + }, + { + "epoch": 1.696969696969697, + "grad_norm": 2.4449899196624756, + "learning_rate": 2.441948991244766e-06, + "loss": 1.4023, + "step": 1652 + }, + { + "epoch": 1.6979969183359014, + "grad_norm": 2.4462502002716064, + "learning_rate": 2.4400456794822994e-06, + "loss": 1.3761, + "step": 1653 + }, + { + "epoch": 1.6990241397021058, + "grad_norm": 2.413308620452881, + "learning_rate": 2.438142367719833e-06, + "loss": 1.4017, + "step": 1654 + }, + { + "epoch": 1.7000513610683101, + "grad_norm": 2.398103952407837, + "learning_rate": 2.4362390559573657e-06, + "loss": 1.3709, + "step": 1655 + }, + { + "epoch": 1.7010785824345147, + "grad_norm": 2.4637506008148193, + "learning_rate": 2.434335744194899e-06, + "loss": 1.3657, + "step": 1656 + }, + { + "epoch": 1.702105803800719, + "grad_norm": 2.419947385787964, + "learning_rate": 2.432432432432433e-06, + "loss": 1.4071, + "step": 1657 + }, + { + "epoch": 1.7031330251669234, + "grad_norm": 2.4477648735046387, + "learning_rate": 2.4305291206699662e-06, + "loss": 1.3288, + "step": 1658 + }, + { + "epoch": 1.704160246533128, + "grad_norm": 2.3359293937683105, + "learning_rate": 2.428625808907499e-06, + "loss": 1.396, + "step": 1659 + }, + { + "epoch": 1.7051874678993322, + "grad_norm": 2.396193742752075, + "learning_rate": 2.4267224971450325e-06, + "loss": 1.4246, + "step": 1660 + }, + { + "epoch": 1.7051874678993322, + "eval_loss": 1.7526435852050781, + "eval_runtime": 37.9178, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 1660 + }, + { + "epoch": 1.7062146892655368, + "grad_norm": 2.511453628540039, + "learning_rate": 2.424819185382566e-06, + "loss": 1.3877, + "step": 1661 + }, + { + "epoch": 1.7072419106317411, + "grad_norm": 2.4472320079803467, + "learning_rate": 2.4229158736200993e-06, + "loss": 1.3346, + "step": 1662 + }, + { + "epoch": 1.7082691319979455, + "grad_norm": 2.2219882011413574, + "learning_rate": 2.4210125618576326e-06, + "loss": 1.3352, + "step": 1663 + }, + { + "epoch": 1.70929635336415, + "grad_norm": 2.440121650695801, + "learning_rate": 2.4191092500951656e-06, + "loss": 1.3644, + "step": 1664 + }, + { + "epoch": 1.7103235747303542, + "grad_norm": 2.5501232147216797, + "learning_rate": 2.417205938332699e-06, + "loss": 1.4073, + "step": 1665 + }, + { + "epoch": 1.7113507960965588, + "grad_norm": 2.4315030574798584, + "learning_rate": 2.4153026265702323e-06, + "loss": 1.3137, + "step": 1666 + }, + { + "epoch": 1.7123780174627632, + "grad_norm": 2.3719727993011475, + "learning_rate": 2.4133993148077657e-06, + "loss": 1.3484, + "step": 1667 + }, + { + "epoch": 1.7134052388289676, + "grad_norm": 2.3800482749938965, + "learning_rate": 2.411496003045299e-06, + "loss": 1.4106, + "step": 1668 + }, + { + "epoch": 1.7144324601951721, + "grad_norm": 2.4554190635681152, + "learning_rate": 2.4095926912828324e-06, + "loss": 1.3603, + "step": 1669 + }, + { + "epoch": 1.7154596815613765, + "grad_norm": 2.2984702587127686, + "learning_rate": 2.4076893795203658e-06, + "loss": 1.3739, + "step": 1670 + }, + { + "epoch": 1.7154596815613765, + "eval_loss": 1.752260446548462, + "eval_runtime": 37.9208, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 1670 + }, + { + "epoch": 1.7164869029275809, + "grad_norm": 2.3967223167419434, + "learning_rate": 2.405786067757899e-06, + "loss": 1.3598, + "step": 1671 + }, + { + "epoch": 1.7175141242937855, + "grad_norm": 2.585935592651367, + "learning_rate": 2.403882755995432e-06, + "loss": 1.4489, + "step": 1672 + }, + { + "epoch": 1.7185413456599896, + "grad_norm": 2.4227662086486816, + "learning_rate": 2.4019794442329654e-06, + "loss": 1.3745, + "step": 1673 + }, + { + "epoch": 1.7195685670261942, + "grad_norm": 2.485344648361206, + "learning_rate": 2.4000761324704988e-06, + "loss": 1.3405, + "step": 1674 + }, + { + "epoch": 1.7205957883923986, + "grad_norm": 2.4055545330047607, + "learning_rate": 2.398172820708032e-06, + "loss": 1.3914, + "step": 1675 + }, + { + "epoch": 1.721623009758603, + "grad_norm": 2.3449783325195312, + "learning_rate": 2.3962695089455655e-06, + "loss": 1.3637, + "step": 1676 + }, + { + "epoch": 1.7226502311248075, + "grad_norm": 2.338590621948242, + "learning_rate": 2.3943661971830984e-06, + "loss": 1.3571, + "step": 1677 + }, + { + "epoch": 1.7236774524910117, + "grad_norm": 2.3274776935577393, + "learning_rate": 2.3924628854206322e-06, + "loss": 1.3632, + "step": 1678 + }, + { + "epoch": 1.7247046738572163, + "grad_norm": 2.3467724323272705, + "learning_rate": 2.3905595736581656e-06, + "loss": 1.3568, + "step": 1679 + }, + { + "epoch": 1.7257318952234206, + "grad_norm": 2.552039861679077, + "learning_rate": 2.388656261895699e-06, + "loss": 1.4099, + "step": 1680 + }, + { + "epoch": 1.7257318952234206, + "eval_loss": 1.745519995689392, + "eval_runtime": 37.9527, + "eval_samples_per_second": 26.349, + "eval_steps_per_second": 1.107, + "step": 1680 + }, + { + "epoch": 1.726759116589625, + "grad_norm": 2.5332300662994385, + "learning_rate": 2.386752950133232e-06, + "loss": 1.4017, + "step": 1681 + }, + { + "epoch": 1.7277863379558296, + "grad_norm": 2.4033122062683105, + "learning_rate": 2.3848496383707653e-06, + "loss": 1.3866, + "step": 1682 + }, + { + "epoch": 1.7288135593220337, + "grad_norm": 2.460071086883545, + "learning_rate": 2.3829463266082986e-06, + "loss": 1.3824, + "step": 1683 + }, + { + "epoch": 1.7298407806882383, + "grad_norm": 2.5089826583862305, + "learning_rate": 2.381043014845832e-06, + "loss": 1.398, + "step": 1684 + }, + { + "epoch": 1.7308680020544427, + "grad_norm": 2.419689893722534, + "learning_rate": 2.3791397030833653e-06, + "loss": 1.3852, + "step": 1685 + }, + { + "epoch": 1.731895223420647, + "grad_norm": 2.314080238342285, + "learning_rate": 2.3772363913208983e-06, + "loss": 1.3666, + "step": 1686 + }, + { + "epoch": 1.7329224447868516, + "grad_norm": 2.324474334716797, + "learning_rate": 2.3753330795584316e-06, + "loss": 1.3495, + "step": 1687 + }, + { + "epoch": 1.733949666153056, + "grad_norm": 2.5079402923583984, + "learning_rate": 2.3734297677959654e-06, + "loss": 1.3731, + "step": 1688 + }, + { + "epoch": 1.7349768875192604, + "grad_norm": 2.492154598236084, + "learning_rate": 2.3715264560334984e-06, + "loss": 1.413, + "step": 1689 + }, + { + "epoch": 1.736004108885465, + "grad_norm": 2.3400940895080566, + "learning_rate": 2.3696231442710317e-06, + "loss": 1.3952, + "step": 1690 + }, + { + "epoch": 1.736004108885465, + "eval_loss": 1.7412214279174805, + "eval_runtime": 37.9159, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 1690 + }, + { + "epoch": 1.7370313302516691, + "grad_norm": 2.3744688034057617, + "learning_rate": 2.367719832508565e-06, + "loss": 1.396, + "step": 1691 + }, + { + "epoch": 1.7380585516178737, + "grad_norm": 2.4762585163116455, + "learning_rate": 2.3658165207460985e-06, + "loss": 1.4113, + "step": 1692 + }, + { + "epoch": 1.739085772984078, + "grad_norm": 2.361391305923462, + "learning_rate": 2.363913208983632e-06, + "loss": 1.3962, + "step": 1693 + }, + { + "epoch": 1.7401129943502824, + "grad_norm": 2.4838762283325195, + "learning_rate": 2.362009897221165e-06, + "loss": 1.4098, + "step": 1694 + }, + { + "epoch": 1.741140215716487, + "grad_norm": 2.4860999584198, + "learning_rate": 2.360106585458698e-06, + "loss": 1.4006, + "step": 1695 + }, + { + "epoch": 1.7421674370826912, + "grad_norm": 2.3834593296051025, + "learning_rate": 2.3582032736962315e-06, + "loss": 1.3445, + "step": 1696 + }, + { + "epoch": 1.7431946584488958, + "grad_norm": 2.5383975505828857, + "learning_rate": 2.356299961933765e-06, + "loss": 1.3849, + "step": 1697 + }, + { + "epoch": 1.7442218798151001, + "grad_norm": 2.414038896560669, + "learning_rate": 2.354396650171298e-06, + "loss": 1.356, + "step": 1698 + }, + { + "epoch": 1.7452491011813045, + "grad_norm": 2.3932273387908936, + "learning_rate": 2.3524933384088316e-06, + "loss": 1.3422, + "step": 1699 + }, + { + "epoch": 1.746276322547509, + "grad_norm": 2.5618491172790527, + "learning_rate": 2.350590026646365e-06, + "loss": 1.4324, + "step": 1700 + }, + { + "epoch": 1.746276322547509, + "eval_loss": 1.7560412883758545, + "eval_runtime": 37.9455, + "eval_samples_per_second": 26.354, + "eval_steps_per_second": 1.107, + "step": 1700 + }, + { + "epoch": 1.7473035439137135, + "grad_norm": 2.450718879699707, + "learning_rate": 2.3486867148838983e-06, + "loss": 1.3631, + "step": 1701 + }, + { + "epoch": 1.7483307652799178, + "grad_norm": 2.4136962890625, + "learning_rate": 2.3467834031214317e-06, + "loss": 1.3528, + "step": 1702 + }, + { + "epoch": 1.7493579866461222, + "grad_norm": 2.4310078620910645, + "learning_rate": 2.3448800913589646e-06, + "loss": 1.369, + "step": 1703 + }, + { + "epoch": 1.7503852080123266, + "grad_norm": 2.5222883224487305, + "learning_rate": 2.342976779596498e-06, + "loss": 1.3897, + "step": 1704 + }, + { + "epoch": 1.7514124293785311, + "grad_norm": 2.3903377056121826, + "learning_rate": 2.3410734678340313e-06, + "loss": 1.3526, + "step": 1705 + }, + { + "epoch": 1.7524396507447355, + "grad_norm": 2.3869972229003906, + "learning_rate": 2.3391701560715647e-06, + "loss": 1.4091, + "step": 1706 + }, + { + "epoch": 1.7534668721109399, + "grad_norm": 2.3368723392486572, + "learning_rate": 2.337266844309098e-06, + "loss": 1.3671, + "step": 1707 + }, + { + "epoch": 1.7544940934771445, + "grad_norm": 2.4195168018341064, + "learning_rate": 2.3353635325466314e-06, + "loss": 1.3701, + "step": 1708 + }, + { + "epoch": 1.7555213148433486, + "grad_norm": 2.4312965869903564, + "learning_rate": 2.3334602207841648e-06, + "loss": 1.3795, + "step": 1709 + }, + { + "epoch": 1.7565485362095532, + "grad_norm": 2.582592725753784, + "learning_rate": 2.331556909021698e-06, + "loss": 1.3743, + "step": 1710 + }, + { + "epoch": 1.7565485362095532, + "eval_loss": 1.7604093551635742, + "eval_runtime": 37.6917, + "eval_samples_per_second": 26.531, + "eval_steps_per_second": 1.114, + "step": 1710 + }, + { + "epoch": 1.7575757575757576, + "grad_norm": 2.4194061756134033, + "learning_rate": 2.3296535972592315e-06, + "loss": 1.4049, + "step": 1711 + }, + { + "epoch": 1.758602978941962, + "grad_norm": 2.4093074798583984, + "learning_rate": 2.3277502854967644e-06, + "loss": 1.352, + "step": 1712 + }, + { + "epoch": 1.7596302003081665, + "grad_norm": 2.509335517883301, + "learning_rate": 2.325846973734298e-06, + "loss": 1.4221, + "step": 1713 + }, + { + "epoch": 1.7606574216743707, + "grad_norm": 2.422558069229126, + "learning_rate": 2.323943661971831e-06, + "loss": 1.3508, + "step": 1714 + }, + { + "epoch": 1.7616846430405753, + "grad_norm": 2.3949813842773438, + "learning_rate": 2.3220403502093645e-06, + "loss": 1.4146, + "step": 1715 + }, + { + "epoch": 1.7627118644067796, + "grad_norm": 2.47383189201355, + "learning_rate": 2.320137038446898e-06, + "loss": 1.413, + "step": 1716 + }, + { + "epoch": 1.763739085772984, + "grad_norm": 2.403531312942505, + "learning_rate": 2.318233726684431e-06, + "loss": 1.4181, + "step": 1717 + }, + { + "epoch": 1.7647663071391886, + "grad_norm": 2.463175058364868, + "learning_rate": 2.3163304149219646e-06, + "loss": 1.4121, + "step": 1718 + }, + { + "epoch": 1.765793528505393, + "grad_norm": 2.4082279205322266, + "learning_rate": 2.314427103159498e-06, + "loss": 1.3844, + "step": 1719 + }, + { + "epoch": 1.7668207498715973, + "grad_norm": 2.2674102783203125, + "learning_rate": 2.312523791397031e-06, + "loss": 1.3647, + "step": 1720 + }, + { + "epoch": 1.7668207498715973, + "eval_loss": 1.7476056814193726, + "eval_runtime": 37.8818, + "eval_samples_per_second": 26.398, + "eval_steps_per_second": 1.109, + "step": 1720 + }, + { + "epoch": 1.767847971237802, + "grad_norm": 2.2334558963775635, + "learning_rate": 2.3106204796345643e-06, + "loss": 1.3425, + "step": 1721 + }, + { + "epoch": 1.768875192604006, + "grad_norm": 2.3458974361419678, + "learning_rate": 2.3087171678720976e-06, + "loss": 1.3707, + "step": 1722 + }, + { + "epoch": 1.7699024139702106, + "grad_norm": 2.437530040740967, + "learning_rate": 2.306813856109631e-06, + "loss": 1.422, + "step": 1723 + }, + { + "epoch": 1.770929635336415, + "grad_norm": 2.467456817626953, + "learning_rate": 2.3049105443471644e-06, + "loss": 1.3795, + "step": 1724 + }, + { + "epoch": 1.7719568567026194, + "grad_norm": 2.485546827316284, + "learning_rate": 2.3030072325846973e-06, + "loss": 1.3648, + "step": 1725 + }, + { + "epoch": 1.772984078068824, + "grad_norm": 2.4228715896606445, + "learning_rate": 2.3011039208222307e-06, + "loss": 1.3508, + "step": 1726 + }, + { + "epoch": 1.774011299435028, + "grad_norm": 2.3490238189697266, + "learning_rate": 2.299200609059764e-06, + "loss": 1.3587, + "step": 1727 + }, + { + "epoch": 1.7750385208012327, + "grad_norm": 2.333501100540161, + "learning_rate": 2.297297297297298e-06, + "loss": 1.3593, + "step": 1728 + }, + { + "epoch": 1.776065742167437, + "grad_norm": 2.459859609603882, + "learning_rate": 2.2953939855348308e-06, + "loss": 1.3333, + "step": 1729 + }, + { + "epoch": 1.7770929635336414, + "grad_norm": 2.476691961288452, + "learning_rate": 2.293490673772364e-06, + "loss": 1.4137, + "step": 1730 + }, + { + "epoch": 1.7770929635336414, + "eval_loss": 1.7561018466949463, + "eval_runtime": 37.9778, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1730 + }, + { + "epoch": 1.778120184899846, + "grad_norm": 2.443805694580078, + "learning_rate": 2.2915873620098975e-06, + "loss": 1.3981, + "step": 1731 + }, + { + "epoch": 1.7791474062660502, + "grad_norm": 2.4164700508117676, + "learning_rate": 2.289684050247431e-06, + "loss": 1.392, + "step": 1732 + }, + { + "epoch": 1.7801746276322548, + "grad_norm": 2.3005855083465576, + "learning_rate": 2.287780738484964e-06, + "loss": 1.3603, + "step": 1733 + }, + { + "epoch": 1.7812018489984591, + "grad_norm": 2.395570755004883, + "learning_rate": 2.285877426722497e-06, + "loss": 1.383, + "step": 1734 + }, + { + "epoch": 1.7822290703646635, + "grad_norm": 2.4991118907928467, + "learning_rate": 2.2839741149600305e-06, + "loss": 1.4121, + "step": 1735 + }, + { + "epoch": 1.783256291730868, + "grad_norm": 2.420830726623535, + "learning_rate": 2.282070803197564e-06, + "loss": 1.3997, + "step": 1736 + }, + { + "epoch": 1.7842835130970724, + "grad_norm": 2.3576571941375732, + "learning_rate": 2.2801674914350972e-06, + "loss": 1.3825, + "step": 1737 + }, + { + "epoch": 1.7853107344632768, + "grad_norm": 2.355811595916748, + "learning_rate": 2.2782641796726306e-06, + "loss": 1.3716, + "step": 1738 + }, + { + "epoch": 1.7863379558294814, + "grad_norm": 2.3596901893615723, + "learning_rate": 2.276360867910164e-06, + "loss": 1.4064, + "step": 1739 + }, + { + "epoch": 1.7873651771956856, + "grad_norm": 2.301748514175415, + "learning_rate": 2.2744575561476973e-06, + "loss": 1.3358, + "step": 1740 + }, + { + "epoch": 1.7873651771956856, + "eval_loss": 1.7705498933792114, + "eval_runtime": 37.9787, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1740 + }, + { + "epoch": 1.7883923985618901, + "grad_norm": 2.3325891494750977, + "learning_rate": 2.2725542443852307e-06, + "loss": 1.3749, + "step": 1741 + }, + { + "epoch": 1.7894196199280945, + "grad_norm": 2.3880693912506104, + "learning_rate": 2.2706509326227636e-06, + "loss": 1.4153, + "step": 1742 + }, + { + "epoch": 1.7904468412942989, + "grad_norm": 2.3157029151916504, + "learning_rate": 2.268747620860297e-06, + "loss": 1.3899, + "step": 1743 + }, + { + "epoch": 1.7914740626605035, + "grad_norm": 2.4292151927948, + "learning_rate": 2.2668443090978303e-06, + "loss": 1.3804, + "step": 1744 + }, + { + "epoch": 1.7925012840267076, + "grad_norm": 2.4811818599700928, + "learning_rate": 2.2649409973353637e-06, + "loss": 1.4036, + "step": 1745 + }, + { + "epoch": 1.7935285053929122, + "grad_norm": 2.260420799255371, + "learning_rate": 2.263037685572897e-06, + "loss": 1.3616, + "step": 1746 + }, + { + "epoch": 1.7945557267591166, + "grad_norm": 2.262769937515259, + "learning_rate": 2.26113437381043e-06, + "loss": 1.351, + "step": 1747 + }, + { + "epoch": 1.795582948125321, + "grad_norm": 2.481602191925049, + "learning_rate": 2.259231062047964e-06, + "loss": 1.3816, + "step": 1748 + }, + { + "epoch": 1.7966101694915255, + "grad_norm": 2.584700345993042, + "learning_rate": 2.257327750285497e-06, + "loss": 1.4026, + "step": 1749 + }, + { + "epoch": 1.7976373908577299, + "grad_norm": 2.3472707271575928, + "learning_rate": 2.2554244385230305e-06, + "loss": 1.37, + "step": 1750 + }, + { + "epoch": 1.7976373908577299, + "eval_loss": 1.7779643535614014, + "eval_runtime": 38.0042, + "eval_samples_per_second": 26.313, + "eval_steps_per_second": 1.105, + "step": 1750 + }, + { + "epoch": 1.7986646122239343, + "grad_norm": 2.3616456985473633, + "learning_rate": 2.2535211267605635e-06, + "loss": 1.3713, + "step": 1751 + }, + { + "epoch": 1.7996918335901386, + "grad_norm": 2.4747135639190674, + "learning_rate": 2.251617814998097e-06, + "loss": 1.3797, + "step": 1752 + }, + { + "epoch": 1.800719054956343, + "grad_norm": 2.396944284439087, + "learning_rate": 2.24971450323563e-06, + "loss": 1.3081, + "step": 1753 + }, + { + "epoch": 1.8017462763225476, + "grad_norm": 2.5272505283355713, + "learning_rate": 2.2478111914731636e-06, + "loss": 1.3888, + "step": 1754 + }, + { + "epoch": 1.802773497688752, + "grad_norm": 2.319617748260498, + "learning_rate": 2.245907879710697e-06, + "loss": 1.2698, + "step": 1755 + }, + { + "epoch": 1.8038007190549563, + "grad_norm": 2.4994258880615234, + "learning_rate": 2.24400456794823e-06, + "loss": 1.4026, + "step": 1756 + }, + { + "epoch": 1.804827940421161, + "grad_norm": 2.3463289737701416, + "learning_rate": 2.2421012561857632e-06, + "loss": 1.3682, + "step": 1757 + }, + { + "epoch": 1.805855161787365, + "grad_norm": 2.3547732830047607, + "learning_rate": 2.240197944423297e-06, + "loss": 1.3486, + "step": 1758 + }, + { + "epoch": 1.8068823831535696, + "grad_norm": 2.381570816040039, + "learning_rate": 2.23829463266083e-06, + "loss": 1.3531, + "step": 1759 + }, + { + "epoch": 1.807909604519774, + "grad_norm": 2.2834391593933105, + "learning_rate": 2.2363913208983633e-06, + "loss": 1.3467, + "step": 1760 + }, + { + "epoch": 1.807909604519774, + "eval_loss": 1.7836403846740723, + "eval_runtime": 37.966, + "eval_samples_per_second": 26.339, + "eval_steps_per_second": 1.106, + "step": 1760 + }, + { + "epoch": 1.8089368258859784, + "grad_norm": 2.3574178218841553, + "learning_rate": 2.2344880091358967e-06, + "loss": 1.3867, + "step": 1761 + }, + { + "epoch": 1.809964047252183, + "grad_norm": 2.3440022468566895, + "learning_rate": 2.23258469737343e-06, + "loss": 1.3652, + "step": 1762 + }, + { + "epoch": 1.810991268618387, + "grad_norm": 2.4115090370178223, + "learning_rate": 2.2306813856109634e-06, + "loss": 1.4059, + "step": 1763 + }, + { + "epoch": 1.8120184899845917, + "grad_norm": 2.3989944458007812, + "learning_rate": 2.2287780738484963e-06, + "loss": 1.3572, + "step": 1764 + }, + { + "epoch": 1.813045711350796, + "grad_norm": 2.394892930984497, + "learning_rate": 2.2268747620860297e-06, + "loss": 1.433, + "step": 1765 + }, + { + "epoch": 1.8140729327170004, + "grad_norm": 2.5118515491485596, + "learning_rate": 2.224971450323563e-06, + "loss": 1.3714, + "step": 1766 + }, + { + "epoch": 1.815100154083205, + "grad_norm": 2.392695426940918, + "learning_rate": 2.2230681385610964e-06, + "loss": 1.3061, + "step": 1767 + }, + { + "epoch": 1.8161273754494094, + "grad_norm": 2.4015321731567383, + "learning_rate": 2.2211648267986298e-06, + "loss": 1.3681, + "step": 1768 + }, + { + "epoch": 1.8171545968156138, + "grad_norm": 2.3518455028533936, + "learning_rate": 2.219261515036163e-06, + "loss": 1.3549, + "step": 1769 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 2.40037202835083, + "learning_rate": 2.2173582032736965e-06, + "loss": 1.3882, + "step": 1770 + }, + { + "epoch": 1.8181818181818183, + "eval_loss": 1.7858384847640991, + "eval_runtime": 37.9787, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1770 + }, + { + "epoch": 1.8192090395480225, + "grad_norm": 2.4189655780792236, + "learning_rate": 2.21545489151123e-06, + "loss": 1.4019, + "step": 1771 + }, + { + "epoch": 1.820236260914227, + "grad_norm": 2.4965226650238037, + "learning_rate": 2.2135515797487632e-06, + "loss": 1.3703, + "step": 1772 + }, + { + "epoch": 1.8212634822804314, + "grad_norm": 2.4330036640167236, + "learning_rate": 2.211648267986296e-06, + "loss": 1.3689, + "step": 1773 + }, + { + "epoch": 1.8222907036466358, + "grad_norm": 2.3654327392578125, + "learning_rate": 2.2097449562238295e-06, + "loss": 1.3677, + "step": 1774 + }, + { + "epoch": 1.8233179250128404, + "grad_norm": 2.4047489166259766, + "learning_rate": 2.207841644461363e-06, + "loss": 1.3626, + "step": 1775 + }, + { + "epoch": 1.8243451463790445, + "grad_norm": 2.563310146331787, + "learning_rate": 2.2059383326988963e-06, + "loss": 1.424, + "step": 1776 + }, + { + "epoch": 1.8253723677452491, + "grad_norm": 2.506540536880493, + "learning_rate": 2.2040350209364296e-06, + "loss": 1.3727, + "step": 1777 + }, + { + "epoch": 1.8263995891114535, + "grad_norm": 2.271742105484009, + "learning_rate": 2.2021317091739626e-06, + "loss": 1.3412, + "step": 1778 + }, + { + "epoch": 1.8274268104776579, + "grad_norm": 2.3172881603240967, + "learning_rate": 2.2002283974114963e-06, + "loss": 1.3812, + "step": 1779 + }, + { + "epoch": 1.8284540318438625, + "grad_norm": 2.3927204608917236, + "learning_rate": 2.1983250856490297e-06, + "loss": 1.3148, + "step": 1780 + }, + { + "epoch": 1.8284540318438625, + "eval_loss": 1.7755651473999023, + "eval_runtime": 37.9896, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1780 + }, + { + "epoch": 1.8294812532100666, + "grad_norm": 2.391972303390503, + "learning_rate": 2.1964217738865626e-06, + "loss": 1.3553, + "step": 1781 + }, + { + "epoch": 1.8305084745762712, + "grad_norm": 2.3727471828460693, + "learning_rate": 2.194518462124096e-06, + "loss": 1.377, + "step": 1782 + }, + { + "epoch": 1.8315356959424756, + "grad_norm": 2.5614147186279297, + "learning_rate": 2.1926151503616294e-06, + "loss": 1.4188, + "step": 1783 + }, + { + "epoch": 1.83256291730868, + "grad_norm": 2.344714641571045, + "learning_rate": 2.1907118385991627e-06, + "loss": 1.3877, + "step": 1784 + }, + { + "epoch": 1.8335901386748845, + "grad_norm": 2.34287428855896, + "learning_rate": 2.188808526836696e-06, + "loss": 1.351, + "step": 1785 + }, + { + "epoch": 1.8346173600410889, + "grad_norm": 2.343554735183716, + "learning_rate": 2.1869052150742295e-06, + "loss": 1.382, + "step": 1786 + }, + { + "epoch": 1.8356445814072933, + "grad_norm": 2.2418034076690674, + "learning_rate": 2.1850019033117624e-06, + "loss": 1.3062, + "step": 1787 + }, + { + "epoch": 1.8366718027734978, + "grad_norm": 2.4358294010162354, + "learning_rate": 2.1830985915492958e-06, + "loss": 1.3681, + "step": 1788 + }, + { + "epoch": 1.837699024139702, + "grad_norm": 2.460731029510498, + "learning_rate": 2.1811952797868295e-06, + "loss": 1.3791, + "step": 1789 + }, + { + "epoch": 1.8387262455059066, + "grad_norm": 2.4244463443756104, + "learning_rate": 2.1792919680243625e-06, + "loss": 1.3528, + "step": 1790 + }, + { + "epoch": 1.8387262455059066, + "eval_loss": 1.764306664466858, + "eval_runtime": 37.9906, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 1790 + }, + { + "epoch": 1.839753466872111, + "grad_norm": 2.3808743953704834, + "learning_rate": 2.177388656261896e-06, + "loss": 1.3882, + "step": 1791 + }, + { + "epoch": 1.8407806882383153, + "grad_norm": 2.443690538406372, + "learning_rate": 2.1754853444994292e-06, + "loss": 1.4542, + "step": 1792 + }, + { + "epoch": 1.84180790960452, + "grad_norm": 2.3968520164489746, + "learning_rate": 2.1735820327369626e-06, + "loss": 1.2815, + "step": 1793 + }, + { + "epoch": 1.842835130970724, + "grad_norm": 2.5139951705932617, + "learning_rate": 2.171678720974496e-06, + "loss": 1.3986, + "step": 1794 + }, + { + "epoch": 1.8438623523369286, + "grad_norm": 2.3268849849700928, + "learning_rate": 2.169775409212029e-06, + "loss": 1.3923, + "step": 1795 + }, + { + "epoch": 1.844889573703133, + "grad_norm": 2.424268960952759, + "learning_rate": 2.1678720974495622e-06, + "loss": 1.4046, + "step": 1796 + }, + { + "epoch": 1.8459167950693374, + "grad_norm": 2.3144543170928955, + "learning_rate": 2.1659687856870956e-06, + "loss": 1.391, + "step": 1797 + }, + { + "epoch": 1.846944016435542, + "grad_norm": 2.4157726764678955, + "learning_rate": 2.164065473924629e-06, + "loss": 1.4084, + "step": 1798 + }, + { + "epoch": 1.8479712378017463, + "grad_norm": 2.3993160724639893, + "learning_rate": 2.1621621621621623e-06, + "loss": 1.3372, + "step": 1799 + }, + { + "epoch": 1.8489984591679507, + "grad_norm": 2.3267669677734375, + "learning_rate": 2.1602588503996957e-06, + "loss": 1.3899, + "step": 1800 + }, + { + "epoch": 1.8489984591679507, + "eval_loss": 1.7611773014068604, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 1800 + }, + { + "epoch": 1.850025680534155, + "grad_norm": 2.488994836807251, + "learning_rate": 2.158355538637229e-06, + "loss": 1.3647, + "step": 1801 + }, + { + "epoch": 1.8510529019003594, + "grad_norm": 2.32352876663208, + "learning_rate": 2.1564522268747624e-06, + "loss": 1.3615, + "step": 1802 + }, + { + "epoch": 1.852080123266564, + "grad_norm": 2.667987585067749, + "learning_rate": 2.1545489151122958e-06, + "loss": 1.4388, + "step": 1803 + }, + { + "epoch": 1.8531073446327684, + "grad_norm": 2.4570565223693848, + "learning_rate": 2.1526456033498287e-06, + "loss": 1.3939, + "step": 1804 + }, + { + "epoch": 1.8541345659989728, + "grad_norm": 2.4037978649139404, + "learning_rate": 2.150742291587362e-06, + "loss": 1.3834, + "step": 1805 + }, + { + "epoch": 1.8551617873651773, + "grad_norm": 2.42080020904541, + "learning_rate": 2.1488389798248954e-06, + "loss": 1.3758, + "step": 1806 + }, + { + "epoch": 1.8561890087313815, + "grad_norm": 2.3361966609954834, + "learning_rate": 2.146935668062429e-06, + "loss": 1.4024, + "step": 1807 + }, + { + "epoch": 1.857216230097586, + "grad_norm": 2.251089334487915, + "learning_rate": 2.145032356299962e-06, + "loss": 1.3365, + "step": 1808 + }, + { + "epoch": 1.8582434514637904, + "grad_norm": 2.4046289920806885, + "learning_rate": 2.1431290445374955e-06, + "loss": 1.3831, + "step": 1809 + }, + { + "epoch": 1.8592706728299948, + "grad_norm": 2.2613766193389893, + "learning_rate": 2.141225732775029e-06, + "loss": 1.3423, + "step": 1810 + }, + { + "epoch": 1.8592706728299948, + "eval_loss": 1.752854347229004, + "eval_runtime": 37.6469, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 1810 + }, + { + "epoch": 1.8602978941961994, + "grad_norm": 2.377535581588745, + "learning_rate": 2.1393224210125623e-06, + "loss": 1.3796, + "step": 1811 + }, + { + "epoch": 1.8613251155624035, + "grad_norm": 2.3959462642669678, + "learning_rate": 2.137419109250095e-06, + "loss": 1.3664, + "step": 1812 + }, + { + "epoch": 1.8623523369286081, + "grad_norm": 2.5317859649658203, + "learning_rate": 2.1355157974876286e-06, + "loss": 1.418, + "step": 1813 + }, + { + "epoch": 1.8633795582948125, + "grad_norm": 2.5266010761260986, + "learning_rate": 2.133612485725162e-06, + "loss": 1.4117, + "step": 1814 + }, + { + "epoch": 1.8644067796610169, + "grad_norm": 2.3911185264587402, + "learning_rate": 2.1317091739626953e-06, + "loss": 1.3206, + "step": 1815 + }, + { + "epoch": 1.8654340010272215, + "grad_norm": 2.358596086502075, + "learning_rate": 2.1298058622002286e-06, + "loss": 1.3475, + "step": 1816 + }, + { + "epoch": 1.8664612223934258, + "grad_norm": 2.348175048828125, + "learning_rate": 2.1279025504377616e-06, + "loss": 1.3769, + "step": 1817 + }, + { + "epoch": 1.8674884437596302, + "grad_norm": 2.417470932006836, + "learning_rate": 2.125999238675295e-06, + "loss": 1.3667, + "step": 1818 + }, + { + "epoch": 1.8685156651258348, + "grad_norm": 2.5871050357818604, + "learning_rate": 2.1240959269128287e-06, + "loss": 1.4139, + "step": 1819 + }, + { + "epoch": 1.869542886492039, + "grad_norm": 2.384218692779541, + "learning_rate": 2.122192615150362e-06, + "loss": 1.3755, + "step": 1820 + }, + { + "epoch": 1.869542886492039, + "eval_loss": 1.7497111558914185, + "eval_runtime": 37.7655, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 1820 + }, + { + "epoch": 1.8705701078582435, + "grad_norm": 2.354877233505249, + "learning_rate": 2.120289303387895e-06, + "loss": 1.3174, + "step": 1821 + }, + { + "epoch": 1.8715973292244479, + "grad_norm": 2.4020509719848633, + "learning_rate": 2.1183859916254284e-06, + "loss": 1.3699, + "step": 1822 + }, + { + "epoch": 1.8726245505906522, + "grad_norm": 2.2876627445220947, + "learning_rate": 2.1164826798629618e-06, + "loss": 1.4053, + "step": 1823 + }, + { + "epoch": 1.8736517719568568, + "grad_norm": 2.2526464462280273, + "learning_rate": 2.114579368100495e-06, + "loss": 1.3569, + "step": 1824 + }, + { + "epoch": 1.874678993323061, + "grad_norm": 2.340773344039917, + "learning_rate": 2.1126760563380285e-06, + "loss": 1.3439, + "step": 1825 + }, + { + "epoch": 1.8757062146892656, + "grad_norm": 2.315812110900879, + "learning_rate": 2.1107727445755614e-06, + "loss": 1.3129, + "step": 1826 + }, + { + "epoch": 1.87673343605547, + "grad_norm": 2.480487108230591, + "learning_rate": 2.1088694328130948e-06, + "loss": 1.4372, + "step": 1827 + }, + { + "epoch": 1.8777606574216743, + "grad_norm": 2.267662286758423, + "learning_rate": 2.106966121050628e-06, + "loss": 1.3429, + "step": 1828 + }, + { + "epoch": 1.878787878787879, + "grad_norm": 2.383023977279663, + "learning_rate": 2.1050628092881615e-06, + "loss": 1.3963, + "step": 1829 + }, + { + "epoch": 1.879815100154083, + "grad_norm": 2.4784996509552, + "learning_rate": 2.103159497525695e-06, + "loss": 1.4405, + "step": 1830 + }, + { + "epoch": 1.879815100154083, + "eval_loss": 1.7609645128250122, + "eval_runtime": 37.904, + "eval_samples_per_second": 26.382, + "eval_steps_per_second": 1.108, + "step": 1830 + }, + { + "epoch": 1.8808423215202876, + "grad_norm": 2.359614133834839, + "learning_rate": 2.1012561857632282e-06, + "loss": 1.3577, + "step": 1831 + }, + { + "epoch": 1.881869542886492, + "grad_norm": 2.452862501144409, + "learning_rate": 2.0993528740007616e-06, + "loss": 1.4082, + "step": 1832 + }, + { + "epoch": 1.8828967642526964, + "grad_norm": 2.384011745452881, + "learning_rate": 2.097449562238295e-06, + "loss": 1.4175, + "step": 1833 + }, + { + "epoch": 1.883923985618901, + "grad_norm": 2.4906508922576904, + "learning_rate": 2.095546250475828e-06, + "loss": 1.3693, + "step": 1834 + }, + { + "epoch": 1.8849512069851053, + "grad_norm": 2.510040521621704, + "learning_rate": 2.0936429387133613e-06, + "loss": 1.3703, + "step": 1835 + }, + { + "epoch": 1.8859784283513097, + "grad_norm": 2.3438405990600586, + "learning_rate": 2.0917396269508946e-06, + "loss": 1.4311, + "step": 1836 + }, + { + "epoch": 1.8870056497175143, + "grad_norm": 2.532226800918579, + "learning_rate": 2.089836315188428e-06, + "loss": 1.3986, + "step": 1837 + }, + { + "epoch": 1.8880328710837184, + "grad_norm": 2.4368748664855957, + "learning_rate": 2.0879330034259614e-06, + "loss": 1.4025, + "step": 1838 + }, + { + "epoch": 1.889060092449923, + "grad_norm": 2.4487054347991943, + "learning_rate": 2.0860296916634947e-06, + "loss": 1.3562, + "step": 1839 + }, + { + "epoch": 1.8900873138161274, + "grad_norm": 2.6636273860931396, + "learning_rate": 2.084126379901028e-06, + "loss": 1.4384, + "step": 1840 + }, + { + "epoch": 1.8900873138161274, + "eval_loss": 1.7675000429153442, + "eval_runtime": 37.9371, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 1840 + }, + { + "epoch": 1.8911145351823317, + "grad_norm": 2.497738838195801, + "learning_rate": 2.0822230681385614e-06, + "loss": 1.3462, + "step": 1841 + }, + { + "epoch": 1.8921417565485363, + "grad_norm": 2.4419288635253906, + "learning_rate": 2.080319756376095e-06, + "loss": 1.3974, + "step": 1842 + }, + { + "epoch": 1.8931689779147405, + "grad_norm": 2.3671085834503174, + "learning_rate": 2.0784164446136277e-06, + "loss": 1.3917, + "step": 1843 + }, + { + "epoch": 1.894196199280945, + "grad_norm": 2.3837625980377197, + "learning_rate": 2.076513132851161e-06, + "loss": 1.3663, + "step": 1844 + }, + { + "epoch": 1.8952234206471494, + "grad_norm": 2.590935468673706, + "learning_rate": 2.0746098210886945e-06, + "loss": 1.3962, + "step": 1845 + }, + { + "epoch": 1.8962506420133538, + "grad_norm": 2.2917118072509766, + "learning_rate": 2.072706509326228e-06, + "loss": 1.3641, + "step": 1846 + }, + { + "epoch": 1.8972778633795584, + "grad_norm": 2.4366025924682617, + "learning_rate": 2.070803197563761e-06, + "loss": 1.3749, + "step": 1847 + }, + { + "epoch": 1.8983050847457628, + "grad_norm": 2.385152816772461, + "learning_rate": 2.068899885801294e-06, + "loss": 1.3655, + "step": 1848 + }, + { + "epoch": 1.8993323061119671, + "grad_norm": 2.473550796508789, + "learning_rate": 2.066996574038828e-06, + "loss": 1.3917, + "step": 1849 + }, + { + "epoch": 1.9003595274781715, + "grad_norm": 2.360003709793091, + "learning_rate": 2.0650932622763613e-06, + "loss": 1.4089, + "step": 1850 + }, + { + "epoch": 1.9003595274781715, + "eval_loss": 1.7553730010986328, + "eval_runtime": 37.9412, + "eval_samples_per_second": 26.357, + "eval_steps_per_second": 1.107, + "step": 1850 + }, + { + "epoch": 1.9013867488443759, + "grad_norm": 2.3891544342041016, + "learning_rate": 2.0631899505138942e-06, + "loss": 1.3033, + "step": 1851 + }, + { + "epoch": 1.9024139702105805, + "grad_norm": 2.330413341522217, + "learning_rate": 2.0612866387514276e-06, + "loss": 1.3446, + "step": 1852 + }, + { + "epoch": 1.9034411915767848, + "grad_norm": 2.3061635494232178, + "learning_rate": 2.059383326988961e-06, + "loss": 1.3872, + "step": 1853 + }, + { + "epoch": 1.9044684129429892, + "grad_norm": 2.4351041316986084, + "learning_rate": 2.0574800152264943e-06, + "loss": 1.362, + "step": 1854 + }, + { + "epoch": 1.9054956343091938, + "grad_norm": 2.447535991668701, + "learning_rate": 2.0555767034640277e-06, + "loss": 1.3417, + "step": 1855 + }, + { + "epoch": 1.906522855675398, + "grad_norm": 2.369133472442627, + "learning_rate": 2.053673391701561e-06, + "loss": 1.3181, + "step": 1856 + }, + { + "epoch": 1.9075500770416025, + "grad_norm": 2.3961751461029053, + "learning_rate": 2.051770079939094e-06, + "loss": 1.3913, + "step": 1857 + }, + { + "epoch": 1.9085772984078069, + "grad_norm": 2.40244460105896, + "learning_rate": 2.0498667681766273e-06, + "loss": 1.3447, + "step": 1858 + }, + { + "epoch": 1.9096045197740112, + "grad_norm": 2.4332613945007324, + "learning_rate": 2.047963456414161e-06, + "loss": 1.3427, + "step": 1859 + }, + { + "epoch": 1.9106317411402158, + "grad_norm": 2.3410897254943848, + "learning_rate": 2.046060144651694e-06, + "loss": 1.3697, + "step": 1860 + }, + { + "epoch": 1.9106317411402158, + "eval_loss": 1.74896240234375, + "eval_runtime": 37.9315, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 1860 + }, + { + "epoch": 1.91165896250642, + "grad_norm": 2.3965389728546143, + "learning_rate": 2.0441568328892274e-06, + "loss": 1.3752, + "step": 1861 + }, + { + "epoch": 1.9126861838726246, + "grad_norm": 2.3173274993896484, + "learning_rate": 2.0422535211267608e-06, + "loss": 1.3807, + "step": 1862 + }, + { + "epoch": 1.913713405238829, + "grad_norm": 2.2883224487304688, + "learning_rate": 2.040350209364294e-06, + "loss": 1.3551, + "step": 1863 + }, + { + "epoch": 1.9147406266050333, + "grad_norm": 2.2770445346832275, + "learning_rate": 2.0384468976018275e-06, + "loss": 1.3405, + "step": 1864 + }, + { + "epoch": 1.915767847971238, + "grad_norm": 2.3694701194763184, + "learning_rate": 2.0365435858393604e-06, + "loss": 1.3538, + "step": 1865 + }, + { + "epoch": 1.9167950693374423, + "grad_norm": 2.4186787605285645, + "learning_rate": 2.034640274076894e-06, + "loss": 1.353, + "step": 1866 + }, + { + "epoch": 1.9178222907036466, + "grad_norm": 2.4059510231018066, + "learning_rate": 2.032736962314427e-06, + "loss": 1.4149, + "step": 1867 + }, + { + "epoch": 1.9188495120698512, + "grad_norm": 2.5529062747955322, + "learning_rate": 2.0308336505519605e-06, + "loss": 1.3669, + "step": 1868 + }, + { + "epoch": 1.9198767334360554, + "grad_norm": 2.3925650119781494, + "learning_rate": 2.028930338789494e-06, + "loss": 1.4178, + "step": 1869 + }, + { + "epoch": 1.92090395480226, + "grad_norm": 2.5203816890716553, + "learning_rate": 2.0270270270270273e-06, + "loss": 1.3418, + "step": 1870 + }, + { + "epoch": 1.92090395480226, + "eval_loss": 1.7473065853118896, + "eval_runtime": 37.9634, + "eval_samples_per_second": 26.341, + "eval_steps_per_second": 1.106, + "step": 1870 + }, + { + "epoch": 1.9219311761684643, + "grad_norm": 2.4792747497558594, + "learning_rate": 2.0251237152645606e-06, + "loss": 1.3611, + "step": 1871 + }, + { + "epoch": 1.9229583975346687, + "grad_norm": 2.357645273208618, + "learning_rate": 2.023220403502094e-06, + "loss": 1.3433, + "step": 1872 + }, + { + "epoch": 1.9239856189008733, + "grad_norm": 2.284149408340454, + "learning_rate": 2.0213170917396273e-06, + "loss": 1.3157, + "step": 1873 + }, + { + "epoch": 1.9250128402670774, + "grad_norm": 2.3971638679504395, + "learning_rate": 2.0194137799771603e-06, + "loss": 1.3435, + "step": 1874 + }, + { + "epoch": 1.926040061633282, + "grad_norm": 2.605325222015381, + "learning_rate": 2.0175104682146937e-06, + "loss": 1.3619, + "step": 1875 + }, + { + "epoch": 1.9270672829994864, + "grad_norm": 2.3957643508911133, + "learning_rate": 2.015607156452227e-06, + "loss": 1.3078, + "step": 1876 + }, + { + "epoch": 1.9280945043656907, + "grad_norm": 2.3887429237365723, + "learning_rate": 2.0137038446897604e-06, + "loss": 1.3447, + "step": 1877 + }, + { + "epoch": 1.9291217257318953, + "grad_norm": 2.31253719329834, + "learning_rate": 2.0118005329272937e-06, + "loss": 1.3467, + "step": 1878 + }, + { + "epoch": 1.9301489470980995, + "grad_norm": 2.3084523677825928, + "learning_rate": 2.0098972211648267e-06, + "loss": 1.3466, + "step": 1879 + }, + { + "epoch": 1.931176168464304, + "grad_norm": 2.3024210929870605, + "learning_rate": 2.0079939094023605e-06, + "loss": 1.3525, + "step": 1880 + }, + { + "epoch": 1.931176168464304, + "eval_loss": 1.7498610019683838, + "eval_runtime": 37.9363, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 1880 + }, + { + "epoch": 1.9322033898305084, + "grad_norm": 2.342660665512085, + "learning_rate": 2.006090597639894e-06, + "loss": 1.4013, + "step": 1881 + }, + { + "epoch": 1.9332306111967128, + "grad_norm": 2.4165151119232178, + "learning_rate": 2.0041872858774268e-06, + "loss": 1.4154, + "step": 1882 + }, + { + "epoch": 1.9342578325629174, + "grad_norm": 2.3674707412719727, + "learning_rate": 2.00228397411496e-06, + "loss": 1.3903, + "step": 1883 + }, + { + "epoch": 1.9352850539291218, + "grad_norm": 2.2478187084198, + "learning_rate": 2.0003806623524935e-06, + "loss": 1.3708, + "step": 1884 + }, + { + "epoch": 1.9363122752953261, + "grad_norm": 2.328073740005493, + "learning_rate": 1.998477350590027e-06, + "loss": 1.3013, + "step": 1885 + }, + { + "epoch": 1.9373394966615307, + "grad_norm": 2.4772889614105225, + "learning_rate": 1.9965740388275602e-06, + "loss": 1.3818, + "step": 1886 + }, + { + "epoch": 1.9383667180277349, + "grad_norm": 2.456451892852783, + "learning_rate": 1.994670727065093e-06, + "loss": 1.3806, + "step": 1887 + }, + { + "epoch": 1.9393939393939394, + "grad_norm": 2.325561046600342, + "learning_rate": 1.9927674153026265e-06, + "loss": 1.3262, + "step": 1888 + }, + { + "epoch": 1.9404211607601438, + "grad_norm": 2.447673797607422, + "learning_rate": 1.99086410354016e-06, + "loss": 1.3776, + "step": 1889 + }, + { + "epoch": 1.9414483821263482, + "grad_norm": 2.4231903553009033, + "learning_rate": 1.9889607917776937e-06, + "loss": 1.3593, + "step": 1890 + }, + { + "epoch": 1.9414483821263482, + "eval_loss": 1.7612230777740479, + "eval_runtime": 37.9427, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 1890 + }, + { + "epoch": 1.9424756034925528, + "grad_norm": 2.5399789810180664, + "learning_rate": 1.9870574800152266e-06, + "loss": 1.3737, + "step": 1891 + }, + { + "epoch": 1.943502824858757, + "grad_norm": 2.3155572414398193, + "learning_rate": 1.98515416825276e-06, + "loss": 1.388, + "step": 1892 + }, + { + "epoch": 1.9445300462249615, + "grad_norm": 2.33693265914917, + "learning_rate": 1.9832508564902933e-06, + "loss": 1.3734, + "step": 1893 + }, + { + "epoch": 1.9455572675911659, + "grad_norm": 2.2426419258117676, + "learning_rate": 1.9813475447278267e-06, + "loss": 1.3011, + "step": 1894 + }, + { + "epoch": 1.9465844889573702, + "grad_norm": 2.4306788444519043, + "learning_rate": 1.97944423296536e-06, + "loss": 1.3721, + "step": 1895 + }, + { + "epoch": 1.9476117103235748, + "grad_norm": 2.329252004623413, + "learning_rate": 1.977540921202893e-06, + "loss": 1.363, + "step": 1896 + }, + { + "epoch": 1.9486389316897792, + "grad_norm": 2.3523757457733154, + "learning_rate": 1.9756376094404264e-06, + "loss": 1.3485, + "step": 1897 + }, + { + "epoch": 1.9496661530559836, + "grad_norm": 2.3240344524383545, + "learning_rate": 1.9737342976779597e-06, + "loss": 1.3258, + "step": 1898 + }, + { + "epoch": 1.950693374422188, + "grad_norm": 2.4068586826324463, + "learning_rate": 1.971830985915493e-06, + "loss": 1.406, + "step": 1899 + }, + { + "epoch": 1.9517205957883923, + "grad_norm": 2.385835886001587, + "learning_rate": 1.9699276741530264e-06, + "loss": 1.3641, + "step": 1900 + }, + { + "epoch": 1.9517205957883923, + "eval_loss": 1.747411847114563, + "eval_runtime": 37.9419, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 1900 + }, + { + "epoch": 1.9527478171545969, + "grad_norm": 2.3512096405029297, + "learning_rate": 1.96802436239056e-06, + "loss": 1.3647, + "step": 1901 + }, + { + "epoch": 1.9537750385208013, + "grad_norm": 2.3813812732696533, + "learning_rate": 1.966121050628093e-06, + "loss": 1.3422, + "step": 1902 + }, + { + "epoch": 1.9548022598870056, + "grad_norm": 2.404499053955078, + "learning_rate": 1.9642177388656265e-06, + "loss": 1.3918, + "step": 1903 + }, + { + "epoch": 1.9558294812532102, + "grad_norm": 2.4429967403411865, + "learning_rate": 1.9623144271031595e-06, + "loss": 1.4049, + "step": 1904 + }, + { + "epoch": 1.9568567026194144, + "grad_norm": 2.505833864212036, + "learning_rate": 1.960411115340693e-06, + "loss": 1.4112, + "step": 1905 + }, + { + "epoch": 1.957883923985619, + "grad_norm": 2.4441304206848145, + "learning_rate": 1.958507803578226e-06, + "loss": 1.3648, + "step": 1906 + }, + { + "epoch": 1.9589111453518233, + "grad_norm": 2.4598183631896973, + "learning_rate": 1.9566044918157596e-06, + "loss": 1.3879, + "step": 1907 + }, + { + "epoch": 1.9599383667180277, + "grad_norm": 2.3862075805664062, + "learning_rate": 1.954701180053293e-06, + "loss": 1.3831, + "step": 1908 + }, + { + "epoch": 1.9609655880842323, + "grad_norm": 2.434634208679199, + "learning_rate": 1.952797868290826e-06, + "loss": 1.3499, + "step": 1909 + }, + { + "epoch": 1.9619928094504364, + "grad_norm": 2.4674196243286133, + "learning_rate": 1.9508945565283596e-06, + "loss": 1.3941, + "step": 1910 + }, + { + "epoch": 1.9619928094504364, + "eval_loss": 1.734555959701538, + "eval_runtime": 37.6492, + "eval_samples_per_second": 26.561, + "eval_steps_per_second": 1.116, + "step": 1910 + }, + { + "epoch": 1.963020030816641, + "grad_norm": 2.336597442626953, + "learning_rate": 1.948991244765893e-06, + "loss": 1.3731, + "step": 1911 + }, + { + "epoch": 1.9640472521828454, + "grad_norm": 2.4431259632110596, + "learning_rate": 1.9470879330034264e-06, + "loss": 1.3926, + "step": 1912 + }, + { + "epoch": 1.9650744735490497, + "grad_norm": 2.4182162284851074, + "learning_rate": 1.9451846212409593e-06, + "loss": 1.3291, + "step": 1913 + }, + { + "epoch": 1.9661016949152543, + "grad_norm": 2.382519483566284, + "learning_rate": 1.9432813094784927e-06, + "loss": 1.3891, + "step": 1914 + }, + { + "epoch": 1.9671289162814587, + "grad_norm": 2.3186089992523193, + "learning_rate": 1.941377997716026e-06, + "loss": 1.3206, + "step": 1915 + }, + { + "epoch": 1.968156137647663, + "grad_norm": 2.32810378074646, + "learning_rate": 1.9394746859535594e-06, + "loss": 1.3564, + "step": 1916 + }, + { + "epoch": 1.9691833590138677, + "grad_norm": 2.3164830207824707, + "learning_rate": 1.9375713741910928e-06, + "loss": 1.3154, + "step": 1917 + }, + { + "epoch": 1.9702105803800718, + "grad_norm": 2.3495121002197266, + "learning_rate": 1.9356680624286257e-06, + "loss": 1.3533, + "step": 1918 + }, + { + "epoch": 1.9712378017462764, + "grad_norm": 2.307860851287842, + "learning_rate": 1.933764750666159e-06, + "loss": 1.3142, + "step": 1919 + }, + { + "epoch": 1.9722650231124808, + "grad_norm": 2.2913570404052734, + "learning_rate": 1.931861438903693e-06, + "loss": 1.3672, + "step": 1920 + }, + { + "epoch": 1.9722650231124808, + "eval_loss": 1.7303361892700195, + "eval_runtime": 37.7597, + "eval_samples_per_second": 26.483, + "eval_steps_per_second": 1.112, + "step": 1920 + }, + { + "epoch": 1.9732922444786851, + "grad_norm": 2.3222689628601074, + "learning_rate": 1.929958127141226e-06, + "loss": 1.3317, + "step": 1921 + }, + { + "epoch": 1.9743194658448897, + "grad_norm": 2.2708919048309326, + "learning_rate": 1.928054815378759e-06, + "loss": 1.3518, + "step": 1922 + }, + { + "epoch": 1.9753466872110939, + "grad_norm": 2.4408884048461914, + "learning_rate": 1.9261515036162925e-06, + "loss": 1.4332, + "step": 1923 + }, + { + "epoch": 1.9763739085772984, + "grad_norm": 2.469506025314331, + "learning_rate": 1.924248191853826e-06, + "loss": 1.3927, + "step": 1924 + }, + { + "epoch": 1.9774011299435028, + "grad_norm": 2.3409810066223145, + "learning_rate": 1.9223448800913592e-06, + "loss": 1.3692, + "step": 1925 + }, + { + "epoch": 1.9784283513097072, + "grad_norm": 2.293306589126587, + "learning_rate": 1.9204415683288926e-06, + "loss": 1.3491, + "step": 1926 + }, + { + "epoch": 1.9794555726759118, + "grad_norm": 2.428431272506714, + "learning_rate": 1.9185382565664255e-06, + "loss": 1.381, + "step": 1927 + }, + { + "epoch": 1.980482794042116, + "grad_norm": 2.4767544269561768, + "learning_rate": 1.916634944803959e-06, + "loss": 1.3953, + "step": 1928 + }, + { + "epoch": 1.9815100154083205, + "grad_norm": 2.44821834564209, + "learning_rate": 1.9147316330414923e-06, + "loss": 1.3746, + "step": 1929 + }, + { + "epoch": 1.9825372367745249, + "grad_norm": 2.349822759628296, + "learning_rate": 1.9128283212790256e-06, + "loss": 1.3619, + "step": 1930 + }, + { + "epoch": 1.9825372367745249, + "eval_loss": 1.7362234592437744, + "eval_runtime": 37.8754, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1930 + }, + { + "epoch": 1.9835644581407292, + "grad_norm": 2.5379276275634766, + "learning_rate": 1.910925009516559e-06, + "loss": 1.4051, + "step": 1931 + }, + { + "epoch": 1.9845916795069338, + "grad_norm": 2.329698324203491, + "learning_rate": 1.9090216977540924e-06, + "loss": 1.3693, + "step": 1932 + }, + { + "epoch": 1.9856189008731382, + "grad_norm": 2.424672842025757, + "learning_rate": 1.9071183859916257e-06, + "loss": 1.4073, + "step": 1933 + }, + { + "epoch": 1.9866461222393426, + "grad_norm": 2.480088949203491, + "learning_rate": 1.9052150742291589e-06, + "loss": 1.3816, + "step": 1934 + }, + { + "epoch": 1.9876733436055471, + "grad_norm": 2.4967830181121826, + "learning_rate": 1.9033117624666922e-06, + "loss": 1.3804, + "step": 1935 + }, + { + "epoch": 1.9887005649717513, + "grad_norm": 2.337944984436035, + "learning_rate": 1.9014084507042254e-06, + "loss": 1.3452, + "step": 1936 + }, + { + "epoch": 1.9897277863379559, + "grad_norm": 2.4424421787261963, + "learning_rate": 1.8995051389417587e-06, + "loss": 1.3382, + "step": 1937 + }, + { + "epoch": 1.9907550077041603, + "grad_norm": 2.2587194442749023, + "learning_rate": 1.8976018271792921e-06, + "loss": 1.3229, + "step": 1938 + }, + { + "epoch": 1.9917822290703646, + "grad_norm": 2.4167850017547607, + "learning_rate": 1.8956985154168253e-06, + "loss": 1.3828, + "step": 1939 + }, + { + "epoch": 1.9928094504365692, + "grad_norm": 2.397630214691162, + "learning_rate": 1.8937952036543588e-06, + "loss": 1.3499, + "step": 1940 + }, + { + "epoch": 1.9928094504365692, + "eval_loss": 1.7392780780792236, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 1940 + }, + { + "epoch": 1.9938366718027734, + "grad_norm": 2.3320090770721436, + "learning_rate": 1.8918918918918922e-06, + "loss": 1.3157, + "step": 1941 + }, + { + "epoch": 1.994863893168978, + "grad_norm": 2.4160780906677246, + "learning_rate": 1.8899885801294253e-06, + "loss": 1.3693, + "step": 1942 + }, + { + "epoch": 1.9958911145351823, + "grad_norm": 2.4204070568084717, + "learning_rate": 1.8880852683669587e-06, + "loss": 1.3787, + "step": 1943 + }, + { + "epoch": 1.9969183359013867, + "grad_norm": 2.3747220039367676, + "learning_rate": 1.886181956604492e-06, + "loss": 1.3449, + "step": 1944 + }, + { + "epoch": 1.9979455572675913, + "grad_norm": 2.454822063446045, + "learning_rate": 1.8842786448420252e-06, + "loss": 1.3478, + "step": 1945 + }, + { + "epoch": 1.9989727786337956, + "grad_norm": 2.3999180793762207, + "learning_rate": 1.8823753330795586e-06, + "loss": 1.3786, + "step": 1946 + }, + { + "epoch": 2.0, + "grad_norm": 2.295583963394165, + "learning_rate": 1.8804720213170917e-06, + "loss": 1.2788, + "step": 1947 + }, + { + "epoch": 2.0010272213662046, + "grad_norm": 2.5762782096862793, + "learning_rate": 1.878568709554625e-06, + "loss": 1.2981, + "step": 1948 + }, + { + "epoch": 2.0020544427324087, + "grad_norm": 2.7369227409362793, + "learning_rate": 1.8766653977921585e-06, + "loss": 1.3315, + "step": 1949 + }, + { + "epoch": 2.0030816640986133, + "grad_norm": 2.5014007091522217, + "learning_rate": 1.874762086029692e-06, + "loss": 1.3112, + "step": 1950 + }, + { + "epoch": 2.0030816640986133, + "eval_loss": 1.7656217813491821, + "eval_runtime": 37.9186, + "eval_samples_per_second": 26.372, + "eval_steps_per_second": 1.108, + "step": 1950 + }, + { + "epoch": 2.0041088854648175, + "grad_norm": 2.7794134616851807, + "learning_rate": 1.8728587742672252e-06, + "loss": 1.3328, + "step": 1951 + }, + { + "epoch": 2.005136106831022, + "grad_norm": 2.9862377643585205, + "learning_rate": 1.8709554625047585e-06, + "loss": 1.3467, + "step": 1952 + }, + { + "epoch": 2.0061633281972266, + "grad_norm": 2.8717260360717773, + "learning_rate": 1.8690521507422917e-06, + "loss": 1.2869, + "step": 1953 + }, + { + "epoch": 2.007190549563431, + "grad_norm": 2.6782584190368652, + "learning_rate": 1.867148838979825e-06, + "loss": 1.3081, + "step": 1954 + }, + { + "epoch": 2.0082177709296354, + "grad_norm": 2.804760456085205, + "learning_rate": 1.8652455272173584e-06, + "loss": 1.3089, + "step": 1955 + }, + { + "epoch": 2.00924499229584, + "grad_norm": 2.8827545642852783, + "learning_rate": 1.8633422154548916e-06, + "loss": 1.352, + "step": 1956 + }, + { + "epoch": 2.010272213662044, + "grad_norm": 2.6346447467803955, + "learning_rate": 1.861438903692425e-06, + "loss": 1.2582, + "step": 1957 + }, + { + "epoch": 2.0112994350282487, + "grad_norm": 2.690293073654175, + "learning_rate": 1.859535591929958e-06, + "loss": 1.2807, + "step": 1958 + }, + { + "epoch": 2.012326656394453, + "grad_norm": 2.8933660984039307, + "learning_rate": 1.8576322801674915e-06, + "loss": 1.3553, + "step": 1959 + }, + { + "epoch": 2.0133538777606574, + "grad_norm": 2.6514511108398438, + "learning_rate": 1.855728968405025e-06, + "loss": 1.2971, + "step": 1960 + }, + { + "epoch": 2.0133538777606574, + "eval_loss": 1.7628012895584106, + "eval_runtime": 37.9132, + "eval_samples_per_second": 26.376, + "eval_steps_per_second": 1.108, + "step": 1960 + }, + { + "epoch": 2.014381099126862, + "grad_norm": 2.8454601764678955, + "learning_rate": 1.8538256566425584e-06, + "loss": 1.3132, + "step": 1961 + }, + { + "epoch": 2.015408320493066, + "grad_norm": 2.715981960296631, + "learning_rate": 1.8519223448800915e-06, + "loss": 1.3503, + "step": 1962 + }, + { + "epoch": 2.0164355418592708, + "grad_norm": 2.688493251800537, + "learning_rate": 1.850019033117625e-06, + "loss": 1.2966, + "step": 1963 + }, + { + "epoch": 2.017462763225475, + "grad_norm": 2.6956987380981445, + "learning_rate": 1.848115721355158e-06, + "loss": 1.2969, + "step": 1964 + }, + { + "epoch": 2.0184899845916795, + "grad_norm": 2.697744607925415, + "learning_rate": 1.8462124095926914e-06, + "loss": 1.3207, + "step": 1965 + }, + { + "epoch": 2.019517205957884, + "grad_norm": 2.7628865242004395, + "learning_rate": 1.8443090978302248e-06, + "loss": 1.3311, + "step": 1966 + }, + { + "epoch": 2.0205444273240882, + "grad_norm": 2.801499366760254, + "learning_rate": 1.842405786067758e-06, + "loss": 1.3411, + "step": 1967 + }, + { + "epoch": 2.021571648690293, + "grad_norm": 2.6009418964385986, + "learning_rate": 1.8405024743052913e-06, + "loss": 1.2985, + "step": 1968 + }, + { + "epoch": 2.022598870056497, + "grad_norm": 2.772385358810425, + "learning_rate": 1.8385991625428247e-06, + "loss": 1.3407, + "step": 1969 + }, + { + "epoch": 2.0236260914227016, + "grad_norm": 2.558302164077759, + "learning_rate": 1.8366958507803578e-06, + "loss": 1.2814, + "step": 1970 + }, + { + "epoch": 2.0236260914227016, + "eval_loss": 1.7421122789382935, + "eval_runtime": 37.9255, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 1970 + }, + { + "epoch": 2.024653312788906, + "grad_norm": 2.813371181488037, + "learning_rate": 1.8347925390178914e-06, + "loss": 1.3309, + "step": 1971 + }, + { + "epoch": 2.0256805341551103, + "grad_norm": 2.742785692214966, + "learning_rate": 1.8328892272554247e-06, + "loss": 1.2711, + "step": 1972 + }, + { + "epoch": 2.026707755521315, + "grad_norm": 2.889450788497925, + "learning_rate": 1.8309859154929579e-06, + "loss": 1.3309, + "step": 1973 + }, + { + "epoch": 2.0277349768875195, + "grad_norm": 2.919278860092163, + "learning_rate": 1.8290826037304913e-06, + "loss": 1.3168, + "step": 1974 + }, + { + "epoch": 2.0287621982537236, + "grad_norm": 2.6912894248962402, + "learning_rate": 1.8271792919680246e-06, + "loss": 1.29, + "step": 1975 + }, + { + "epoch": 2.029789419619928, + "grad_norm": 2.584214687347412, + "learning_rate": 1.8252759802055578e-06, + "loss": 1.2873, + "step": 1976 + }, + { + "epoch": 2.0308166409861323, + "grad_norm": 2.6445112228393555, + "learning_rate": 1.8233726684430911e-06, + "loss": 1.3335, + "step": 1977 + }, + { + "epoch": 2.031843862352337, + "grad_norm": 2.6441738605499268, + "learning_rate": 1.8214693566806243e-06, + "loss": 1.2795, + "step": 1978 + }, + { + "epoch": 2.0328710837185415, + "grad_norm": 2.7365334033966064, + "learning_rate": 1.8195660449181576e-06, + "loss": 1.3142, + "step": 1979 + }, + { + "epoch": 2.0338983050847457, + "grad_norm": 2.7914040088653564, + "learning_rate": 1.817662733155691e-06, + "loss": 1.3175, + "step": 1980 + }, + { + "epoch": 2.0338983050847457, + "eval_loss": 1.7680878639221191, + "eval_runtime": 37.9012, + "eval_samples_per_second": 26.384, + "eval_steps_per_second": 1.108, + "step": 1980 + }, + { + "epoch": 2.0349255264509503, + "grad_norm": 2.6883442401885986, + "learning_rate": 1.8157594213932244e-06, + "loss": 1.3157, + "step": 1981 + }, + { + "epoch": 2.0359527478171544, + "grad_norm": 2.7459399700164795, + "learning_rate": 1.8138561096307577e-06, + "loss": 1.33, + "step": 1982 + }, + { + "epoch": 2.036979969183359, + "grad_norm": 2.620150327682495, + "learning_rate": 1.811952797868291e-06, + "loss": 1.2664, + "step": 1983 + }, + { + "epoch": 2.0380071905495636, + "grad_norm": 2.6702964305877686, + "learning_rate": 1.8100494861058242e-06, + "loss": 1.3174, + "step": 1984 + }, + { + "epoch": 2.0390344119157677, + "grad_norm": 2.6688029766082764, + "learning_rate": 1.8081461743433576e-06, + "loss": 1.3224, + "step": 1985 + }, + { + "epoch": 2.0400616332819723, + "grad_norm": 2.7634220123291016, + "learning_rate": 1.806242862580891e-06, + "loss": 1.2804, + "step": 1986 + }, + { + "epoch": 2.0410888546481765, + "grad_norm": 2.8583552837371826, + "learning_rate": 1.8043395508184241e-06, + "loss": 1.3242, + "step": 1987 + }, + { + "epoch": 2.042116076014381, + "grad_norm": 2.776536703109741, + "learning_rate": 1.8024362390559575e-06, + "loss": 1.3101, + "step": 1988 + }, + { + "epoch": 2.0431432973805856, + "grad_norm": 2.8194644451141357, + "learning_rate": 1.8005329272934906e-06, + "loss": 1.2875, + "step": 1989 + }, + { + "epoch": 2.04417051874679, + "grad_norm": 3.063290596008301, + "learning_rate": 1.798629615531024e-06, + "loss": 1.365, + "step": 1990 + }, + { + "epoch": 2.04417051874679, + "eval_loss": 1.7815766334533691, + "eval_runtime": 37.9473, + "eval_samples_per_second": 26.352, + "eval_steps_per_second": 1.107, + "step": 1990 + }, + { + "epoch": 2.0451977401129944, + "grad_norm": 2.8954226970672607, + "learning_rate": 1.7967263037685576e-06, + "loss": 1.3109, + "step": 1991 + }, + { + "epoch": 2.046224961479199, + "grad_norm": 2.9319002628326416, + "learning_rate": 1.794822992006091e-06, + "loss": 1.2983, + "step": 1992 + }, + { + "epoch": 2.047252182845403, + "grad_norm": 2.920145034790039, + "learning_rate": 1.792919680243624e-06, + "loss": 1.3571, + "step": 1993 + }, + { + "epoch": 2.0482794042116077, + "grad_norm": 2.6473276615142822, + "learning_rate": 1.7910163684811574e-06, + "loss": 1.2752, + "step": 1994 + }, + { + "epoch": 2.049306625577812, + "grad_norm": 2.6453239917755127, + "learning_rate": 1.7891130567186906e-06, + "loss": 1.311, + "step": 1995 + }, + { + "epoch": 2.0503338469440164, + "grad_norm": 2.6123287677764893, + "learning_rate": 1.787209744956224e-06, + "loss": 1.2817, + "step": 1996 + }, + { + "epoch": 2.051361068310221, + "grad_norm": 2.6542904376983643, + "learning_rate": 1.7853064331937573e-06, + "loss": 1.3067, + "step": 1997 + }, + { + "epoch": 2.052388289676425, + "grad_norm": 2.799492835998535, + "learning_rate": 1.7834031214312905e-06, + "loss": 1.3436, + "step": 1998 + }, + { + "epoch": 2.0534155110426298, + "grad_norm": 2.6220905780792236, + "learning_rate": 1.7814998096688238e-06, + "loss": 1.2819, + "step": 1999 + }, + { + "epoch": 2.054442732408834, + "grad_norm": 2.691195487976074, + "learning_rate": 1.779596497906357e-06, + "loss": 1.3099, + "step": 2000 + }, + { + "epoch": 2.054442732408834, + "eval_loss": 1.770631194114685, + "eval_runtime": 37.9652, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 2000 + }, + { + "epoch": 2.0554699537750385, + "grad_norm": 2.8053042888641357, + "learning_rate": 1.7776931861438906e-06, + "loss": 1.3599, + "step": 2001 + }, + { + "epoch": 2.056497175141243, + "grad_norm": 2.784608840942383, + "learning_rate": 1.775789874381424e-06, + "loss": 1.3527, + "step": 2002 + }, + { + "epoch": 2.0575243965074472, + "grad_norm": 2.8933472633361816, + "learning_rate": 1.7738865626189573e-06, + "loss": 1.3363, + "step": 2003 + }, + { + "epoch": 2.058551617873652, + "grad_norm": 2.691600799560547, + "learning_rate": 1.7719832508564904e-06, + "loss": 1.2678, + "step": 2004 + }, + { + "epoch": 2.0595788392398564, + "grad_norm": 2.627678394317627, + "learning_rate": 1.7700799390940238e-06, + "loss": 1.3002, + "step": 2005 + }, + { + "epoch": 2.0606060606060606, + "grad_norm": 2.724674940109253, + "learning_rate": 1.768176627331557e-06, + "loss": 1.3237, + "step": 2006 + }, + { + "epoch": 2.061633281972265, + "grad_norm": 2.59698224067688, + "learning_rate": 1.7662733155690903e-06, + "loss": 1.3165, + "step": 2007 + }, + { + "epoch": 2.0626605033384693, + "grad_norm": 2.7686543464660645, + "learning_rate": 1.7643700038066237e-06, + "loss": 1.3478, + "step": 2008 + }, + { + "epoch": 2.063687724704674, + "grad_norm": 2.6561412811279297, + "learning_rate": 1.7624666920441568e-06, + "loss": 1.2851, + "step": 2009 + }, + { + "epoch": 2.0647149460708785, + "grad_norm": 2.722731351852417, + "learning_rate": 1.7605633802816902e-06, + "loss": 1.2726, + "step": 2010 + }, + { + "epoch": 2.0647149460708785, + "eval_loss": 1.7728410959243774, + "eval_runtime": 37.6541, + "eval_samples_per_second": 26.558, + "eval_steps_per_second": 1.115, + "step": 2010 + }, + { + "epoch": 2.0657421674370826, + "grad_norm": 2.910609483718872, + "learning_rate": 1.7586600685192238e-06, + "loss": 1.3603, + "step": 2011 + }, + { + "epoch": 2.066769388803287, + "grad_norm": 2.9910264015197754, + "learning_rate": 1.756756756756757e-06, + "loss": 1.3705, + "step": 2012 + }, + { + "epoch": 2.0677966101694913, + "grad_norm": 2.6477038860321045, + "learning_rate": 1.7548534449942903e-06, + "loss": 1.3372, + "step": 2013 + }, + { + "epoch": 2.068823831535696, + "grad_norm": 2.667626142501831, + "learning_rate": 1.7529501332318236e-06, + "loss": 1.3068, + "step": 2014 + }, + { + "epoch": 2.0698510529019005, + "grad_norm": 2.7895913124084473, + "learning_rate": 1.7510468214693568e-06, + "loss": 1.2813, + "step": 2015 + }, + { + "epoch": 2.0708782742681047, + "grad_norm": 2.5626792907714844, + "learning_rate": 1.7491435097068902e-06, + "loss": 1.2852, + "step": 2016 + }, + { + "epoch": 2.0719054956343093, + "grad_norm": 2.71805739402771, + "learning_rate": 1.7472401979444233e-06, + "loss": 1.3203, + "step": 2017 + }, + { + "epoch": 2.0729327170005134, + "grad_norm": 2.6200578212738037, + "learning_rate": 1.7453368861819567e-06, + "loss": 1.2617, + "step": 2018 + }, + { + "epoch": 2.073959938366718, + "grad_norm": 2.7212417125701904, + "learning_rate": 1.74343357441949e-06, + "loss": 1.3082, + "step": 2019 + }, + { + "epoch": 2.0749871597329226, + "grad_norm": 2.983325958251953, + "learning_rate": 1.7415302626570232e-06, + "loss": 1.3699, + "step": 2020 + }, + { + "epoch": 2.0749871597329226, + "eval_loss": 1.756274700164795, + "eval_runtime": 37.7628, + "eval_samples_per_second": 26.481, + "eval_steps_per_second": 1.112, + "step": 2020 + }, + { + "epoch": 2.0760143810991267, + "grad_norm": 2.6345975399017334, + "learning_rate": 1.7396269508945568e-06, + "loss": 1.3024, + "step": 2021 + }, + { + "epoch": 2.0770416024653313, + "grad_norm": 2.670807123184204, + "learning_rate": 1.7377236391320901e-06, + "loss": 1.2737, + "step": 2022 + }, + { + "epoch": 2.078068823831536, + "grad_norm": 2.788457155227661, + "learning_rate": 1.7358203273696233e-06, + "loss": 1.3459, + "step": 2023 + }, + { + "epoch": 2.07909604519774, + "grad_norm": 2.740338087081909, + "learning_rate": 1.7339170156071566e-06, + "loss": 1.3019, + "step": 2024 + }, + { + "epoch": 2.0801232665639446, + "grad_norm": 2.605855941772461, + "learning_rate": 1.73201370384469e-06, + "loss": 1.3096, + "step": 2025 + }, + { + "epoch": 2.081150487930149, + "grad_norm": 2.640096664428711, + "learning_rate": 1.7301103920822231e-06, + "loss": 1.3168, + "step": 2026 + }, + { + "epoch": 2.0821777092963534, + "grad_norm": 2.7602169513702393, + "learning_rate": 1.7282070803197565e-06, + "loss": 1.3121, + "step": 2027 + }, + { + "epoch": 2.083204930662558, + "grad_norm": 2.658534526824951, + "learning_rate": 1.7263037685572897e-06, + "loss": 1.3071, + "step": 2028 + }, + { + "epoch": 2.084232152028762, + "grad_norm": 2.7146220207214355, + "learning_rate": 1.724400456794823e-06, + "loss": 1.3336, + "step": 2029 + }, + { + "epoch": 2.0852593733949667, + "grad_norm": 2.694387197494507, + "learning_rate": 1.7224971450323564e-06, + "loss": 1.3013, + "step": 2030 + }, + { + "epoch": 2.0852593733949667, + "eval_loss": 1.7371397018432617, + "eval_runtime": 37.848, + "eval_samples_per_second": 26.421, + "eval_steps_per_second": 1.11, + "step": 2030 + }, + { + "epoch": 2.086286594761171, + "grad_norm": 2.7298598289489746, + "learning_rate": 1.72059383326989e-06, + "loss": 1.3048, + "step": 2031 + }, + { + "epoch": 2.0873138161273754, + "grad_norm": 2.9527862071990967, + "learning_rate": 1.7186905215074231e-06, + "loss": 1.3561, + "step": 2032 + }, + { + "epoch": 2.08834103749358, + "grad_norm": 2.7970757484436035, + "learning_rate": 1.7167872097449565e-06, + "loss": 1.2897, + "step": 2033 + }, + { + "epoch": 2.089368258859784, + "grad_norm": 2.7579336166381836, + "learning_rate": 1.7148838979824896e-06, + "loss": 1.3092, + "step": 2034 + }, + { + "epoch": 2.0903954802259888, + "grad_norm": 2.8146934509277344, + "learning_rate": 1.712980586220023e-06, + "loss": 1.3056, + "step": 2035 + }, + { + "epoch": 2.091422701592193, + "grad_norm": 2.812089443206787, + "learning_rate": 1.7110772744575563e-06, + "loss": 1.2924, + "step": 2036 + }, + { + "epoch": 2.0924499229583975, + "grad_norm": 2.7205023765563965, + "learning_rate": 1.7091739626950895e-06, + "loss": 1.3058, + "step": 2037 + }, + { + "epoch": 2.093477144324602, + "grad_norm": 2.692681312561035, + "learning_rate": 1.7072706509326229e-06, + "loss": 1.3336, + "step": 2038 + }, + { + "epoch": 2.0945043656908062, + "grad_norm": 2.8189735412597656, + "learning_rate": 1.7053673391701562e-06, + "loss": 1.3171, + "step": 2039 + }, + { + "epoch": 2.095531587057011, + "grad_norm": 2.664710760116577, + "learning_rate": 1.7034640274076894e-06, + "loss": 1.285, + "step": 2040 + }, + { + "epoch": 2.095531587057011, + "eval_loss": 1.7483630180358887, + "eval_runtime": 37.9016, + "eval_samples_per_second": 26.384, + "eval_steps_per_second": 1.108, + "step": 2040 + }, + { + "epoch": 2.0965588084232154, + "grad_norm": 2.664710760116577, + "learning_rate": 1.7034640274076894e-06, + "loss": 1.35, + "step": 2041 + }, + { + "epoch": 2.0975860297894195, + "grad_norm": 2.6864893436431885, + "learning_rate": 1.701560715645223e-06, + "loss": 1.2723, + "step": 2042 + }, + { + "epoch": 2.098613251155624, + "grad_norm": 2.751150608062744, + "learning_rate": 1.6996574038827563e-06, + "loss": 1.3154, + "step": 2043 + }, + { + "epoch": 2.0996404725218283, + "grad_norm": 2.7600581645965576, + "learning_rate": 1.6977540921202895e-06, + "loss": 1.3043, + "step": 2044 + }, + { + "epoch": 2.100667693888033, + "grad_norm": 2.7397212982177734, + "learning_rate": 1.6958507803578228e-06, + "loss": 1.3183, + "step": 2045 + }, + { + "epoch": 2.1016949152542375, + "grad_norm": 2.7808914184570312, + "learning_rate": 1.693947468595356e-06, + "loss": 1.3367, + "step": 2046 + }, + { + "epoch": 2.1027221366204416, + "grad_norm": 2.534454584121704, + "learning_rate": 1.6920441568328893e-06, + "loss": 1.2638, + "step": 2047 + }, + { + "epoch": 2.103749357986646, + "grad_norm": 2.6857175827026367, + "learning_rate": 1.6901408450704227e-06, + "loss": 1.2955, + "step": 2048 + }, + { + "epoch": 2.1047765793528503, + "grad_norm": 2.689323663711548, + "learning_rate": 1.6882375333079559e-06, + "loss": 1.3482, + "step": 2049 + }, + { + "epoch": 2.105803800719055, + "grad_norm": 2.7330009937286377, + "learning_rate": 1.6863342215454892e-06, + "loss": 1.3405, + "step": 2050 + }, + { + "epoch": 2.105803800719055, + "eval_loss": 1.757278323173523, + "eval_runtime": 37.9263, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 2050 + }, + { + "epoch": 2.1068310220852595, + "grad_norm": 2.847092628479004, + "learning_rate": 1.6844309097830226e-06, + "loss": 1.3338, + "step": 2051 + }, + { + "epoch": 2.1078582434514637, + "grad_norm": 2.8156185150146484, + "learning_rate": 1.682527598020556e-06, + "loss": 1.3446, + "step": 2052 + }, + { + "epoch": 2.1088854648176683, + "grad_norm": 2.665828227996826, + "learning_rate": 1.6806242862580893e-06, + "loss": 1.2962, + "step": 2053 + }, + { + "epoch": 2.109912686183873, + "grad_norm": 2.6218175888061523, + "learning_rate": 1.6787209744956227e-06, + "loss": 1.2798, + "step": 2054 + }, + { + "epoch": 2.110939907550077, + "grad_norm": 2.622291326522827, + "learning_rate": 1.6768176627331558e-06, + "loss": 1.3243, + "step": 2055 + }, + { + "epoch": 2.1119671289162816, + "grad_norm": 2.7171177864074707, + "learning_rate": 1.6749143509706892e-06, + "loss": 1.2919, + "step": 2056 + }, + { + "epoch": 2.1129943502824857, + "grad_norm": 2.655845880508423, + "learning_rate": 1.6730110392082225e-06, + "loss": 1.3152, + "step": 2057 + }, + { + "epoch": 2.1140215716486903, + "grad_norm": 2.7783730030059814, + "learning_rate": 1.6711077274457557e-06, + "loss": 1.2694, + "step": 2058 + }, + { + "epoch": 2.115048793014895, + "grad_norm": 2.819627046585083, + "learning_rate": 1.669204415683289e-06, + "loss": 1.3335, + "step": 2059 + }, + { + "epoch": 2.116076014381099, + "grad_norm": 2.817685127258301, + "learning_rate": 1.6673011039208222e-06, + "loss": 1.3172, + "step": 2060 + }, + { + "epoch": 2.116076014381099, + "eval_loss": 1.7683359384536743, + "eval_runtime": 37.8941, + "eval_samples_per_second": 26.389, + "eval_steps_per_second": 1.108, + "step": 2060 + }, + { + "epoch": 2.1171032357473036, + "grad_norm": 2.9053642749786377, + "learning_rate": 1.6653977921583556e-06, + "loss": 1.3356, + "step": 2061 + }, + { + "epoch": 2.118130457113508, + "grad_norm": 2.6939432621002197, + "learning_rate": 1.663494480395889e-06, + "loss": 1.2915, + "step": 2062 + }, + { + "epoch": 2.1191576784797124, + "grad_norm": 2.8870749473571777, + "learning_rate": 1.6615911686334225e-06, + "loss": 1.3188, + "step": 2063 + }, + { + "epoch": 2.120184899845917, + "grad_norm": 2.6949965953826904, + "learning_rate": 1.6596878568709557e-06, + "loss": 1.2984, + "step": 2064 + }, + { + "epoch": 2.121212121212121, + "grad_norm": 2.6951143741607666, + "learning_rate": 1.657784545108489e-06, + "loss": 1.3328, + "step": 2065 + }, + { + "epoch": 2.1222393425783257, + "grad_norm": 2.6718437671661377, + "learning_rate": 1.6558812333460222e-06, + "loss": 1.3327, + "step": 2066 + }, + { + "epoch": 2.12326656394453, + "grad_norm": 2.7002549171447754, + "learning_rate": 1.6539779215835555e-06, + "loss": 1.2941, + "step": 2067 + }, + { + "epoch": 2.1242937853107344, + "grad_norm": 2.695836305618286, + "learning_rate": 1.652074609821089e-06, + "loss": 1.301, + "step": 2068 + }, + { + "epoch": 2.125321006676939, + "grad_norm": 2.7710680961608887, + "learning_rate": 1.650171298058622e-06, + "loss": 1.3084, + "step": 2069 + }, + { + "epoch": 2.126348228043143, + "grad_norm": 2.7416014671325684, + "learning_rate": 1.6482679862961554e-06, + "loss": 1.2984, + "step": 2070 + }, + { + "epoch": 2.126348228043143, + "eval_loss": 1.7566925287246704, + "eval_runtime": 37.9783, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 2070 + }, + { + "epoch": 2.1273754494093478, + "grad_norm": 2.570422887802124, + "learning_rate": 1.6463646745336886e-06, + "loss": 1.2817, + "step": 2071 + }, + { + "epoch": 2.128402670775552, + "grad_norm": 2.658071517944336, + "learning_rate": 1.644461362771222e-06, + "loss": 1.308, + "step": 2072 + }, + { + "epoch": 2.1294298921417565, + "grad_norm": 2.65232515335083, + "learning_rate": 1.6425580510087555e-06, + "loss": 1.2854, + "step": 2073 + }, + { + "epoch": 2.130457113507961, + "grad_norm": 2.6072998046875, + "learning_rate": 1.6406547392462889e-06, + "loss": 1.2836, + "step": 2074 + }, + { + "epoch": 2.131484334874165, + "grad_norm": 2.6997647285461426, + "learning_rate": 1.638751427483822e-06, + "loss": 1.3258, + "step": 2075 + }, + { + "epoch": 2.13251155624037, + "grad_norm": 2.5479812622070312, + "learning_rate": 1.6368481157213554e-06, + "loss": 1.2987, + "step": 2076 + }, + { + "epoch": 2.1335387776065744, + "grad_norm": 2.801560878753662, + "learning_rate": 1.6349448039588885e-06, + "loss": 1.3542, + "step": 2077 + }, + { + "epoch": 2.1345659989727785, + "grad_norm": 2.6710598468780518, + "learning_rate": 1.6330414921964219e-06, + "loss": 1.2832, + "step": 2078 + }, + { + "epoch": 2.135593220338983, + "grad_norm": 2.8505167961120605, + "learning_rate": 1.6311381804339553e-06, + "loss": 1.327, + "step": 2079 + }, + { + "epoch": 2.1366204417051873, + "grad_norm": 2.6584553718566895, + "learning_rate": 1.6292348686714884e-06, + "loss": 1.3003, + "step": 2080 + }, + { + "epoch": 2.1366204417051873, + "eval_loss": 1.7610517740249634, + "eval_runtime": 37.9139, + "eval_samples_per_second": 26.376, + "eval_steps_per_second": 1.108, + "step": 2080 + }, + { + "epoch": 2.137647663071392, + "grad_norm": 2.7930376529693604, + "learning_rate": 1.6273315569090218e-06, + "loss": 1.3759, + "step": 2081 + }, + { + "epoch": 2.1386748844375965, + "grad_norm": 2.6514129638671875, + "learning_rate": 1.625428245146555e-06, + "loss": 1.3512, + "step": 2082 + }, + { + "epoch": 2.1397021058038006, + "grad_norm": 2.626838445663452, + "learning_rate": 1.6235249333840885e-06, + "loss": 1.3074, + "step": 2083 + }, + { + "epoch": 2.140729327170005, + "grad_norm": 2.741288661956787, + "learning_rate": 1.6216216216216219e-06, + "loss": 1.3327, + "step": 2084 + }, + { + "epoch": 2.1417565485362093, + "grad_norm": 2.7371652126312256, + "learning_rate": 1.6197183098591552e-06, + "loss": 1.3193, + "step": 2085 + }, + { + "epoch": 2.142783769902414, + "grad_norm": 2.631516218185425, + "learning_rate": 1.6178149980966884e-06, + "loss": 1.2899, + "step": 2086 + }, + { + "epoch": 2.1438109912686185, + "grad_norm": 2.7853667736053467, + "learning_rate": 1.6159116863342217e-06, + "loss": 1.356, + "step": 2087 + }, + { + "epoch": 2.1448382126348227, + "grad_norm": 2.598886489868164, + "learning_rate": 1.6140083745717549e-06, + "loss": 1.2885, + "step": 2088 + }, + { + "epoch": 2.1458654340010272, + "grad_norm": 2.700669288635254, + "learning_rate": 1.6121050628092882e-06, + "loss": 1.3471, + "step": 2089 + }, + { + "epoch": 2.146892655367232, + "grad_norm": 2.6174018383026123, + "learning_rate": 1.6102017510468216e-06, + "loss": 1.2635, + "step": 2090 + }, + { + "epoch": 2.146892655367232, + "eval_loss": 1.764804482460022, + "eval_runtime": 37.8935, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 2090 + }, + { + "epoch": 2.147919876733436, + "grad_norm": 2.8135688304901123, + "learning_rate": 1.6082984392843548e-06, + "loss": 1.3178, + "step": 2091 + }, + { + "epoch": 2.1489470980996406, + "grad_norm": 2.8416011333465576, + "learning_rate": 1.6063951275218881e-06, + "loss": 1.3652, + "step": 2092 + }, + { + "epoch": 2.1499743194658447, + "grad_norm": 2.780001640319824, + "learning_rate": 1.6044918157594217e-06, + "loss": 1.2921, + "step": 2093 + }, + { + "epoch": 2.1510015408320493, + "grad_norm": 2.634460926055908, + "learning_rate": 1.6025885039969548e-06, + "loss": 1.3224, + "step": 2094 + }, + { + "epoch": 2.152028762198254, + "grad_norm": 2.6783623695373535, + "learning_rate": 1.6006851922344882e-06, + "loss": 1.3297, + "step": 2095 + }, + { + "epoch": 2.153055983564458, + "grad_norm": 2.7781169414520264, + "learning_rate": 1.5987818804720216e-06, + "loss": 1.3324, + "step": 2096 + }, + { + "epoch": 2.1540832049306626, + "grad_norm": 2.587174415588379, + "learning_rate": 1.5968785687095547e-06, + "loss": 1.2801, + "step": 2097 + }, + { + "epoch": 2.1551104262968668, + "grad_norm": 2.654083728790283, + "learning_rate": 1.594975256947088e-06, + "loss": 1.3323, + "step": 2098 + }, + { + "epoch": 2.1561376476630714, + "grad_norm": 2.6695969104766846, + "learning_rate": 1.5930719451846212e-06, + "loss": 1.289, + "step": 2099 + }, + { + "epoch": 2.157164869029276, + "grad_norm": 2.7060039043426514, + "learning_rate": 1.5911686334221546e-06, + "loss": 1.3425, + "step": 2100 + }, + { + "epoch": 2.157164869029276, + "eval_loss": 1.7727104425430298, + "eval_runtime": 37.9001, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 2100 + }, + { + "epoch": 2.15819209039548, + "grad_norm": 2.536756753921509, + "learning_rate": 1.589265321659688e-06, + "loss": 1.2538, + "step": 2101 + }, + { + "epoch": 2.1592193117616847, + "grad_norm": 2.655916213989258, + "learning_rate": 1.5873620098972211e-06, + "loss": 1.2925, + "step": 2102 + }, + { + "epoch": 2.1602465331278893, + "grad_norm": 2.686228036880493, + "learning_rate": 1.5854586981347547e-06, + "loss": 1.3386, + "step": 2103 + }, + { + "epoch": 2.1612737544940934, + "grad_norm": 2.794640302658081, + "learning_rate": 1.583555386372288e-06, + "loss": 1.348, + "step": 2104 + }, + { + "epoch": 2.162300975860298, + "grad_norm": 2.6761186122894287, + "learning_rate": 1.5816520746098212e-06, + "loss": 1.3029, + "step": 2105 + }, + { + "epoch": 2.163328197226502, + "grad_norm": 2.5585744380950928, + "learning_rate": 1.5797487628473546e-06, + "loss": 1.2802, + "step": 2106 + }, + { + "epoch": 2.1643554185927067, + "grad_norm": 2.6511101722717285, + "learning_rate": 1.577845451084888e-06, + "loss": 1.3048, + "step": 2107 + }, + { + "epoch": 2.1653826399589113, + "grad_norm": 2.601940870285034, + "learning_rate": 1.575942139322421e-06, + "loss": 1.2433, + "step": 2108 + }, + { + "epoch": 2.1664098613251155, + "grad_norm": 2.676440954208374, + "learning_rate": 1.5740388275599544e-06, + "loss": 1.3296, + "step": 2109 + }, + { + "epoch": 2.16743708269132, + "grad_norm": 2.7463326454162598, + "learning_rate": 1.5721355157974876e-06, + "loss": 1.2931, + "step": 2110 + }, + { + "epoch": 2.16743708269132, + "eval_loss": 1.7691689729690552, + "eval_runtime": 37.6043, + "eval_samples_per_second": 26.593, + "eval_steps_per_second": 1.117, + "step": 2110 + }, + { + "epoch": 2.168464304057524, + "grad_norm": 2.895216464996338, + "learning_rate": 1.570232204035021e-06, + "loss": 1.3091, + "step": 2111 + }, + { + "epoch": 2.169491525423729, + "grad_norm": 2.765704870223999, + "learning_rate": 1.5683288922725543e-06, + "loss": 1.3197, + "step": 2112 + }, + { + "epoch": 2.1705187467899334, + "grad_norm": 2.750681161880493, + "learning_rate": 1.5664255805100879e-06, + "loss": 1.3016, + "step": 2113 + }, + { + "epoch": 2.1715459681561375, + "grad_norm": 2.6807963848114014, + "learning_rate": 1.564522268747621e-06, + "loss": 1.3176, + "step": 2114 + }, + { + "epoch": 2.172573189522342, + "grad_norm": 2.618666410446167, + "learning_rate": 1.5626189569851544e-06, + "loss": 1.2391, + "step": 2115 + }, + { + "epoch": 2.1736004108885463, + "grad_norm": 2.745319128036499, + "learning_rate": 1.5607156452226875e-06, + "loss": 1.3264, + "step": 2116 + }, + { + "epoch": 2.174627632254751, + "grad_norm": 2.787904977798462, + "learning_rate": 1.558812333460221e-06, + "loss": 1.319, + "step": 2117 + }, + { + "epoch": 2.1756548536209555, + "grad_norm": 2.6901752948760986, + "learning_rate": 1.5569090216977543e-06, + "loss": 1.3216, + "step": 2118 + }, + { + "epoch": 2.1766820749871596, + "grad_norm": 2.723172426223755, + "learning_rate": 1.5550057099352874e-06, + "loss": 1.3241, + "step": 2119 + }, + { + "epoch": 2.177709296353364, + "grad_norm": 2.6102678775787354, + "learning_rate": 1.5531023981728208e-06, + "loss": 1.2892, + "step": 2120 + }, + { + "epoch": 2.177709296353364, + "eval_loss": 1.7888449430465698, + "eval_runtime": 37.7503, + "eval_samples_per_second": 26.49, + "eval_steps_per_second": 1.113, + "step": 2120 + }, + { + "epoch": 2.1787365177195683, + "grad_norm": 2.653026819229126, + "learning_rate": 1.5511990864103542e-06, + "loss": 1.322, + "step": 2121 + }, + { + "epoch": 2.179763739085773, + "grad_norm": 2.806363344192505, + "learning_rate": 1.5492957746478873e-06, + "loss": 1.3412, + "step": 2122 + }, + { + "epoch": 2.1807909604519775, + "grad_norm": 2.682968854904175, + "learning_rate": 1.5473924628854209e-06, + "loss": 1.2953, + "step": 2123 + }, + { + "epoch": 2.1818181818181817, + "grad_norm": 2.7165629863739014, + "learning_rate": 1.5454891511229542e-06, + "loss": 1.312, + "step": 2124 + }, + { + "epoch": 2.1828454031843862, + "grad_norm": 2.8055427074432373, + "learning_rate": 1.5435858393604874e-06, + "loss": 1.3318, + "step": 2125 + }, + { + "epoch": 2.183872624550591, + "grad_norm": 2.731459617614746, + "learning_rate": 1.5416825275980208e-06, + "loss": 1.3059, + "step": 2126 + }, + { + "epoch": 2.184899845916795, + "grad_norm": 2.6671993732452393, + "learning_rate": 1.539779215835554e-06, + "loss": 1.2945, + "step": 2127 + }, + { + "epoch": 2.1859270672829996, + "grad_norm": 2.8046534061431885, + "learning_rate": 1.5378759040730873e-06, + "loss": 1.3437, + "step": 2128 + }, + { + "epoch": 2.1869542886492037, + "grad_norm": 2.717656135559082, + "learning_rate": 1.5359725923106206e-06, + "loss": 1.3221, + "step": 2129 + }, + { + "epoch": 2.1879815100154083, + "grad_norm": 2.746224880218506, + "learning_rate": 1.5340692805481538e-06, + "loss": 1.2857, + "step": 2130 + }, + { + "epoch": 2.1879815100154083, + "eval_loss": 1.777937889099121, + "eval_runtime": 37.8481, + "eval_samples_per_second": 26.421, + "eval_steps_per_second": 1.11, + "step": 2130 + }, + { + "epoch": 2.189008731381613, + "grad_norm": 2.6477184295654297, + "learning_rate": 1.5321659687856871e-06, + "loss": 1.2811, + "step": 2131 + }, + { + "epoch": 2.190035952747817, + "grad_norm": 2.664757251739502, + "learning_rate": 1.5302626570232205e-06, + "loss": 1.2542, + "step": 2132 + }, + { + "epoch": 2.1910631741140216, + "grad_norm": 2.8133883476257324, + "learning_rate": 1.5283593452607539e-06, + "loss": 1.3485, + "step": 2133 + }, + { + "epoch": 2.1920903954802258, + "grad_norm": 2.723867893218994, + "learning_rate": 1.5264560334982872e-06, + "loss": 1.2974, + "step": 2134 + }, + { + "epoch": 2.1931176168464304, + "grad_norm": 2.776700496673584, + "learning_rate": 1.5245527217358206e-06, + "loss": 1.3219, + "step": 2135 + }, + { + "epoch": 2.194144838212635, + "grad_norm": 2.547466993331909, + "learning_rate": 1.5226494099733537e-06, + "loss": 1.2923, + "step": 2136 + }, + { + "epoch": 2.195172059578839, + "grad_norm": 2.7371773719787598, + "learning_rate": 1.520746098210887e-06, + "loss": 1.3466, + "step": 2137 + }, + { + "epoch": 2.1961992809450437, + "grad_norm": 2.4609546661376953, + "learning_rate": 1.5188427864484205e-06, + "loss": 1.2865, + "step": 2138 + }, + { + "epoch": 2.1972265023112483, + "grad_norm": 2.6336286067962646, + "learning_rate": 1.5169394746859536e-06, + "loss": 1.3071, + "step": 2139 + }, + { + "epoch": 2.1982537236774524, + "grad_norm": 2.6983938217163086, + "learning_rate": 1.515036162923487e-06, + "loss": 1.3216, + "step": 2140 + }, + { + "epoch": 2.1982537236774524, + "eval_loss": 1.7778681516647339, + "eval_runtime": 37.873, + "eval_samples_per_second": 26.404, + "eval_steps_per_second": 1.109, + "step": 2140 + }, + { + "epoch": 2.199280945043657, + "grad_norm": 2.8055241107940674, + "learning_rate": 1.5131328511610201e-06, + "loss": 1.3517, + "step": 2141 + }, + { + "epoch": 2.200308166409861, + "grad_norm": 2.760460615158081, + "learning_rate": 1.5112295393985535e-06, + "loss": 1.3242, + "step": 2142 + }, + { + "epoch": 2.2013353877760657, + "grad_norm": 2.574289560317993, + "learning_rate": 1.509326227636087e-06, + "loss": 1.2595, + "step": 2143 + }, + { + "epoch": 2.2023626091422703, + "grad_norm": 2.827829360961914, + "learning_rate": 1.5074229158736204e-06, + "loss": 1.3167, + "step": 2144 + }, + { + "epoch": 2.2033898305084745, + "grad_norm": 2.7533328533172607, + "learning_rate": 1.5055196041111536e-06, + "loss": 1.3161, + "step": 2145 + }, + { + "epoch": 2.204417051874679, + "grad_norm": 2.865976095199585, + "learning_rate": 1.503616292348687e-06, + "loss": 1.3444, + "step": 2146 + }, + { + "epoch": 2.205444273240883, + "grad_norm": 2.5228123664855957, + "learning_rate": 1.50171298058622e-06, + "loss": 1.2566, + "step": 2147 + }, + { + "epoch": 2.206471494607088, + "grad_norm": 2.682549238204956, + "learning_rate": 1.4998096688237535e-06, + "loss": 1.3284, + "step": 2148 + }, + { + "epoch": 2.2074987159732924, + "grad_norm": 2.8158624172210693, + "learning_rate": 1.4979063570612868e-06, + "loss": 1.3148, + "step": 2149 + }, + { + "epoch": 2.2085259373394965, + "grad_norm": 2.7644388675689697, + "learning_rate": 1.49600304529882e-06, + "loss": 1.2914, + "step": 2150 + }, + { + "epoch": 2.2085259373394965, + "eval_loss": 1.7767399549484253, + "eval_runtime": 37.8771, + "eval_samples_per_second": 26.401, + "eval_steps_per_second": 1.109, + "step": 2150 + }, + { + "epoch": 2.209553158705701, + "grad_norm": 2.7370104789733887, + "learning_rate": 1.4940997335363533e-06, + "loss": 1.3165, + "step": 2151 + }, + { + "epoch": 2.2105803800719057, + "grad_norm": 2.692795753479004, + "learning_rate": 1.4921964217738865e-06, + "loss": 1.3038, + "step": 2152 + }, + { + "epoch": 2.21160760143811, + "grad_norm": 2.8083877563476562, + "learning_rate": 1.49029311001142e-06, + "loss": 1.3315, + "step": 2153 + }, + { + "epoch": 2.2126348228043144, + "grad_norm": 2.816497325897217, + "learning_rate": 1.4883897982489534e-06, + "loss": 1.3319, + "step": 2154 + }, + { + "epoch": 2.2136620441705186, + "grad_norm": 2.7338058948516846, + "learning_rate": 1.4864864864864868e-06, + "loss": 1.3057, + "step": 2155 + }, + { + "epoch": 2.214689265536723, + "grad_norm": 2.631532669067383, + "learning_rate": 1.48458317472402e-06, + "loss": 1.3123, + "step": 2156 + }, + { + "epoch": 2.2157164869029278, + "grad_norm": 2.511723279953003, + "learning_rate": 1.4826798629615533e-06, + "loss": 1.2714, + "step": 2157 + }, + { + "epoch": 2.216743708269132, + "grad_norm": 2.6720175743103027, + "learning_rate": 1.4807765511990864e-06, + "loss": 1.2806, + "step": 2158 + }, + { + "epoch": 2.2177709296353365, + "grad_norm": 2.6232240200042725, + "learning_rate": 1.4788732394366198e-06, + "loss": 1.3057, + "step": 2159 + }, + { + "epoch": 2.2187981510015407, + "grad_norm": 2.7055530548095703, + "learning_rate": 1.4769699276741532e-06, + "loss": 1.2896, + "step": 2160 + }, + { + "epoch": 2.2187981510015407, + "eval_loss": 1.780350923538208, + "eval_runtime": 37.9284, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 2160 + }, + { + "epoch": 2.2198253723677452, + "grad_norm": 2.7227275371551514, + "learning_rate": 1.4750666159116863e-06, + "loss": 1.2688, + "step": 2161 + }, + { + "epoch": 2.22085259373395, + "grad_norm": 2.741649866104126, + "learning_rate": 1.4731633041492197e-06, + "loss": 1.3318, + "step": 2162 + }, + { + "epoch": 2.221879815100154, + "grad_norm": 2.901906967163086, + "learning_rate": 1.4712599923867528e-06, + "loss": 1.3502, + "step": 2163 + }, + { + "epoch": 2.2229070364663586, + "grad_norm": 2.6306073665618896, + "learning_rate": 1.4693566806242864e-06, + "loss": 1.2753, + "step": 2164 + }, + { + "epoch": 2.2239342578325627, + "grad_norm": 2.806844472885132, + "learning_rate": 1.4674533688618198e-06, + "loss": 1.3006, + "step": 2165 + }, + { + "epoch": 2.2249614791987673, + "grad_norm": 2.754244089126587, + "learning_rate": 1.4655500570993531e-06, + "loss": 1.3036, + "step": 2166 + }, + { + "epoch": 2.225988700564972, + "grad_norm": 2.7528069019317627, + "learning_rate": 1.4636467453368863e-06, + "loss": 1.3551, + "step": 2167 + }, + { + "epoch": 2.227015921931176, + "grad_norm": 2.9147045612335205, + "learning_rate": 1.4617434335744197e-06, + "loss": 1.3444, + "step": 2168 + }, + { + "epoch": 2.2280431432973806, + "grad_norm": 2.7143068313598633, + "learning_rate": 1.4598401218119528e-06, + "loss": 1.3333, + "step": 2169 + }, + { + "epoch": 2.2290703646635848, + "grad_norm": 2.7245426177978516, + "learning_rate": 1.4579368100494862e-06, + "loss": 1.32, + "step": 2170 + }, + { + "epoch": 2.2290703646635848, + "eval_loss": 1.7687398195266724, + "eval_runtime": 37.931, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 2170 + }, + { + "epoch": 2.2300975860297894, + "grad_norm": 2.614847421646118, + "learning_rate": 1.4560334982870195e-06, + "loss": 1.316, + "step": 2171 + }, + { + "epoch": 2.231124807395994, + "grad_norm": 2.6008236408233643, + "learning_rate": 1.4541301865245527e-06, + "loss": 1.3271, + "step": 2172 + }, + { + "epoch": 2.232152028762198, + "grad_norm": 2.6043577194213867, + "learning_rate": 1.452226874762086e-06, + "loss": 1.2838, + "step": 2173 + }, + { + "epoch": 2.2331792501284027, + "grad_norm": 2.800450325012207, + "learning_rate": 1.4503235629996196e-06, + "loss": 1.3542, + "step": 2174 + }, + { + "epoch": 2.2342064714946073, + "grad_norm": 2.6808807849884033, + "learning_rate": 1.4484202512371528e-06, + "loss": 1.2573, + "step": 2175 + }, + { + "epoch": 2.2352336928608114, + "grad_norm": 3.005553722381592, + "learning_rate": 1.4465169394746861e-06, + "loss": 1.3598, + "step": 2176 + }, + { + "epoch": 2.236260914227016, + "grad_norm": 2.8075437545776367, + "learning_rate": 1.4446136277122195e-06, + "loss": 1.2687, + "step": 2177 + }, + { + "epoch": 2.23728813559322, + "grad_norm": 2.828524589538574, + "learning_rate": 1.4427103159497526e-06, + "loss": 1.2906, + "step": 2178 + }, + { + "epoch": 2.2383153569594247, + "grad_norm": 2.7086875438690186, + "learning_rate": 1.440807004187286e-06, + "loss": 1.3199, + "step": 2179 + }, + { + "epoch": 2.2393425783256293, + "grad_norm": 2.8826253414154053, + "learning_rate": 1.4389036924248192e-06, + "loss": 1.3414, + "step": 2180 + }, + { + "epoch": 2.2393425783256293, + "eval_loss": 1.7614880800247192, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 2180 + }, + { + "epoch": 2.2403697996918335, + "grad_norm": 2.8024802207946777, + "learning_rate": 1.4370003806623525e-06, + "loss": 1.3127, + "step": 2181 + }, + { + "epoch": 2.241397021058038, + "grad_norm": 2.7175004482269287, + "learning_rate": 1.4350970688998859e-06, + "loss": 1.3589, + "step": 2182 + }, + { + "epoch": 2.242424242424242, + "grad_norm": 2.7450461387634277, + "learning_rate": 1.433193757137419e-06, + "loss": 1.2934, + "step": 2183 + }, + { + "epoch": 2.243451463790447, + "grad_norm": 2.627926826477051, + "learning_rate": 1.4312904453749526e-06, + "loss": 1.2605, + "step": 2184 + }, + { + "epoch": 2.2444786851566514, + "grad_norm": 2.7466957569122314, + "learning_rate": 1.429387133612486e-06, + "loss": 1.2993, + "step": 2185 + }, + { + "epoch": 2.2455059065228555, + "grad_norm": 2.8719801902770996, + "learning_rate": 1.4274838218500191e-06, + "loss": 1.3251, + "step": 2186 + }, + { + "epoch": 2.24653312788906, + "grad_norm": 3.052253007888794, + "learning_rate": 1.4255805100875525e-06, + "loss": 1.3638, + "step": 2187 + }, + { + "epoch": 2.2475603492552647, + "grad_norm": 2.860090970993042, + "learning_rate": 1.4236771983250858e-06, + "loss": 1.3121, + "step": 2188 + }, + { + "epoch": 2.248587570621469, + "grad_norm": 2.656618356704712, + "learning_rate": 1.421773886562619e-06, + "loss": 1.2857, + "step": 2189 + }, + { + "epoch": 2.2496147919876734, + "grad_norm": 2.6252174377441406, + "learning_rate": 1.4198705748001524e-06, + "loss": 1.3094, + "step": 2190 + }, + { + "epoch": 2.2496147919876734, + "eval_loss": 1.7660510540008545, + "eval_runtime": 37.9085, + "eval_samples_per_second": 26.379, + "eval_steps_per_second": 1.108, + "step": 2190 + }, + { + "epoch": 2.2506420133538776, + "grad_norm": 2.610424757003784, + "learning_rate": 1.4179672630376855e-06, + "loss": 1.2809, + "step": 2191 + }, + { + "epoch": 2.251669234720082, + "grad_norm": 2.5957748889923096, + "learning_rate": 1.4160639512752189e-06, + "loss": 1.3009, + "step": 2192 + }, + { + "epoch": 2.2526964560862868, + "grad_norm": 2.6947720050811768, + "learning_rate": 1.4141606395127522e-06, + "loss": 1.2697, + "step": 2193 + }, + { + "epoch": 2.253723677452491, + "grad_norm": 2.5192782878875732, + "learning_rate": 1.4122573277502858e-06, + "loss": 1.3245, + "step": 2194 + }, + { + "epoch": 2.2547508988186955, + "grad_norm": 2.7121758460998535, + "learning_rate": 1.410354015987819e-06, + "loss": 1.3012, + "step": 2195 + }, + { + "epoch": 2.2557781201848996, + "grad_norm": 2.8090286254882812, + "learning_rate": 1.4084507042253523e-06, + "loss": 1.3411, + "step": 2196 + }, + { + "epoch": 2.2568053415511042, + "grad_norm": 2.6683340072631836, + "learning_rate": 1.4065473924628855e-06, + "loss": 1.3059, + "step": 2197 + }, + { + "epoch": 2.257832562917309, + "grad_norm": 2.782543659210205, + "learning_rate": 1.4046440807004188e-06, + "loss": 1.2978, + "step": 2198 + }, + { + "epoch": 2.258859784283513, + "grad_norm": 2.693094253540039, + "learning_rate": 1.4027407689379522e-06, + "loss": 1.2785, + "step": 2199 + }, + { + "epoch": 2.2598870056497176, + "grad_norm": 2.7780873775482178, + "learning_rate": 1.4008374571754854e-06, + "loss": 1.3178, + "step": 2200 + }, + { + "epoch": 2.2598870056497176, + "eval_loss": 1.7748886346817017, + "eval_runtime": 37.9075, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 2200 + }, + { + "epoch": 2.260914227015922, + "grad_norm": 2.670476198196411, + "learning_rate": 1.3989341454130187e-06, + "loss": 1.3095, + "step": 2201 + }, + { + "epoch": 2.2619414483821263, + "grad_norm": 2.7604143619537354, + "learning_rate": 1.397030833650552e-06, + "loss": 1.3342, + "step": 2202 + }, + { + "epoch": 2.262968669748331, + "grad_norm": 2.7799160480499268, + "learning_rate": 1.3951275218880852e-06, + "loss": 1.3734, + "step": 2203 + }, + { + "epoch": 2.263995891114535, + "grad_norm": 2.5651214122772217, + "learning_rate": 1.3932242101256188e-06, + "loss": 1.2775, + "step": 2204 + }, + { + "epoch": 2.2650231124807396, + "grad_norm": 2.8254587650299072, + "learning_rate": 1.3913208983631522e-06, + "loss": 1.3148, + "step": 2205 + }, + { + "epoch": 2.266050333846944, + "grad_norm": 2.694624662399292, + "learning_rate": 1.3894175866006853e-06, + "loss": 1.2702, + "step": 2206 + }, + { + "epoch": 2.2670775552131484, + "grad_norm": 2.789055824279785, + "learning_rate": 1.3875142748382187e-06, + "loss": 1.2516, + "step": 2207 + }, + { + "epoch": 2.268104776579353, + "grad_norm": 2.848062753677368, + "learning_rate": 1.385610963075752e-06, + "loss": 1.3162, + "step": 2208 + }, + { + "epoch": 2.269131997945557, + "grad_norm": 2.939532518386841, + "learning_rate": 1.3837076513132852e-06, + "loss": 1.3428, + "step": 2209 + }, + { + "epoch": 2.2701592193117617, + "grad_norm": 2.9084253311157227, + "learning_rate": 1.3818043395508186e-06, + "loss": 1.3277, + "step": 2210 + }, + { + "epoch": 2.2701592193117617, + "eval_loss": 1.786512017250061, + "eval_runtime": 37.685, + "eval_samples_per_second": 26.536, + "eval_steps_per_second": 1.115, + "step": 2210 + }, + { + "epoch": 2.2711864406779663, + "grad_norm": 2.6281025409698486, + "learning_rate": 1.3799010277883517e-06, + "loss": 1.3128, + "step": 2211 + }, + { + "epoch": 2.2722136620441704, + "grad_norm": 2.748809337615967, + "learning_rate": 1.377997716025885e-06, + "loss": 1.3114, + "step": 2212 + }, + { + "epoch": 2.273240883410375, + "grad_norm": 2.5523533821105957, + "learning_rate": 1.3760944042634184e-06, + "loss": 1.3401, + "step": 2213 + }, + { + "epoch": 2.2742681047765796, + "grad_norm": 2.6466875076293945, + "learning_rate": 1.3741910925009518e-06, + "loss": 1.3338, + "step": 2214 + }, + { + "epoch": 2.2752953261427837, + "grad_norm": 2.7305641174316406, + "learning_rate": 1.3722877807384852e-06, + "loss": 1.3321, + "step": 2215 + }, + { + "epoch": 2.2763225475089883, + "grad_norm": 2.6950535774230957, + "learning_rate": 1.3703844689760185e-06, + "loss": 1.3053, + "step": 2216 + }, + { + "epoch": 2.2773497688751925, + "grad_norm": 2.5823051929473877, + "learning_rate": 1.3684811572135517e-06, + "loss": 1.2622, + "step": 2217 + }, + { + "epoch": 2.278376990241397, + "grad_norm": 2.891983985900879, + "learning_rate": 1.366577845451085e-06, + "loss": 1.3346, + "step": 2218 + }, + { + "epoch": 2.279404211607601, + "grad_norm": 2.8419189453125, + "learning_rate": 1.3646745336886184e-06, + "loss": 1.3336, + "step": 2219 + }, + { + "epoch": 2.280431432973806, + "grad_norm": 2.632622480392456, + "learning_rate": 1.3627712219261515e-06, + "loss": 1.2895, + "step": 2220 + }, + { + "epoch": 2.280431432973806, + "eval_loss": 1.7908138036727905, + "eval_runtime": 37.8557, + "eval_samples_per_second": 26.416, + "eval_steps_per_second": 1.109, + "step": 2220 + }, + { + "epoch": 2.2814586543400104, + "grad_norm": 2.6359457969665527, + "learning_rate": 1.360867910163685e-06, + "loss": 1.3073, + "step": 2221 + }, + { + "epoch": 2.2824858757062145, + "grad_norm": 2.6809372901916504, + "learning_rate": 1.358964598401218e-06, + "loss": 1.3131, + "step": 2222 + }, + { + "epoch": 2.283513097072419, + "grad_norm": 2.579667806625366, + "learning_rate": 1.3570612866387514e-06, + "loss": 1.2958, + "step": 2223 + }, + { + "epoch": 2.2845403184386237, + "grad_norm": 2.8612430095672607, + "learning_rate": 1.355157974876285e-06, + "loss": 1.3478, + "step": 2224 + }, + { + "epoch": 2.285567539804828, + "grad_norm": 2.728100061416626, + "learning_rate": 1.3532546631138184e-06, + "loss": 1.3238, + "step": 2225 + }, + { + "epoch": 2.2865947611710324, + "grad_norm": 2.665339469909668, + "learning_rate": 1.3513513513513515e-06, + "loss": 1.2984, + "step": 2226 + }, + { + "epoch": 2.2876219825372366, + "grad_norm": 2.677293062210083, + "learning_rate": 1.3494480395888849e-06, + "loss": 1.3052, + "step": 2227 + }, + { + "epoch": 2.288649203903441, + "grad_norm": 2.8480992317199707, + "learning_rate": 1.347544727826418e-06, + "loss": 1.2887, + "step": 2228 + }, + { + "epoch": 2.2896764252696458, + "grad_norm": 2.6627776622772217, + "learning_rate": 1.3456414160639514e-06, + "loss": 1.3243, + "step": 2229 + }, + { + "epoch": 2.29070364663585, + "grad_norm": 2.711862087249756, + "learning_rate": 1.3437381043014847e-06, + "loss": 1.3112, + "step": 2230 + }, + { + "epoch": 2.29070364663585, + "eval_loss": 1.781941533088684, + "eval_runtime": 37.9343, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 2230 + }, + { + "epoch": 2.2917308680020545, + "grad_norm": 2.6429977416992188, + "learning_rate": 1.341834792539018e-06, + "loss": 1.2945, + "step": 2231 + }, + { + "epoch": 2.2927580893682586, + "grad_norm": 2.557021379470825, + "learning_rate": 1.3399314807765513e-06, + "loss": 1.32, + "step": 2232 + }, + { + "epoch": 2.2937853107344632, + "grad_norm": 2.5697803497314453, + "learning_rate": 1.3380281690140844e-06, + "loss": 1.3161, + "step": 2233 + }, + { + "epoch": 2.294812532100668, + "grad_norm": 2.5973010063171387, + "learning_rate": 1.336124857251618e-06, + "loss": 1.2822, + "step": 2234 + }, + { + "epoch": 2.295839753466872, + "grad_norm": 2.622098207473755, + "learning_rate": 1.3342215454891513e-06, + "loss": 1.2853, + "step": 2235 + }, + { + "epoch": 2.2968669748330766, + "grad_norm": 2.75445294380188, + "learning_rate": 1.3323182337266847e-06, + "loss": 1.3009, + "step": 2236 + }, + { + "epoch": 2.297894196199281, + "grad_norm": 2.8000597953796387, + "learning_rate": 1.3304149219642179e-06, + "loss": 1.3547, + "step": 2237 + }, + { + "epoch": 2.2989214175654853, + "grad_norm": 2.7219057083129883, + "learning_rate": 1.3285116102017512e-06, + "loss": 1.3128, + "step": 2238 + }, + { + "epoch": 2.29994863893169, + "grad_norm": 2.5688209533691406, + "learning_rate": 1.3266082984392844e-06, + "loss": 1.3077, + "step": 2239 + }, + { + "epoch": 2.300975860297894, + "grad_norm": 2.629265546798706, + "learning_rate": 1.3247049866768177e-06, + "loss": 1.2929, + "step": 2240 + }, + { + "epoch": 2.300975860297894, + "eval_loss": 1.7657849788665771, + "eval_runtime": 37.9938, + "eval_samples_per_second": 26.32, + "eval_steps_per_second": 1.105, + "step": 2240 + }, + { + "epoch": 2.3020030816640986, + "grad_norm": 2.6893677711486816, + "learning_rate": 1.322801674914351e-06, + "loss": 1.2815, + "step": 2241 + }, + { + "epoch": 2.303030303030303, + "grad_norm": 2.6499500274658203, + "learning_rate": 1.3208983631518843e-06, + "loss": 1.3231, + "step": 2242 + }, + { + "epoch": 2.3040575243965074, + "grad_norm": 2.6495258808135986, + "learning_rate": 1.3189950513894176e-06, + "loss": 1.3084, + "step": 2243 + }, + { + "epoch": 2.305084745762712, + "grad_norm": 2.6610732078552246, + "learning_rate": 1.3170917396269512e-06, + "loss": 1.3182, + "step": 2244 + }, + { + "epoch": 2.306111967128916, + "grad_norm": 2.8128137588500977, + "learning_rate": 1.3151884278644843e-06, + "loss": 1.327, + "step": 2245 + }, + { + "epoch": 2.3071391884951207, + "grad_norm": 2.7301197052001953, + "learning_rate": 1.3132851161020177e-06, + "loss": 1.2781, + "step": 2246 + }, + { + "epoch": 2.3081664098613253, + "grad_norm": 2.9533233642578125, + "learning_rate": 1.311381804339551e-06, + "loss": 1.3728, + "step": 2247 + }, + { + "epoch": 2.3091936312275294, + "grad_norm": 2.6429927349090576, + "learning_rate": 1.3094784925770842e-06, + "loss": 1.2931, + "step": 2248 + }, + { + "epoch": 2.310220852593734, + "grad_norm": 2.649885416030884, + "learning_rate": 1.3075751808146176e-06, + "loss": 1.3158, + "step": 2249 + }, + { + "epoch": 2.3112480739599386, + "grad_norm": 2.6279773712158203, + "learning_rate": 1.3056718690521507e-06, + "loss": 1.2875, + "step": 2250 + }, + { + "epoch": 2.3112480739599386, + "eval_loss": 1.777553915977478, + "eval_runtime": 38.0475, + "eval_samples_per_second": 26.283, + "eval_steps_per_second": 1.104, + "step": 2250 + }, + { + "epoch": 2.3122752953261427, + "grad_norm": 2.6415603160858154, + "learning_rate": 1.303768557289684e-06, + "loss": 1.289, + "step": 2251 + }, + { + "epoch": 2.3133025166923473, + "grad_norm": 2.586108684539795, + "learning_rate": 1.3018652455272175e-06, + "loss": 1.2985, + "step": 2252 + }, + { + "epoch": 2.3143297380585515, + "grad_norm": 2.592040538787842, + "learning_rate": 1.2999619337647506e-06, + "loss": 1.2964, + "step": 2253 + }, + { + "epoch": 2.315356959424756, + "grad_norm": 2.6993772983551025, + "learning_rate": 1.2980586220022842e-06, + "loss": 1.3257, + "step": 2254 + }, + { + "epoch": 2.3163841807909606, + "grad_norm": 2.7006278038024902, + "learning_rate": 1.2961553102398175e-06, + "loss": 1.3463, + "step": 2255 + }, + { + "epoch": 2.317411402157165, + "grad_norm": 2.749055862426758, + "learning_rate": 1.2942519984773507e-06, + "loss": 1.3183, + "step": 2256 + }, + { + "epoch": 2.3184386235233694, + "grad_norm": 2.8093061447143555, + "learning_rate": 1.292348686714884e-06, + "loss": 1.2923, + "step": 2257 + }, + { + "epoch": 2.3194658448895735, + "grad_norm": 2.6441662311553955, + "learning_rate": 1.2904453749524174e-06, + "loss": 1.3332, + "step": 2258 + }, + { + "epoch": 2.320493066255778, + "grad_norm": 2.588270425796509, + "learning_rate": 1.2885420631899506e-06, + "loss": 1.2563, + "step": 2259 + }, + { + "epoch": 2.3215202876219827, + "grad_norm": 2.659374952316284, + "learning_rate": 1.286638751427484e-06, + "loss": 1.2948, + "step": 2260 + }, + { + "epoch": 2.3215202876219827, + "eval_loss": 1.7904608249664307, + "eval_runtime": 38.0335, + "eval_samples_per_second": 26.293, + "eval_steps_per_second": 1.104, + "step": 2260 + }, + { + "epoch": 2.322547508988187, + "grad_norm": 2.6419613361358643, + "learning_rate": 1.284735439665017e-06, + "loss": 1.3183, + "step": 2261 + }, + { + "epoch": 2.3235747303543914, + "grad_norm": 2.7432727813720703, + "learning_rate": 1.2828321279025504e-06, + "loss": 1.3301, + "step": 2262 + }, + { + "epoch": 2.324601951720596, + "grad_norm": 2.7030932903289795, + "learning_rate": 1.2809288161400838e-06, + "loss": 1.3091, + "step": 2263 + }, + { + "epoch": 2.3256291730868, + "grad_norm": 2.621903896331787, + "learning_rate": 1.279025504377617e-06, + "loss": 1.2925, + "step": 2264 + }, + { + "epoch": 2.3266563944530048, + "grad_norm": 2.811548948287964, + "learning_rate": 1.2771221926151505e-06, + "loss": 1.3613, + "step": 2265 + }, + { + "epoch": 2.327683615819209, + "grad_norm": 2.543182134628296, + "learning_rate": 1.275218880852684e-06, + "loss": 1.2973, + "step": 2266 + }, + { + "epoch": 2.3287108371854135, + "grad_norm": 2.6845998764038086, + "learning_rate": 1.273315569090217e-06, + "loss": 1.2844, + "step": 2267 + }, + { + "epoch": 2.3297380585516176, + "grad_norm": 2.812823534011841, + "learning_rate": 1.2714122573277504e-06, + "loss": 1.3607, + "step": 2268 + }, + { + "epoch": 2.3307652799178222, + "grad_norm": 3.091813325881958, + "learning_rate": 1.2695089455652838e-06, + "loss": 1.3648, + "step": 2269 + }, + { + "epoch": 2.331792501284027, + "grad_norm": 2.7097294330596924, + "learning_rate": 1.267605633802817e-06, + "loss": 1.326, + "step": 2270 + }, + { + "epoch": 2.331792501284027, + "eval_loss": 1.791210412979126, + "eval_runtime": 38.0649, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 2270 + }, + { + "epoch": 2.332819722650231, + "grad_norm": 2.68989896774292, + "learning_rate": 1.2657023220403503e-06, + "loss": 1.3186, + "step": 2271 + }, + { + "epoch": 2.3338469440164356, + "grad_norm": 2.7425694465637207, + "learning_rate": 1.2637990102778836e-06, + "loss": 1.3014, + "step": 2272 + }, + { + "epoch": 2.33487416538264, + "grad_norm": 2.641493797302246, + "learning_rate": 1.2618956985154168e-06, + "loss": 1.3023, + "step": 2273 + }, + { + "epoch": 2.3359013867488443, + "grad_norm": 2.6478893756866455, + "learning_rate": 1.2599923867529502e-06, + "loss": 1.3508, + "step": 2274 + }, + { + "epoch": 2.336928608115049, + "grad_norm": 2.744187831878662, + "learning_rate": 1.2580890749904837e-06, + "loss": 1.3164, + "step": 2275 + }, + { + "epoch": 2.337955829481253, + "grad_norm": 2.823047161102295, + "learning_rate": 1.2561857632280169e-06, + "loss": 1.3418, + "step": 2276 + }, + { + "epoch": 2.3389830508474576, + "grad_norm": 2.9462826251983643, + "learning_rate": 1.2542824514655502e-06, + "loss": 1.3739, + "step": 2277 + }, + { + "epoch": 2.340010272213662, + "grad_norm": 2.7174856662750244, + "learning_rate": 1.2523791397030834e-06, + "loss": 1.3254, + "step": 2278 + }, + { + "epoch": 2.3410374935798663, + "grad_norm": 2.6504011154174805, + "learning_rate": 1.2504758279406168e-06, + "loss": 1.306, + "step": 2279 + }, + { + "epoch": 2.342064714946071, + "grad_norm": 2.701664447784424, + "learning_rate": 1.2485725161781501e-06, + "loss": 1.3302, + "step": 2280 + }, + { + "epoch": 2.342064714946071, + "eval_loss": 1.770766019821167, + "eval_runtime": 38.0551, + "eval_samples_per_second": 26.278, + "eval_steps_per_second": 1.104, + "step": 2280 + }, + { + "epoch": 2.343091936312275, + "grad_norm": 2.7651360034942627, + "learning_rate": 1.2466692044156833e-06, + "loss": 1.4107, + "step": 2281 + }, + { + "epoch": 2.3441191576784797, + "grad_norm": 2.60256028175354, + "learning_rate": 1.2447658926532168e-06, + "loss": 1.2769, + "step": 2282 + }, + { + "epoch": 2.3451463790446843, + "grad_norm": 2.6077115535736084, + "learning_rate": 1.24286258089075e-06, + "loss": 1.3148, + "step": 2283 + }, + { + "epoch": 2.3461736004108884, + "grad_norm": 2.6873903274536133, + "learning_rate": 1.2409592691282834e-06, + "loss": 1.2795, + "step": 2284 + }, + { + "epoch": 2.347200821777093, + "grad_norm": 2.9477450847625732, + "learning_rate": 1.2390559573658165e-06, + "loss": 1.3257, + "step": 2285 + }, + { + "epoch": 2.3482280431432976, + "grad_norm": 2.799549102783203, + "learning_rate": 1.2371526456033499e-06, + "loss": 1.3274, + "step": 2286 + }, + { + "epoch": 2.3492552645095017, + "grad_norm": 2.7725980281829834, + "learning_rate": 1.2352493338408832e-06, + "loss": 1.2803, + "step": 2287 + }, + { + "epoch": 2.3502824858757063, + "grad_norm": 3.027376413345337, + "learning_rate": 1.2333460220784166e-06, + "loss": 1.3907, + "step": 2288 + }, + { + "epoch": 2.3513097072419105, + "grad_norm": 2.7917747497558594, + "learning_rate": 1.23144271031595e-06, + "loss": 1.3458, + "step": 2289 + }, + { + "epoch": 2.352336928608115, + "grad_norm": 2.7375993728637695, + "learning_rate": 1.2295393985534831e-06, + "loss": 1.3496, + "step": 2290 + }, + { + "epoch": 2.352336928608115, + "eval_loss": 1.7680095434188843, + "eval_runtime": 38.0425, + "eval_samples_per_second": 26.286, + "eval_steps_per_second": 1.104, + "step": 2290 + }, + { + "epoch": 2.3533641499743196, + "grad_norm": 2.593547821044922, + "learning_rate": 1.2276360867910165e-06, + "loss": 1.2395, + "step": 2291 + }, + { + "epoch": 2.354391371340524, + "grad_norm": 2.7914979457855225, + "learning_rate": 1.2257327750285498e-06, + "loss": 1.3629, + "step": 2292 + }, + { + "epoch": 2.3554185927067284, + "grad_norm": 2.802846670150757, + "learning_rate": 1.2238294632660832e-06, + "loss": 1.2883, + "step": 2293 + }, + { + "epoch": 2.3564458140729325, + "grad_norm": 2.5035150051116943, + "learning_rate": 1.2219261515036164e-06, + "loss": 1.2894, + "step": 2294 + }, + { + "epoch": 2.357473035439137, + "grad_norm": 2.745405435562134, + "learning_rate": 1.2200228397411497e-06, + "loss": 1.3327, + "step": 2295 + }, + { + "epoch": 2.3585002568053417, + "grad_norm": 2.9135396480560303, + "learning_rate": 1.2181195279786829e-06, + "loss": 1.3636, + "step": 2296 + }, + { + "epoch": 2.359527478171546, + "grad_norm": 2.8527722358703613, + "learning_rate": 1.2162162162162164e-06, + "loss": 1.3059, + "step": 2297 + }, + { + "epoch": 2.3605546995377504, + "grad_norm": 2.641160249710083, + "learning_rate": 1.2143129044537496e-06, + "loss": 1.295, + "step": 2298 + }, + { + "epoch": 2.361581920903955, + "grad_norm": 2.6421735286712646, + "learning_rate": 1.212409592691283e-06, + "loss": 1.3052, + "step": 2299 + }, + { + "epoch": 2.362609142270159, + "grad_norm": 2.6424500942230225, + "learning_rate": 1.2105062809288163e-06, + "loss": 1.3108, + "step": 2300 + }, + { + "epoch": 2.362609142270159, + "eval_loss": 1.765913724899292, + "eval_runtime": 38.0324, + "eval_samples_per_second": 26.293, + "eval_steps_per_second": 1.104, + "step": 2300 + }, + { + "epoch": 2.3636363636363638, + "grad_norm": 2.814763307571411, + "learning_rate": 1.2086029691663495e-06, + "loss": 1.3234, + "step": 2301 + }, + { + "epoch": 2.364663585002568, + "grad_norm": 2.5212297439575195, + "learning_rate": 1.2066996574038828e-06, + "loss": 1.2656, + "step": 2302 + }, + { + "epoch": 2.3656908063687725, + "grad_norm": 2.8822665214538574, + "learning_rate": 1.2047963456414162e-06, + "loss": 1.3229, + "step": 2303 + }, + { + "epoch": 2.366718027734977, + "grad_norm": 2.6476027965545654, + "learning_rate": 1.2028930338789496e-06, + "loss": 1.3194, + "step": 2304 + }, + { + "epoch": 2.3677452491011812, + "grad_norm": 2.634824514389038, + "learning_rate": 1.2009897221164827e-06, + "loss": 1.2621, + "step": 2305 + }, + { + "epoch": 2.368772470467386, + "grad_norm": 2.7401492595672607, + "learning_rate": 1.199086410354016e-06, + "loss": 1.3255, + "step": 2306 + }, + { + "epoch": 2.36979969183359, + "grad_norm": 2.6971330642700195, + "learning_rate": 1.1971830985915492e-06, + "loss": 1.3003, + "step": 2307 + }, + { + "epoch": 2.3708269131997946, + "grad_norm": 2.7432987689971924, + "learning_rate": 1.1952797868290828e-06, + "loss": 1.2978, + "step": 2308 + }, + { + "epoch": 2.371854134565999, + "grad_norm": 2.7680413722991943, + "learning_rate": 1.193376475066616e-06, + "loss": 1.3168, + "step": 2309 + }, + { + "epoch": 2.3728813559322033, + "grad_norm": 2.7269768714904785, + "learning_rate": 1.1914731633041493e-06, + "loss": 1.3253, + "step": 2310 + }, + { + "epoch": 2.3728813559322033, + "eval_loss": 1.7663168907165527, + "eval_runtime": 37.7171, + "eval_samples_per_second": 26.513, + "eval_steps_per_second": 1.114, + "step": 2310 + }, + { + "epoch": 2.373908577298408, + "grad_norm": 2.6725943088531494, + "learning_rate": 1.1895698515416827e-06, + "loss": 1.2929, + "step": 2311 + }, + { + "epoch": 2.3749357986646125, + "grad_norm": 2.5516929626464844, + "learning_rate": 1.1876665397792158e-06, + "loss": 1.2936, + "step": 2312 + }, + { + "epoch": 2.3759630200308166, + "grad_norm": 2.7443480491638184, + "learning_rate": 1.1857632280167492e-06, + "loss": 1.3228, + "step": 2313 + }, + { + "epoch": 2.376990241397021, + "grad_norm": 2.7275967597961426, + "learning_rate": 1.1838599162542825e-06, + "loss": 1.2625, + "step": 2314 + }, + { + "epoch": 2.3780174627632253, + "grad_norm": 2.8085899353027344, + "learning_rate": 1.181956604491816e-06, + "loss": 1.3307, + "step": 2315 + }, + { + "epoch": 2.37904468412943, + "grad_norm": 2.6398675441741943, + "learning_rate": 1.180053292729349e-06, + "loss": 1.3144, + "step": 2316 + }, + { + "epoch": 2.380071905495634, + "grad_norm": 2.800874948501587, + "learning_rate": 1.1781499809668824e-06, + "loss": 1.3722, + "step": 2317 + }, + { + "epoch": 2.3810991268618387, + "grad_norm": 2.65503191947937, + "learning_rate": 1.1762466692044158e-06, + "loss": 1.3142, + "step": 2318 + }, + { + "epoch": 2.3821263482280433, + "grad_norm": 2.7648980617523193, + "learning_rate": 1.1743433574419491e-06, + "loss": 1.3473, + "step": 2319 + }, + { + "epoch": 2.3831535695942474, + "grad_norm": 2.656358242034912, + "learning_rate": 1.1724400456794823e-06, + "loss": 1.2912, + "step": 2320 + }, + { + "epoch": 2.3831535695942474, + "eval_loss": 1.7656192779541016, + "eval_runtime": 37.8498, + "eval_samples_per_second": 26.42, + "eval_steps_per_second": 1.11, + "step": 2320 + }, + { + "epoch": 2.384180790960452, + "grad_norm": 2.813964605331421, + "learning_rate": 1.1705367339170157e-06, + "loss": 1.3367, + "step": 2321 + }, + { + "epoch": 2.3852080123266566, + "grad_norm": 2.7210686206817627, + "learning_rate": 1.168633422154549e-06, + "loss": 1.3138, + "step": 2322 + }, + { + "epoch": 2.3862352336928607, + "grad_norm": 2.65240216255188, + "learning_rate": 1.1667301103920824e-06, + "loss": 1.2812, + "step": 2323 + }, + { + "epoch": 2.3872624550590653, + "grad_norm": 2.672790288925171, + "learning_rate": 1.1648267986296157e-06, + "loss": 1.2484, + "step": 2324 + }, + { + "epoch": 2.3882896764252695, + "grad_norm": 2.737819194793701, + "learning_rate": 1.162923486867149e-06, + "loss": 1.3027, + "step": 2325 + }, + { + "epoch": 2.389316897791474, + "grad_norm": 2.662963390350342, + "learning_rate": 1.1610201751046823e-06, + "loss": 1.3115, + "step": 2326 + }, + { + "epoch": 2.3903441191576786, + "grad_norm": 2.83809232711792, + "learning_rate": 1.1591168633422154e-06, + "loss": 1.3755, + "step": 2327 + }, + { + "epoch": 2.391371340523883, + "grad_norm": 2.722337484359741, + "learning_rate": 1.157213551579749e-06, + "loss": 1.3285, + "step": 2328 + }, + { + "epoch": 2.3923985618900874, + "grad_norm": 2.699765205383301, + "learning_rate": 1.1553102398172821e-06, + "loss": 1.3057, + "step": 2329 + }, + { + "epoch": 2.3934257832562915, + "grad_norm": 2.5885703563690186, + "learning_rate": 1.1534069280548155e-06, + "loss": 1.2963, + "step": 2330 + }, + { + "epoch": 2.3934257832562915, + "eval_loss": 1.7690107822418213, + "eval_runtime": 37.9639, + "eval_samples_per_second": 26.341, + "eval_steps_per_second": 1.106, + "step": 2330 + }, + { + "epoch": 2.394453004622496, + "grad_norm": 2.732694387435913, + "learning_rate": 1.1515036162923487e-06, + "loss": 1.3144, + "step": 2331 + }, + { + "epoch": 2.3954802259887007, + "grad_norm": 2.6117560863494873, + "learning_rate": 1.149600304529882e-06, + "loss": 1.283, + "step": 2332 + }, + { + "epoch": 2.396507447354905, + "grad_norm": 2.6792120933532715, + "learning_rate": 1.1476969927674154e-06, + "loss": 1.2959, + "step": 2333 + }, + { + "epoch": 2.3975346687211094, + "grad_norm": 2.7566049098968506, + "learning_rate": 1.1457936810049487e-06, + "loss": 1.309, + "step": 2334 + }, + { + "epoch": 2.398561890087314, + "grad_norm": 2.9152936935424805, + "learning_rate": 1.143890369242482e-06, + "loss": 1.3432, + "step": 2335 + }, + { + "epoch": 2.399589111453518, + "grad_norm": 2.6911866664886475, + "learning_rate": 1.1419870574800153e-06, + "loss": 1.2709, + "step": 2336 + }, + { + "epoch": 2.4006163328197228, + "grad_norm": 2.67596697807312, + "learning_rate": 1.1400837457175486e-06, + "loss": 1.2974, + "step": 2337 + }, + { + "epoch": 2.401643554185927, + "grad_norm": 2.5220348834991455, + "learning_rate": 1.138180433955082e-06, + "loss": 1.2691, + "step": 2338 + }, + { + "epoch": 2.4026707755521315, + "grad_norm": 2.7079286575317383, + "learning_rate": 1.1362771221926153e-06, + "loss": 1.3438, + "step": 2339 + }, + { + "epoch": 2.403697996918336, + "grad_norm": 2.5796804428100586, + "learning_rate": 1.1343738104301485e-06, + "loss": 1.3093, + "step": 2340 + }, + { + "epoch": 2.403697996918336, + "eval_loss": 1.7833822965621948, + "eval_runtime": 38.0179, + "eval_samples_per_second": 26.303, + "eval_steps_per_second": 1.105, + "step": 2340 + }, + { + "epoch": 2.4047252182845402, + "grad_norm": 2.533360242843628, + "learning_rate": 1.1324704986676819e-06, + "loss": 1.2876, + "step": 2341 + }, + { + "epoch": 2.405752439650745, + "grad_norm": 2.8825149536132812, + "learning_rate": 1.130567186905215e-06, + "loss": 1.3615, + "step": 2342 + }, + { + "epoch": 2.406779661016949, + "grad_norm": 2.5276525020599365, + "learning_rate": 1.1286638751427486e-06, + "loss": 1.2974, + "step": 2343 + }, + { + "epoch": 2.4078068823831535, + "grad_norm": 2.660332202911377, + "learning_rate": 1.1267605633802817e-06, + "loss": 1.3046, + "step": 2344 + }, + { + "epoch": 2.408834103749358, + "grad_norm": 2.81973934173584, + "learning_rate": 1.124857251617815e-06, + "loss": 1.329, + "step": 2345 + }, + { + "epoch": 2.4098613251155623, + "grad_norm": 2.6832475662231445, + "learning_rate": 1.1229539398553485e-06, + "loss": 1.3621, + "step": 2346 + }, + { + "epoch": 2.410888546481767, + "grad_norm": 2.5246076583862305, + "learning_rate": 1.1210506280928816e-06, + "loss": 1.3043, + "step": 2347 + }, + { + "epoch": 2.4119157678479715, + "grad_norm": 2.889202117919922, + "learning_rate": 1.119147316330415e-06, + "loss": 1.3412, + "step": 2348 + }, + { + "epoch": 2.4129429892141756, + "grad_norm": 2.761444091796875, + "learning_rate": 1.1172440045679483e-06, + "loss": 1.3069, + "step": 2349 + }, + { + "epoch": 2.41397021058038, + "grad_norm": 2.7676186561584473, + "learning_rate": 1.1153406928054817e-06, + "loss": 1.3057, + "step": 2350 + }, + { + "epoch": 2.41397021058038, + "eval_loss": 1.7964602708816528, + "eval_runtime": 38.0018, + "eval_samples_per_second": 26.315, + "eval_steps_per_second": 1.105, + "step": 2350 + }, + { + "epoch": 2.4149974319465843, + "grad_norm": 2.761202096939087, + "learning_rate": 1.1134373810430148e-06, + "loss": 1.3355, + "step": 2351 + }, + { + "epoch": 2.416024653312789, + "grad_norm": 2.74330735206604, + "learning_rate": 1.1115340692805482e-06, + "loss": 1.3415, + "step": 2352 + }, + { + "epoch": 2.4170518746789935, + "grad_norm": 2.6388099193573, + "learning_rate": 1.1096307575180816e-06, + "loss": 1.314, + "step": 2353 + }, + { + "epoch": 2.4180790960451977, + "grad_norm": 2.6058437824249268, + "learning_rate": 1.107727445755615e-06, + "loss": 1.2745, + "step": 2354 + }, + { + "epoch": 2.4191063174114023, + "grad_norm": 2.674837350845337, + "learning_rate": 1.105824133993148e-06, + "loss": 1.2943, + "step": 2355 + }, + { + "epoch": 2.4201335387776064, + "grad_norm": 2.777331829071045, + "learning_rate": 1.1039208222306814e-06, + "loss": 1.3136, + "step": 2356 + }, + { + "epoch": 2.421160760143811, + "grad_norm": 2.7025792598724365, + "learning_rate": 1.1020175104682148e-06, + "loss": 1.3051, + "step": 2357 + }, + { + "epoch": 2.4221879815100156, + "grad_norm": 2.570185899734497, + "learning_rate": 1.1001141987057482e-06, + "loss": 1.2964, + "step": 2358 + }, + { + "epoch": 2.4232152028762197, + "grad_norm": 2.6053781509399414, + "learning_rate": 1.0982108869432813e-06, + "loss": 1.3073, + "step": 2359 + }, + { + "epoch": 2.4242424242424243, + "grad_norm": 2.680983543395996, + "learning_rate": 1.0963075751808147e-06, + "loss": 1.3415, + "step": 2360 + }, + { + "epoch": 2.4242424242424243, + "eval_loss": 1.7697505950927734, + "eval_runtime": 37.9457, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 2360 + }, + { + "epoch": 2.425269645608629, + "grad_norm": 2.558598518371582, + "learning_rate": 1.094404263418348e-06, + "loss": 1.3074, + "step": 2361 + }, + { + "epoch": 2.426296866974833, + "grad_norm": 2.6231720447540283, + "learning_rate": 1.0925009516558812e-06, + "loss": 1.3031, + "step": 2362 + }, + { + "epoch": 2.4273240883410376, + "grad_norm": 2.730914831161499, + "learning_rate": 1.0905976398934148e-06, + "loss": 1.3017, + "step": 2363 + }, + { + "epoch": 2.428351309707242, + "grad_norm": 2.82137393951416, + "learning_rate": 1.088694328130948e-06, + "loss": 1.3154, + "step": 2364 + }, + { + "epoch": 2.4293785310734464, + "grad_norm": 2.6977908611297607, + "learning_rate": 1.0867910163684813e-06, + "loss": 1.2774, + "step": 2365 + }, + { + "epoch": 2.4304057524396505, + "grad_norm": 2.784348964691162, + "learning_rate": 1.0848877046060144e-06, + "loss": 1.3035, + "step": 2366 + }, + { + "epoch": 2.431432973805855, + "grad_norm": 2.663001537322998, + "learning_rate": 1.0829843928435478e-06, + "loss": 1.3026, + "step": 2367 + }, + { + "epoch": 2.4324601951720597, + "grad_norm": 2.7743303775787354, + "learning_rate": 1.0810810810810812e-06, + "loss": 1.3085, + "step": 2368 + }, + { + "epoch": 2.433487416538264, + "grad_norm": 2.797743082046509, + "learning_rate": 1.0791777693186145e-06, + "loss": 1.3254, + "step": 2369 + }, + { + "epoch": 2.4345146379044684, + "grad_norm": 2.625260829925537, + "learning_rate": 1.0772744575561479e-06, + "loss": 1.3098, + "step": 2370 + }, + { + "epoch": 2.4345146379044684, + "eval_loss": 1.7611181735992432, + "eval_runtime": 37.9845, + "eval_samples_per_second": 26.327, + "eval_steps_per_second": 1.106, + "step": 2370 + }, + { + "epoch": 2.435541859270673, + "grad_norm": 2.580679178237915, + "learning_rate": 1.075371145793681e-06, + "loss": 1.2925, + "step": 2371 + }, + { + "epoch": 2.436569080636877, + "grad_norm": 2.655032157897949, + "learning_rate": 1.0734678340312144e-06, + "loss": 1.3122, + "step": 2372 + }, + { + "epoch": 2.4375963020030817, + "grad_norm": 2.6039280891418457, + "learning_rate": 1.0715645222687478e-06, + "loss": 1.3037, + "step": 2373 + }, + { + "epoch": 2.438623523369286, + "grad_norm": 2.532874822616577, + "learning_rate": 1.0696612105062811e-06, + "loss": 1.2921, + "step": 2374 + }, + { + "epoch": 2.4396507447354905, + "grad_norm": 2.71315336227417, + "learning_rate": 1.0677578987438143e-06, + "loss": 1.3249, + "step": 2375 + }, + { + "epoch": 2.440677966101695, + "grad_norm": 2.797964334487915, + "learning_rate": 1.0658545869813476e-06, + "loss": 1.2849, + "step": 2376 + }, + { + "epoch": 2.441705187467899, + "grad_norm": 2.8070008754730225, + "learning_rate": 1.0639512752188808e-06, + "loss": 1.2956, + "step": 2377 + }, + { + "epoch": 2.442732408834104, + "grad_norm": 2.6451804637908936, + "learning_rate": 1.0620479634564144e-06, + "loss": 1.3076, + "step": 2378 + }, + { + "epoch": 2.443759630200308, + "grad_norm": 2.6175758838653564, + "learning_rate": 1.0601446516939475e-06, + "loss": 1.2766, + "step": 2379 + }, + { + "epoch": 2.4447868515665125, + "grad_norm": 2.5483524799346924, + "learning_rate": 1.0582413399314809e-06, + "loss": 1.2777, + "step": 2380 + }, + { + "epoch": 2.4447868515665125, + "eval_loss": 1.7605587244033813, + "eval_runtime": 37.9393, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 2380 + }, + { + "epoch": 2.445814072932717, + "grad_norm": 2.6982779502868652, + "learning_rate": 1.0563380281690142e-06, + "loss": 1.2826, + "step": 2381 + }, + { + "epoch": 2.4468412942989213, + "grad_norm": 2.575270891189575, + "learning_rate": 1.0544347164065474e-06, + "loss": 1.2855, + "step": 2382 + }, + { + "epoch": 2.447868515665126, + "grad_norm": 2.7389488220214844, + "learning_rate": 1.0525314046440808e-06, + "loss": 1.3179, + "step": 2383 + }, + { + "epoch": 2.4488957370313305, + "grad_norm": 2.555800199508667, + "learning_rate": 1.0506280928816141e-06, + "loss": 1.2919, + "step": 2384 + }, + { + "epoch": 2.4499229583975346, + "grad_norm": 2.580599546432495, + "learning_rate": 1.0487247811191475e-06, + "loss": 1.2943, + "step": 2385 + }, + { + "epoch": 2.450950179763739, + "grad_norm": 2.763772964477539, + "learning_rate": 1.0468214693566806e-06, + "loss": 1.3054, + "step": 2386 + }, + { + "epoch": 2.4519774011299433, + "grad_norm": 2.7483937740325928, + "learning_rate": 1.044918157594214e-06, + "loss": 1.3428, + "step": 2387 + }, + { + "epoch": 2.453004622496148, + "grad_norm": 2.644343376159668, + "learning_rate": 1.0430148458317474e-06, + "loss": 1.299, + "step": 2388 + }, + { + "epoch": 2.4540318438623525, + "grad_norm": 2.6011316776275635, + "learning_rate": 1.0411115340692807e-06, + "loss": 1.271, + "step": 2389 + }, + { + "epoch": 2.4550590652285567, + "grad_norm": 2.8445608615875244, + "learning_rate": 1.0392082223068139e-06, + "loss": 1.3426, + "step": 2390 + }, + { + "epoch": 2.4550590652285567, + "eval_loss": 1.764119267463684, + "eval_runtime": 37.9678, + "eval_samples_per_second": 26.338, + "eval_steps_per_second": 1.106, + "step": 2390 + }, + { + "epoch": 2.4560862865947612, + "grad_norm": 2.7232890129089355, + "learning_rate": 1.0373049105443472e-06, + "loss": 1.3083, + "step": 2391 + }, + { + "epoch": 2.4571135079609654, + "grad_norm": 2.7028329372406006, + "learning_rate": 1.0354015987818806e-06, + "loss": 1.3206, + "step": 2392 + }, + { + "epoch": 2.45814072932717, + "grad_norm": 2.7255351543426514, + "learning_rate": 1.033498287019414e-06, + "loss": 1.2851, + "step": 2393 + }, + { + "epoch": 2.4591679506933746, + "grad_norm": 2.9048759937286377, + "learning_rate": 1.0315949752569471e-06, + "loss": 1.335, + "step": 2394 + }, + { + "epoch": 2.4601951720595787, + "grad_norm": 2.680952310562134, + "learning_rate": 1.0296916634944805e-06, + "loss": 1.2832, + "step": 2395 + }, + { + "epoch": 2.4612223934257833, + "grad_norm": 2.6702933311462402, + "learning_rate": 1.0277883517320138e-06, + "loss": 1.3055, + "step": 2396 + }, + { + "epoch": 2.462249614791988, + "grad_norm": 2.851332664489746, + "learning_rate": 1.025885039969547e-06, + "loss": 1.3422, + "step": 2397 + }, + { + "epoch": 2.463276836158192, + "grad_norm": 2.5233068466186523, + "learning_rate": 1.0239817282070806e-06, + "loss": 1.2412, + "step": 2398 + }, + { + "epoch": 2.4643040575243966, + "grad_norm": 2.6488661766052246, + "learning_rate": 1.0220784164446137e-06, + "loss": 1.2725, + "step": 2399 + }, + { + "epoch": 2.4653312788906008, + "grad_norm": 2.558196544647217, + "learning_rate": 1.020175104682147e-06, + "loss": 1.3043, + "step": 2400 + }, + { + "epoch": 2.4653312788906008, + "eval_loss": 1.7678213119506836, + "eval_runtime": 37.9906, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 2400 + }, + { + "epoch": 2.4663585002568054, + "grad_norm": 2.601803779602051, + "learning_rate": 1.0182717929196802e-06, + "loss": 1.2979, + "step": 2401 + }, + { + "epoch": 2.4673857216230095, + "grad_norm": 2.6790895462036133, + "learning_rate": 1.0163684811572136e-06, + "loss": 1.3215, + "step": 2402 + }, + { + "epoch": 2.468412942989214, + "grad_norm": 2.6814472675323486, + "learning_rate": 1.014465169394747e-06, + "loss": 1.2561, + "step": 2403 + }, + { + "epoch": 2.4694401643554187, + "grad_norm": 2.7114996910095215, + "learning_rate": 1.0125618576322803e-06, + "loss": 1.2966, + "step": 2404 + }, + { + "epoch": 2.470467385721623, + "grad_norm": 2.6394622325897217, + "learning_rate": 1.0106585458698137e-06, + "loss": 1.278, + "step": 2405 + }, + { + "epoch": 2.4714946070878274, + "grad_norm": 2.656283140182495, + "learning_rate": 1.0087552341073468e-06, + "loss": 1.2798, + "step": 2406 + }, + { + "epoch": 2.472521828454032, + "grad_norm": 2.719639539718628, + "learning_rate": 1.0068519223448802e-06, + "loss": 1.3656, + "step": 2407 + }, + { + "epoch": 2.473549049820236, + "grad_norm": 2.7754440307617188, + "learning_rate": 1.0049486105824133e-06, + "loss": 1.3384, + "step": 2408 + }, + { + "epoch": 2.4745762711864407, + "grad_norm": 2.6474952697753906, + "learning_rate": 1.003045298819947e-06, + "loss": 1.2859, + "step": 2409 + }, + { + "epoch": 2.4756034925526453, + "grad_norm": 2.6622884273529053, + "learning_rate": 1.00114198705748e-06, + "loss": 1.2753, + "step": 2410 + }, + { + "epoch": 2.4756034925526453, + "eval_loss": 1.7663556337356567, + "eval_runtime": 37.6695, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 2410 + }, + { + "epoch": 2.4766307139188495, + "grad_norm": 2.721719741821289, + "learning_rate": 9.992386752950134e-07, + "loss": 1.2807, + "step": 2411 + }, + { + "epoch": 2.477657935285054, + "grad_norm": 2.5972471237182617, + "learning_rate": 9.973353635325466e-07, + "loss": 1.269, + "step": 2412 + }, + { + "epoch": 2.478685156651258, + "grad_norm": 2.616769790649414, + "learning_rate": 9.9543205177008e-07, + "loss": 1.2789, + "step": 2413 + }, + { + "epoch": 2.479712378017463, + "grad_norm": 2.573221206665039, + "learning_rate": 9.935287400076133e-07, + "loss": 1.282, + "step": 2414 + }, + { + "epoch": 2.480739599383667, + "grad_norm": 2.6644439697265625, + "learning_rate": 9.916254282451467e-07, + "loss": 1.2676, + "step": 2415 + }, + { + "epoch": 2.4817668207498715, + "grad_norm": 2.6775522232055664, + "learning_rate": 9.8972211648268e-07, + "loss": 1.2999, + "step": 2416 + }, + { + "epoch": 2.482794042116076, + "grad_norm": 2.785191297531128, + "learning_rate": 9.878188047202132e-07, + "loss": 1.3202, + "step": 2417 + }, + { + "epoch": 2.4838212634822803, + "grad_norm": 2.7016355991363525, + "learning_rate": 9.859154929577465e-07, + "loss": 1.295, + "step": 2418 + }, + { + "epoch": 2.484848484848485, + "grad_norm": 2.6089084148406982, + "learning_rate": 9.8401218119528e-07, + "loss": 1.2846, + "step": 2419 + }, + { + "epoch": 2.4858757062146895, + "grad_norm": 2.8124489784240723, + "learning_rate": 9.821088694328133e-07, + "loss": 1.2883, + "step": 2420 + }, + { + "epoch": 2.4858757062146895, + "eval_loss": 1.7576653957366943, + "eval_runtime": 37.867, + "eval_samples_per_second": 26.408, + "eval_steps_per_second": 1.109, + "step": 2420 + }, + { + "epoch": 2.4869029275808936, + "grad_norm": 2.7730648517608643, + "learning_rate": 9.802055576703464e-07, + "loss": 1.3321, + "step": 2421 + }, + { + "epoch": 2.487930148947098, + "grad_norm": 2.725834846496582, + "learning_rate": 9.783022459078798e-07, + "loss": 1.2931, + "step": 2422 + }, + { + "epoch": 2.4889573703133023, + "grad_norm": 2.607725143432617, + "learning_rate": 9.76398934145413e-07, + "loss": 1.3002, + "step": 2423 + }, + { + "epoch": 2.489984591679507, + "grad_norm": 2.6080846786499023, + "learning_rate": 9.744956223829465e-07, + "loss": 1.3102, + "step": 2424 + }, + { + "epoch": 2.4910118130457115, + "grad_norm": 2.6172633171081543, + "learning_rate": 9.725923106204797e-07, + "loss": 1.3234, + "step": 2425 + }, + { + "epoch": 2.4920390344119157, + "grad_norm": 2.6781582832336426, + "learning_rate": 9.70688998858013e-07, + "loss": 1.3442, + "step": 2426 + }, + { + "epoch": 2.4930662557781202, + "grad_norm": 2.8291099071502686, + "learning_rate": 9.687856870955464e-07, + "loss": 1.3401, + "step": 2427 + }, + { + "epoch": 2.4940934771443244, + "grad_norm": 2.589317560195923, + "learning_rate": 9.668823753330795e-07, + "loss": 1.3241, + "step": 2428 + }, + { + "epoch": 2.495120698510529, + "grad_norm": 2.5891406536102295, + "learning_rate": 9.64979063570613e-07, + "loss": 1.2589, + "step": 2429 + }, + { + "epoch": 2.4961479198767336, + "grad_norm": 2.589101791381836, + "learning_rate": 9.630757518081463e-07, + "loss": 1.3036, + "step": 2430 + }, + { + "epoch": 2.4961479198767336, + "eval_loss": 1.7622722387313843, + "eval_runtime": 38.0143, + "eval_samples_per_second": 26.306, + "eval_steps_per_second": 1.105, + "step": 2430 + }, + { + "epoch": 2.4971751412429377, + "grad_norm": 2.75529146194458, + "learning_rate": 9.611724400456796e-07, + "loss": 1.2969, + "step": 2431 + }, + { + "epoch": 2.4982023626091423, + "grad_norm": 2.557976722717285, + "learning_rate": 9.592691282832128e-07, + "loss": 1.2881, + "step": 2432 + }, + { + "epoch": 2.499229583975347, + "grad_norm": 2.638200521469116, + "learning_rate": 9.573658165207461e-07, + "loss": 1.3007, + "step": 2433 + }, + { + "epoch": 2.500256805341551, + "grad_norm": 2.6940360069274902, + "learning_rate": 9.554625047582795e-07, + "loss": 1.316, + "step": 2434 + }, + { + "epoch": 2.5012840267077556, + "grad_norm": 2.517789602279663, + "learning_rate": 9.535591929958129e-07, + "loss": 1.2844, + "step": 2435 + }, + { + "epoch": 2.50231124807396, + "grad_norm": 2.5775527954101562, + "learning_rate": 9.516558812333461e-07, + "loss": 1.3084, + "step": 2436 + }, + { + "epoch": 2.5033384694401644, + "grad_norm": 2.605240821838379, + "learning_rate": 9.497525694708794e-07, + "loss": 1.3242, + "step": 2437 + }, + { + "epoch": 2.5043656908063685, + "grad_norm": 2.56835675239563, + "learning_rate": 9.478492577084126e-07, + "loss": 1.2799, + "step": 2438 + }, + { + "epoch": 2.505392912172573, + "grad_norm": 2.564911127090454, + "learning_rate": 9.459459459459461e-07, + "loss": 1.2987, + "step": 2439 + }, + { + "epoch": 2.5064201335387777, + "grad_norm": 2.6561388969421387, + "learning_rate": 9.440426341834794e-07, + "loss": 1.29, + "step": 2440 + }, + { + "epoch": 2.5064201335387777, + "eval_loss": 1.766339898109436, + "eval_runtime": 38.0448, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 2440 + }, + { + "epoch": 2.507447354904982, + "grad_norm": 2.5807695388793945, + "learning_rate": 9.421393224210126e-07, + "loss": 1.3058, + "step": 2441 + }, + { + "epoch": 2.5084745762711864, + "grad_norm": 2.6242141723632812, + "learning_rate": 9.402360106585459e-07, + "loss": 1.3012, + "step": 2442 + }, + { + "epoch": 2.509501797637391, + "grad_norm": 2.6174166202545166, + "learning_rate": 9.383326988960792e-07, + "loss": 1.2716, + "step": 2443 + }, + { + "epoch": 2.510529019003595, + "grad_norm": 2.5244762897491455, + "learning_rate": 9.364293871336126e-07, + "loss": 1.2625, + "step": 2444 + }, + { + "epoch": 2.5115562403697997, + "grad_norm": 2.61104679107666, + "learning_rate": 9.345260753711458e-07, + "loss": 1.3049, + "step": 2445 + }, + { + "epoch": 2.5125834617360043, + "grad_norm": 2.735579013824463, + "learning_rate": 9.326227636086792e-07, + "loss": 1.3159, + "step": 2446 + }, + { + "epoch": 2.5136106831022085, + "grad_norm": 2.5533411502838135, + "learning_rate": 9.307194518462125e-07, + "loss": 1.2611, + "step": 2447 + }, + { + "epoch": 2.514637904468413, + "grad_norm": 2.705458164215088, + "learning_rate": 9.288161400837457e-07, + "loss": 1.3754, + "step": 2448 + }, + { + "epoch": 2.515665125834617, + "grad_norm": 2.5326483249664307, + "learning_rate": 9.269128283212792e-07, + "loss": 1.272, + "step": 2449 + }, + { + "epoch": 2.516692347200822, + "grad_norm": 2.6174025535583496, + "learning_rate": 9.250095165588125e-07, + "loss": 1.3067, + "step": 2450 + }, + { + "epoch": 2.516692347200822, + "eval_loss": 1.7679108381271362, + "eval_runtime": 38.0209, + "eval_samples_per_second": 26.301, + "eval_steps_per_second": 1.105, + "step": 2450 + }, + { + "epoch": 2.517719568567026, + "grad_norm": 2.6596055030822754, + "learning_rate": 9.231062047963457e-07, + "loss": 1.3485, + "step": 2451 + }, + { + "epoch": 2.5187467899332305, + "grad_norm": 2.7891628742218018, + "learning_rate": 9.21202893033879e-07, + "loss": 1.3559, + "step": 2452 + }, + { + "epoch": 2.519774011299435, + "grad_norm": 2.560696840286255, + "learning_rate": 9.192995812714123e-07, + "loss": 1.3187, + "step": 2453 + }, + { + "epoch": 2.5208012326656393, + "grad_norm": 2.6536800861358643, + "learning_rate": 9.173962695089457e-07, + "loss": 1.3129, + "step": 2454 + }, + { + "epoch": 2.521828454031844, + "grad_norm": 2.6149590015411377, + "learning_rate": 9.154929577464789e-07, + "loss": 1.3111, + "step": 2455 + }, + { + "epoch": 2.5228556753980484, + "grad_norm": 2.672088861465454, + "learning_rate": 9.135896459840123e-07, + "loss": 1.3172, + "step": 2456 + }, + { + "epoch": 2.5238828967642526, + "grad_norm": 2.7675814628601074, + "learning_rate": 9.116863342215456e-07, + "loss": 1.3029, + "step": 2457 + }, + { + "epoch": 2.524910118130457, + "grad_norm": 2.931568145751953, + "learning_rate": 9.097830224590788e-07, + "loss": 1.3424, + "step": 2458 + }, + { + "epoch": 2.5259373394966618, + "grad_norm": 2.6001431941986084, + "learning_rate": 9.078797106966122e-07, + "loss": 1.2604, + "step": 2459 + }, + { + "epoch": 2.526964560862866, + "grad_norm": 2.8028078079223633, + "learning_rate": 9.059763989341455e-07, + "loss": 1.3386, + "step": 2460 + }, + { + "epoch": 2.526964560862866, + "eval_loss": 1.7729586362838745, + "eval_runtime": 38.0609, + "eval_samples_per_second": 26.274, + "eval_steps_per_second": 1.103, + "step": 2460 + }, + { + "epoch": 2.5279917822290705, + "grad_norm": 2.6733815670013428, + "learning_rate": 9.040730871716788e-07, + "loss": 1.2824, + "step": 2461 + }, + { + "epoch": 2.5290190035952747, + "grad_norm": 2.6084516048431396, + "learning_rate": 9.021697754092121e-07, + "loss": 1.3075, + "step": 2462 + }, + { + "epoch": 2.5300462249614792, + "grad_norm": 2.6287853717803955, + "learning_rate": 9.002664636467453e-07, + "loss": 1.3112, + "step": 2463 + }, + { + "epoch": 2.5310734463276834, + "grad_norm": 2.702615976333618, + "learning_rate": 8.983631518842788e-07, + "loss": 1.3282, + "step": 2464 + }, + { + "epoch": 2.532100667693888, + "grad_norm": 2.60542893409729, + "learning_rate": 8.96459840121812e-07, + "loss": 1.3386, + "step": 2465 + }, + { + "epoch": 2.5331278890600926, + "grad_norm": 2.6633386611938477, + "learning_rate": 8.945565283593453e-07, + "loss": 1.3019, + "step": 2466 + }, + { + "epoch": 2.5341551104262967, + "grad_norm": 2.5968427658081055, + "learning_rate": 8.926532165968787e-07, + "loss": 1.3085, + "step": 2467 + }, + { + "epoch": 2.5351823317925013, + "grad_norm": 2.746767044067383, + "learning_rate": 8.907499048344119e-07, + "loss": 1.3233, + "step": 2468 + }, + { + "epoch": 2.536209553158706, + "grad_norm": 2.589085578918457, + "learning_rate": 8.888465930719453e-07, + "loss": 1.258, + "step": 2469 + }, + { + "epoch": 2.53723677452491, + "grad_norm": 2.61661434173584, + "learning_rate": 8.869432813094786e-07, + "loss": 1.2664, + "step": 2470 + }, + { + "epoch": 2.53723677452491, + "eval_loss": 1.7717350721359253, + "eval_runtime": 38.0362, + "eval_samples_per_second": 26.291, + "eval_steps_per_second": 1.104, + "step": 2470 + }, + { + "epoch": 2.5382639958911146, + "grad_norm": 2.657177448272705, + "learning_rate": 8.850399695470119e-07, + "loss": 1.3052, + "step": 2471 + }, + { + "epoch": 2.539291217257319, + "grad_norm": 2.603573799133301, + "learning_rate": 8.831366577845452e-07, + "loss": 1.3235, + "step": 2472 + }, + { + "epoch": 2.5403184386235234, + "grad_norm": 2.6154067516326904, + "learning_rate": 8.812333460220784e-07, + "loss": 1.2742, + "step": 2473 + }, + { + "epoch": 2.541345659989728, + "grad_norm": 2.515085220336914, + "learning_rate": 8.793300342596119e-07, + "loss": 1.2535, + "step": 2474 + }, + { + "epoch": 2.542372881355932, + "grad_norm": 2.6949517726898193, + "learning_rate": 8.774267224971451e-07, + "loss": 1.3088, + "step": 2475 + }, + { + "epoch": 2.5434001027221367, + "grad_norm": 2.891838312149048, + "learning_rate": 8.755234107346784e-07, + "loss": 1.3416, + "step": 2476 + }, + { + "epoch": 2.544427324088341, + "grad_norm": 2.623758554458618, + "learning_rate": 8.736200989722117e-07, + "loss": 1.2775, + "step": 2477 + }, + { + "epoch": 2.5454545454545454, + "grad_norm": 2.65680193901062, + "learning_rate": 8.71716787209745e-07, + "loss": 1.2999, + "step": 2478 + }, + { + "epoch": 2.54648176682075, + "grad_norm": 2.85903000831604, + "learning_rate": 8.698134754472784e-07, + "loss": 1.3453, + "step": 2479 + }, + { + "epoch": 2.547508988186954, + "grad_norm": 2.7487850189208984, + "learning_rate": 8.679101636848116e-07, + "loss": 1.2432, + "step": 2480 + }, + { + "epoch": 2.547508988186954, + "eval_loss": 1.7745047807693481, + "eval_runtime": 37.9926, + "eval_samples_per_second": 26.321, + "eval_steps_per_second": 1.105, + "step": 2480 + }, + { + "epoch": 2.5485362095531587, + "grad_norm": 2.737975835800171, + "learning_rate": 8.66006851922345e-07, + "loss": 1.3116, + "step": 2481 + }, + { + "epoch": 2.5495634309193633, + "grad_norm": 2.6670281887054443, + "learning_rate": 8.641035401598783e-07, + "loss": 1.2989, + "step": 2482 + }, + { + "epoch": 2.5505906522855675, + "grad_norm": 2.6732940673828125, + "learning_rate": 8.622002283974115e-07, + "loss": 1.2907, + "step": 2483 + }, + { + "epoch": 2.551617873651772, + "grad_norm": 2.7414166927337646, + "learning_rate": 8.60296916634945e-07, + "loss": 1.3347, + "step": 2484 + }, + { + "epoch": 2.5526450950179767, + "grad_norm": 2.557168960571289, + "learning_rate": 8.583936048724782e-07, + "loss": 1.2939, + "step": 2485 + }, + { + "epoch": 2.553672316384181, + "grad_norm": 2.5460598468780518, + "learning_rate": 8.564902931100115e-07, + "loss": 1.285, + "step": 2486 + }, + { + "epoch": 2.554699537750385, + "grad_norm": 2.7185349464416504, + "learning_rate": 8.545869813475448e-07, + "loss": 1.3066, + "step": 2487 + }, + { + "epoch": 2.5557267591165895, + "grad_norm": 2.6216139793395996, + "learning_rate": 8.526836695850781e-07, + "loss": 1.3032, + "step": 2488 + }, + { + "epoch": 2.556753980482794, + "grad_norm": 2.7842869758605957, + "learning_rate": 8.507803578226115e-07, + "loss": 1.3279, + "step": 2489 + }, + { + "epoch": 2.5577812018489983, + "grad_norm": 2.6502983570098877, + "learning_rate": 8.488770460601447e-07, + "loss": 1.3119, + "step": 2490 + }, + { + "epoch": 2.5577812018489983, + "eval_loss": 1.7768635749816895, + "eval_runtime": 37.9867, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 2490 + }, + { + "epoch": 2.558808423215203, + "grad_norm": 2.695314645767212, + "learning_rate": 8.46973734297678e-07, + "loss": 1.2769, + "step": 2491 + }, + { + "epoch": 2.5598356445814074, + "grad_norm": 2.661893844604492, + "learning_rate": 8.450704225352114e-07, + "loss": 1.2892, + "step": 2492 + }, + { + "epoch": 2.5608628659476116, + "grad_norm": 2.686737537384033, + "learning_rate": 8.431671107727446e-07, + "loss": 1.2883, + "step": 2493 + }, + { + "epoch": 2.561890087313816, + "grad_norm": 2.661504030227661, + "learning_rate": 8.41263799010278e-07, + "loss": 1.3139, + "step": 2494 + }, + { + "epoch": 2.5629173086800208, + "grad_norm": 2.6665518283843994, + "learning_rate": 8.393604872478113e-07, + "loss": 1.2981, + "step": 2495 + }, + { + "epoch": 2.563944530046225, + "grad_norm": 2.722047805786133, + "learning_rate": 8.374571754853446e-07, + "loss": 1.2862, + "step": 2496 + }, + { + "epoch": 2.5649717514124295, + "grad_norm": 2.6358418464660645, + "learning_rate": 8.355538637228778e-07, + "loss": 1.2934, + "step": 2497 + }, + { + "epoch": 2.5659989727786336, + "grad_norm": 2.7137503623962402, + "learning_rate": 8.336505519604111e-07, + "loss": 1.2874, + "step": 2498 + }, + { + "epoch": 2.5670261941448382, + "grad_norm": 2.589345693588257, + "learning_rate": 8.317472401979445e-07, + "loss": 1.3229, + "step": 2499 + }, + { + "epoch": 2.5680534155110424, + "grad_norm": 2.597588062286377, + "learning_rate": 8.298439284354778e-07, + "loss": 1.3481, + "step": 2500 + }, + { + "epoch": 2.5680534155110424, + "eval_loss": 1.7751270532608032, + "eval_runtime": 37.9958, + "eval_samples_per_second": 26.319, + "eval_steps_per_second": 1.105, + "step": 2500 + }, + { + "epoch": 2.569080636877247, + "grad_norm": 2.7178173065185547, + "learning_rate": 8.279406166730111e-07, + "loss": 1.3318, + "step": 2501 + }, + { + "epoch": 2.5701078582434516, + "grad_norm": 2.6459460258483887, + "learning_rate": 8.260373049105444e-07, + "loss": 1.332, + "step": 2502 + }, + { + "epoch": 2.5711350796096557, + "grad_norm": 2.6359541416168213, + "learning_rate": 8.241339931480777e-07, + "loss": 1.311, + "step": 2503 + }, + { + "epoch": 2.5721623009758603, + "grad_norm": 2.706570863723755, + "learning_rate": 8.22230681385611e-07, + "loss": 1.3391, + "step": 2504 + }, + { + "epoch": 2.573189522342065, + "grad_norm": 2.591161012649536, + "learning_rate": 8.203273696231444e-07, + "loss": 1.3058, + "step": 2505 + }, + { + "epoch": 2.574216743708269, + "grad_norm": 2.595672607421875, + "learning_rate": 8.184240578606777e-07, + "loss": 1.2646, + "step": 2506 + }, + { + "epoch": 2.5752439650744736, + "grad_norm": 2.7604939937591553, + "learning_rate": 8.165207460982109e-07, + "loss": 1.3191, + "step": 2507 + }, + { + "epoch": 2.576271186440678, + "grad_norm": 2.6298060417175293, + "learning_rate": 8.146174343357442e-07, + "loss": 1.2735, + "step": 2508 + }, + { + "epoch": 2.5772984078068824, + "grad_norm": 2.7220308780670166, + "learning_rate": 8.127141225732775e-07, + "loss": 1.3164, + "step": 2509 + }, + { + "epoch": 2.578325629173087, + "grad_norm": 2.875459671020508, + "learning_rate": 8.108108108108109e-07, + "loss": 1.3236, + "step": 2510 + }, + { + "epoch": 2.578325629173087, + "eval_loss": 1.7866631746292114, + "eval_runtime": 37.6739, + "eval_samples_per_second": 26.544, + "eval_steps_per_second": 1.115, + "step": 2510 + }, + { + "epoch": 2.579352850539291, + "grad_norm": 2.747443675994873, + "learning_rate": 8.089074990483442e-07, + "loss": 1.3147, + "step": 2511 + }, + { + "epoch": 2.5803800719054957, + "grad_norm": 2.7403504848480225, + "learning_rate": 8.070041872858774e-07, + "loss": 1.3016, + "step": 2512 + }, + { + "epoch": 2.5814072932717, + "grad_norm": 2.700188159942627, + "learning_rate": 8.051008755234108e-07, + "loss": 1.3118, + "step": 2513 + }, + { + "epoch": 2.5824345146379044, + "grad_norm": 2.575113296508789, + "learning_rate": 8.031975637609441e-07, + "loss": 1.255, + "step": 2514 + }, + { + "epoch": 2.583461736004109, + "grad_norm": 2.7600820064544678, + "learning_rate": 8.012942519984774e-07, + "loss": 1.305, + "step": 2515 + }, + { + "epoch": 2.584488957370313, + "grad_norm": 2.5697903633117676, + "learning_rate": 7.993909402360108e-07, + "loss": 1.327, + "step": 2516 + }, + { + "epoch": 2.5855161787365177, + "grad_norm": 2.617521047592163, + "learning_rate": 7.97487628473544e-07, + "loss": 1.3196, + "step": 2517 + }, + { + "epoch": 2.5865434001027223, + "grad_norm": 2.708566427230835, + "learning_rate": 7.955843167110773e-07, + "loss": 1.2993, + "step": 2518 + }, + { + "epoch": 2.5875706214689265, + "grad_norm": 2.683959722518921, + "learning_rate": 7.936810049486106e-07, + "loss": 1.2748, + "step": 2519 + }, + { + "epoch": 2.588597842835131, + "grad_norm": 2.6184964179992676, + "learning_rate": 7.91777693186144e-07, + "loss": 1.2935, + "step": 2520 + }, + { + "epoch": 2.588597842835131, + "eval_loss": 1.7722079753875732, + "eval_runtime": 37.7975, + "eval_samples_per_second": 26.457, + "eval_steps_per_second": 1.111, + "step": 2520 + }, + { + "epoch": 2.5896250642013356, + "grad_norm": 2.538283586502075, + "learning_rate": 7.898743814236773e-07, + "loss": 1.3149, + "step": 2521 + }, + { + "epoch": 2.59065228556754, + "grad_norm": 2.5143439769744873, + "learning_rate": 7.879710696612105e-07, + "loss": 1.2907, + "step": 2522 + }, + { + "epoch": 2.5916795069337444, + "grad_norm": 2.5969698429107666, + "learning_rate": 7.860677578987438e-07, + "loss": 1.3575, + "step": 2523 + }, + { + "epoch": 2.5927067282999485, + "grad_norm": 2.617640733718872, + "learning_rate": 7.841644461362772e-07, + "loss": 1.3152, + "step": 2524 + }, + { + "epoch": 2.593733949666153, + "grad_norm": 2.649030923843384, + "learning_rate": 7.822611343738105e-07, + "loss": 1.2949, + "step": 2525 + }, + { + "epoch": 2.5947611710323573, + "grad_norm": 2.694195032119751, + "learning_rate": 7.803578226113438e-07, + "loss": 1.3333, + "step": 2526 + }, + { + "epoch": 2.595788392398562, + "grad_norm": 2.559192657470703, + "learning_rate": 7.784545108488771e-07, + "loss": 1.282, + "step": 2527 + }, + { + "epoch": 2.5968156137647664, + "grad_norm": 2.5788307189941406, + "learning_rate": 7.765511990864104e-07, + "loss": 1.3033, + "step": 2528 + }, + { + "epoch": 2.5978428351309706, + "grad_norm": 2.8756697177886963, + "learning_rate": 7.746478873239437e-07, + "loss": 1.3569, + "step": 2529 + }, + { + "epoch": 2.598870056497175, + "grad_norm": 2.632779598236084, + "learning_rate": 7.727445755614771e-07, + "loss": 1.319, + "step": 2530 + }, + { + "epoch": 2.598870056497175, + "eval_loss": 1.7722663879394531, + "eval_runtime": 37.9166, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 2530 + }, + { + "epoch": 2.5998972778633798, + "grad_norm": 2.76119065284729, + "learning_rate": 7.708412637990104e-07, + "loss": 1.3142, + "step": 2531 + }, + { + "epoch": 2.600924499229584, + "grad_norm": 2.593604803085327, + "learning_rate": 7.689379520365436e-07, + "loss": 1.3286, + "step": 2532 + }, + { + "epoch": 2.6019517205957885, + "grad_norm": 2.6926052570343018, + "learning_rate": 7.670346402740769e-07, + "loss": 1.2933, + "step": 2533 + }, + { + "epoch": 2.602978941961993, + "grad_norm": 2.734140157699585, + "learning_rate": 7.651313285116103e-07, + "loss": 1.3187, + "step": 2534 + }, + { + "epoch": 2.6040061633281972, + "grad_norm": 2.7742362022399902, + "learning_rate": 7.632280167491436e-07, + "loss": 1.3427, + "step": 2535 + }, + { + "epoch": 2.6050333846944014, + "grad_norm": 2.5474061965942383, + "learning_rate": 7.613247049866769e-07, + "loss": 1.2652, + "step": 2536 + }, + { + "epoch": 2.606060606060606, + "grad_norm": 2.704885959625244, + "learning_rate": 7.594213932242102e-07, + "loss": 1.2405, + "step": 2537 + }, + { + "epoch": 2.6070878274268106, + "grad_norm": 2.67887020111084, + "learning_rate": 7.575180814617435e-07, + "loss": 1.2702, + "step": 2538 + }, + { + "epoch": 2.6081150487930147, + "grad_norm": 2.6307291984558105, + "learning_rate": 7.556147696992767e-07, + "loss": 1.3046, + "step": 2539 + }, + { + "epoch": 2.6091422701592193, + "grad_norm": 2.7046597003936768, + "learning_rate": 7.537114579368102e-07, + "loss": 1.3121, + "step": 2540 + }, + { + "epoch": 2.6091422701592193, + "eval_loss": 1.7705127000808716, + "eval_runtime": 38.0131, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 2540 + }, + { + "epoch": 2.610169491525424, + "grad_norm": 2.564473867416382, + "learning_rate": 7.518081461743435e-07, + "loss": 1.28, + "step": 2541 + }, + { + "epoch": 2.611196712891628, + "grad_norm": 2.63580322265625, + "learning_rate": 7.499048344118767e-07, + "loss": 1.2945, + "step": 2542 + }, + { + "epoch": 2.6122239342578326, + "grad_norm": 2.6448638439178467, + "learning_rate": 7.4800152264941e-07, + "loss": 1.3265, + "step": 2543 + }, + { + "epoch": 2.613251155624037, + "grad_norm": 2.490447521209717, + "learning_rate": 7.460982108869432e-07, + "loss": 1.3035, + "step": 2544 + }, + { + "epoch": 2.6142783769902413, + "grad_norm": 2.5429506301879883, + "learning_rate": 7.441948991244767e-07, + "loss": 1.288, + "step": 2545 + }, + { + "epoch": 2.615305598356446, + "grad_norm": 2.6845016479492188, + "learning_rate": 7.4229158736201e-07, + "loss": 1.3119, + "step": 2546 + }, + { + "epoch": 2.61633281972265, + "grad_norm": 2.6247217655181885, + "learning_rate": 7.403882755995432e-07, + "loss": 1.3213, + "step": 2547 + }, + { + "epoch": 2.6173600410888547, + "grad_norm": 2.5123143196105957, + "learning_rate": 7.384849638370766e-07, + "loss": 1.2831, + "step": 2548 + }, + { + "epoch": 2.618387262455059, + "grad_norm": 2.5071399211883545, + "learning_rate": 7.365816520746098e-07, + "loss": 1.2393, + "step": 2549 + }, + { + "epoch": 2.6194144838212634, + "grad_norm": 2.744621515274048, + "learning_rate": 7.346783403121432e-07, + "loss": 1.3083, + "step": 2550 + }, + { + "epoch": 2.6194144838212634, + "eval_loss": 1.7846120595932007, + "eval_runtime": 37.9987, + "eval_samples_per_second": 26.317, + "eval_steps_per_second": 1.105, + "step": 2550 + }, + { + "epoch": 2.620441705187468, + "grad_norm": 2.714101552963257, + "learning_rate": 7.327750285496766e-07, + "loss": 1.287, + "step": 2551 + }, + { + "epoch": 2.621468926553672, + "grad_norm": 2.792046546936035, + "learning_rate": 7.308717167872098e-07, + "loss": 1.2873, + "step": 2552 + }, + { + "epoch": 2.6224961479198767, + "grad_norm": 2.493668556213379, + "learning_rate": 7.289684050247431e-07, + "loss": 1.255, + "step": 2553 + }, + { + "epoch": 2.6235233692860813, + "grad_norm": 2.631716012954712, + "learning_rate": 7.270650932622763e-07, + "loss": 1.3418, + "step": 2554 + }, + { + "epoch": 2.6245505906522855, + "grad_norm": 2.443617820739746, + "learning_rate": 7.251617814998098e-07, + "loss": 1.2512, + "step": 2555 + }, + { + "epoch": 2.62557781201849, + "grad_norm": 2.4819979667663574, + "learning_rate": 7.232584697373431e-07, + "loss": 1.3131, + "step": 2556 + }, + { + "epoch": 2.6266050333846946, + "grad_norm": 2.5260822772979736, + "learning_rate": 7.213551579748763e-07, + "loss": 1.2812, + "step": 2557 + }, + { + "epoch": 2.627632254750899, + "grad_norm": 2.614377498626709, + "learning_rate": 7.194518462124096e-07, + "loss": 1.3065, + "step": 2558 + }, + { + "epoch": 2.6286594761171034, + "grad_norm": 2.619166374206543, + "learning_rate": 7.175485344499429e-07, + "loss": 1.3813, + "step": 2559 + }, + { + "epoch": 2.6296866974833075, + "grad_norm": 2.6223137378692627, + "learning_rate": 7.156452226874763e-07, + "loss": 1.2856, + "step": 2560 + }, + { + "epoch": 2.6296866974833075, + "eval_loss": 1.7834736108779907, + "eval_runtime": 38.0086, + "eval_samples_per_second": 26.31, + "eval_steps_per_second": 1.105, + "step": 2560 + }, + { + "epoch": 2.630713918849512, + "grad_norm": 2.7283148765563965, + "learning_rate": 7.137419109250096e-07, + "loss": 1.375, + "step": 2561 + }, + { + "epoch": 2.6317411402157163, + "grad_norm": 2.629539728164673, + "learning_rate": 7.118385991625429e-07, + "loss": 1.2915, + "step": 2562 + }, + { + "epoch": 2.632768361581921, + "grad_norm": 2.576698064804077, + "learning_rate": 7.099352874000762e-07, + "loss": 1.2818, + "step": 2563 + }, + { + "epoch": 2.6337955829481254, + "grad_norm": 2.5807430744171143, + "learning_rate": 7.080319756376094e-07, + "loss": 1.285, + "step": 2564 + }, + { + "epoch": 2.6348228043143296, + "grad_norm": 2.461386203765869, + "learning_rate": 7.061286638751429e-07, + "loss": 1.2594, + "step": 2565 + }, + { + "epoch": 2.635850025680534, + "grad_norm": 2.6482107639312744, + "learning_rate": 7.042253521126762e-07, + "loss": 1.3038, + "step": 2566 + }, + { + "epoch": 2.6368772470467388, + "grad_norm": 2.637989044189453, + "learning_rate": 7.023220403502094e-07, + "loss": 1.2453, + "step": 2567 + }, + { + "epoch": 2.637904468412943, + "grad_norm": 2.6575927734375, + "learning_rate": 7.004187285877427e-07, + "loss": 1.3413, + "step": 2568 + }, + { + "epoch": 2.6389316897791475, + "grad_norm": 2.616055727005005, + "learning_rate": 6.98515416825276e-07, + "loss": 1.3212, + "step": 2569 + }, + { + "epoch": 2.639958911145352, + "grad_norm": 2.5758297443389893, + "learning_rate": 6.966121050628094e-07, + "loss": 1.2999, + "step": 2570 + }, + { + "epoch": 2.639958911145352, + "eval_loss": 1.7819379568099976, + "eval_runtime": 38.0358, + "eval_samples_per_second": 26.291, + "eval_steps_per_second": 1.104, + "step": 2570 + }, + { + "epoch": 2.6409861325115562, + "grad_norm": 2.7511165142059326, + "learning_rate": 6.947087933003427e-07, + "loss": 1.3519, + "step": 2571 + }, + { + "epoch": 2.642013353877761, + "grad_norm": 2.6978421211242676, + "learning_rate": 6.92805481537876e-07, + "loss": 1.2944, + "step": 2572 + }, + { + "epoch": 2.643040575243965, + "grad_norm": 2.7035439014434814, + "learning_rate": 6.909021697754093e-07, + "loss": 1.3496, + "step": 2573 + }, + { + "epoch": 2.6440677966101696, + "grad_norm": 2.6517820358276367, + "learning_rate": 6.889988580129425e-07, + "loss": 1.3233, + "step": 2574 + }, + { + "epoch": 2.6450950179763737, + "grad_norm": 2.65956974029541, + "learning_rate": 6.870955462504759e-07, + "loss": 1.2826, + "step": 2575 + }, + { + "epoch": 2.6461222393425783, + "grad_norm": 2.671339988708496, + "learning_rate": 6.851922344880093e-07, + "loss": 1.3257, + "step": 2576 + }, + { + "epoch": 2.647149460708783, + "grad_norm": 2.648261785507202, + "learning_rate": 6.832889227255425e-07, + "loss": 1.3015, + "step": 2577 + }, + { + "epoch": 2.648176682074987, + "grad_norm": 2.700082540512085, + "learning_rate": 6.813856109630758e-07, + "loss": 1.3129, + "step": 2578 + }, + { + "epoch": 2.6492039034411916, + "grad_norm": 2.7364165782928467, + "learning_rate": 6.79482299200609e-07, + "loss": 1.3185, + "step": 2579 + }, + { + "epoch": 2.650231124807396, + "grad_norm": 2.5831615924835205, + "learning_rate": 6.775789874381425e-07, + "loss": 1.2796, + "step": 2580 + }, + { + "epoch": 2.650231124807396, + "eval_loss": 1.7886762619018555, + "eval_runtime": 38.0048, + "eval_samples_per_second": 26.312, + "eval_steps_per_second": 1.105, + "step": 2580 + }, + { + "epoch": 2.6512583461736003, + "grad_norm": 2.6888351440429688, + "learning_rate": 6.756756756756758e-07, + "loss": 1.3447, + "step": 2581 + }, + { + "epoch": 2.652285567539805, + "grad_norm": 2.643916606903076, + "learning_rate": 6.73772363913209e-07, + "loss": 1.3234, + "step": 2582 + }, + { + "epoch": 2.653312788906009, + "grad_norm": 2.593222141265869, + "learning_rate": 6.718690521507424e-07, + "loss": 1.3164, + "step": 2583 + }, + { + "epoch": 2.6543400102722137, + "grad_norm": 2.535433053970337, + "learning_rate": 6.699657403882756e-07, + "loss": 1.2637, + "step": 2584 + }, + { + "epoch": 2.655367231638418, + "grad_norm": 2.5683724880218506, + "learning_rate": 6.68062428625809e-07, + "loss": 1.328, + "step": 2585 + }, + { + "epoch": 2.6563944530046224, + "grad_norm": 2.6119470596313477, + "learning_rate": 6.661591168633424e-07, + "loss": 1.3393, + "step": 2586 + }, + { + "epoch": 2.657421674370827, + "grad_norm": 2.646026134490967, + "learning_rate": 6.642558051008756e-07, + "loss": 1.3219, + "step": 2587 + }, + { + "epoch": 2.658448895737031, + "grad_norm": 2.7056374549865723, + "learning_rate": 6.623524933384089e-07, + "loss": 1.326, + "step": 2588 + }, + { + "epoch": 2.6594761171032357, + "grad_norm": 2.638749361038208, + "learning_rate": 6.604491815759421e-07, + "loss": 1.2979, + "step": 2589 + }, + { + "epoch": 2.6605033384694403, + "grad_norm": 2.6330184936523438, + "learning_rate": 6.585458698134756e-07, + "loss": 1.292, + "step": 2590 + }, + { + "epoch": 2.6605033384694403, + "eval_loss": 1.7978336811065674, + "eval_runtime": 38.019, + "eval_samples_per_second": 26.303, + "eval_steps_per_second": 1.105, + "step": 2590 + }, + { + "epoch": 2.6615305598356445, + "grad_norm": 2.577609062194824, + "learning_rate": 6.566425580510089e-07, + "loss": 1.2541, + "step": 2591 + }, + { + "epoch": 2.662557781201849, + "grad_norm": 2.666790008544922, + "learning_rate": 6.547392462885421e-07, + "loss": 1.3183, + "step": 2592 + }, + { + "epoch": 2.6635850025680536, + "grad_norm": 2.6628904342651367, + "learning_rate": 6.528359345260754e-07, + "loss": 1.3303, + "step": 2593 + }, + { + "epoch": 2.664612223934258, + "grad_norm": 2.6553924083709717, + "learning_rate": 6.509326227636087e-07, + "loss": 1.3019, + "step": 2594 + }, + { + "epoch": 2.6656394453004624, + "grad_norm": 2.7442100048065186, + "learning_rate": 6.490293110011421e-07, + "loss": 1.3111, + "step": 2595 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 2.6254689693450928, + "learning_rate": 6.471259992386753e-07, + "loss": 1.2986, + "step": 2596 + }, + { + "epoch": 2.667693888032871, + "grad_norm": 2.5442848205566406, + "learning_rate": 6.452226874762087e-07, + "loss": 1.3471, + "step": 2597 + }, + { + "epoch": 2.6687211093990753, + "grad_norm": 2.5967025756835938, + "learning_rate": 6.43319375713742e-07, + "loss": 1.2879, + "step": 2598 + }, + { + "epoch": 2.66974833076528, + "grad_norm": 2.69215989112854, + "learning_rate": 6.414160639512752e-07, + "loss": 1.3201, + "step": 2599 + }, + { + "epoch": 2.6707755521314844, + "grad_norm": 2.7470815181732178, + "learning_rate": 6.395127521888085e-07, + "loss": 1.31, + "step": 2600 + }, + { + "epoch": 2.6707755521314844, + "eval_loss": 1.8076120615005493, + "eval_runtime": 38.023, + "eval_samples_per_second": 26.3, + "eval_steps_per_second": 1.105, + "step": 2600 + }, + { + "epoch": 2.6718027734976886, + "grad_norm": 2.681523561477661, + "learning_rate": 6.37609440426342e-07, + "loss": 1.2818, + "step": 2601 + }, + { + "epoch": 2.672829994863893, + "grad_norm": 2.7073915004730225, + "learning_rate": 6.357061286638752e-07, + "loss": 1.3465, + "step": 2602 + }, + { + "epoch": 2.6738572162300978, + "grad_norm": 2.7676303386688232, + "learning_rate": 6.338028169014085e-07, + "loss": 1.3072, + "step": 2603 + }, + { + "epoch": 2.674884437596302, + "grad_norm": 2.7673637866973877, + "learning_rate": 6.318995051389418e-07, + "loss": 1.3415, + "step": 2604 + }, + { + "epoch": 2.6759116589625065, + "grad_norm": 2.518784284591675, + "learning_rate": 6.299961933764751e-07, + "loss": 1.2292, + "step": 2605 + }, + { + "epoch": 2.676938880328711, + "grad_norm": 2.569060802459717, + "learning_rate": 6.280928816140084e-07, + "loss": 1.3314, + "step": 2606 + }, + { + "epoch": 2.6779661016949152, + "grad_norm": 2.70438814163208, + "learning_rate": 6.261895698515417e-07, + "loss": 1.3806, + "step": 2607 + }, + { + "epoch": 2.67899332306112, + "grad_norm": 2.6343674659729004, + "learning_rate": 6.242862580890751e-07, + "loss": 1.326, + "step": 2608 + }, + { + "epoch": 2.680020544427324, + "grad_norm": 2.5554616451263428, + "learning_rate": 6.223829463266084e-07, + "loss": 1.2799, + "step": 2609 + }, + { + "epoch": 2.6810477657935285, + "grad_norm": 2.6620726585388184, + "learning_rate": 6.204796345641417e-07, + "loss": 1.2466, + "step": 2610 + }, + { + "epoch": 2.6810477657935285, + "eval_loss": 1.8074779510498047, + "eval_runtime": 37.6679, + "eval_samples_per_second": 26.548, + "eval_steps_per_second": 1.115, + "step": 2610 + }, + { + "epoch": 2.6820749871597327, + "grad_norm": 2.7202115058898926, + "learning_rate": 6.185763228016749e-07, + "loss": 1.3034, + "step": 2611 + }, + { + "epoch": 2.6831022085259373, + "grad_norm": 2.912074089050293, + "learning_rate": 6.166730110392083e-07, + "loss": 1.3341, + "step": 2612 + }, + { + "epoch": 2.684129429892142, + "grad_norm": 3.0352418422698975, + "learning_rate": 6.147696992767416e-07, + "loss": 1.3551, + "step": 2613 + }, + { + "epoch": 2.685156651258346, + "grad_norm": 2.603231191635132, + "learning_rate": 6.128663875142749e-07, + "loss": 1.3205, + "step": 2614 + }, + { + "epoch": 2.6861838726245506, + "grad_norm": 2.6780762672424316, + "learning_rate": 6.109630757518082e-07, + "loss": 1.2837, + "step": 2615 + }, + { + "epoch": 2.687211093990755, + "grad_norm": 2.8105902671813965, + "learning_rate": 6.090597639893414e-07, + "loss": 1.3635, + "step": 2616 + }, + { + "epoch": 2.6882383153569593, + "grad_norm": 2.609379291534424, + "learning_rate": 6.071564522268748e-07, + "loss": 1.2684, + "step": 2617 + }, + { + "epoch": 2.689265536723164, + "grad_norm": 2.6364426612854004, + "learning_rate": 6.052531404644082e-07, + "loss": 1.3132, + "step": 2618 + }, + { + "epoch": 2.6902927580893685, + "grad_norm": 2.7625892162323, + "learning_rate": 6.033498287019414e-07, + "loss": 1.3828, + "step": 2619 + }, + { + "epoch": 2.6913199794555727, + "grad_norm": 2.5761687755584717, + "learning_rate": 6.014465169394748e-07, + "loss": 1.2953, + "step": 2620 + }, + { + "epoch": 2.6913199794555727, + "eval_loss": 1.7990498542785645, + "eval_runtime": 37.8153, + "eval_samples_per_second": 26.444, + "eval_steps_per_second": 1.111, + "step": 2620 + }, + { + "epoch": 2.6923472008217773, + "grad_norm": 2.7242608070373535, + "learning_rate": 5.99543205177008e-07, + "loss": 1.2924, + "step": 2621 + }, + { + "epoch": 2.6933744221879814, + "grad_norm": 2.589207410812378, + "learning_rate": 5.976398934145414e-07, + "loss": 1.2918, + "step": 2622 + }, + { + "epoch": 2.694401643554186, + "grad_norm": 2.5671799182891846, + "learning_rate": 5.957365816520747e-07, + "loss": 1.288, + "step": 2623 + }, + { + "epoch": 2.69542886492039, + "grad_norm": 2.6533076763153076, + "learning_rate": 5.938332698896079e-07, + "loss": 1.3131, + "step": 2624 + }, + { + "epoch": 2.6964560862865947, + "grad_norm": 2.60552716255188, + "learning_rate": 5.919299581271413e-07, + "loss": 1.3049, + "step": 2625 + }, + { + "epoch": 2.6974833076527993, + "grad_norm": 2.6435978412628174, + "learning_rate": 5.900266463646745e-07, + "loss": 1.3068, + "step": 2626 + }, + { + "epoch": 2.6985105290190035, + "grad_norm": 2.551628589630127, + "learning_rate": 5.881233346022079e-07, + "loss": 1.2941, + "step": 2627 + }, + { + "epoch": 2.699537750385208, + "grad_norm": 2.829988479614258, + "learning_rate": 5.862200228397411e-07, + "loss": 1.3361, + "step": 2628 + }, + { + "epoch": 2.7005649717514126, + "grad_norm": 2.6916370391845703, + "learning_rate": 5.843167110772745e-07, + "loss": 1.3516, + "step": 2629 + }, + { + "epoch": 2.701592193117617, + "grad_norm": 2.6523451805114746, + "learning_rate": 5.824133993148079e-07, + "loss": 1.3286, + "step": 2630 + }, + { + "epoch": 2.701592193117617, + "eval_loss": 1.7828145027160645, + "eval_runtime": 37.9485, + "eval_samples_per_second": 26.351, + "eval_steps_per_second": 1.107, + "step": 2630 + }, + { + "epoch": 2.7026194144838214, + "grad_norm": 2.6095120906829834, + "learning_rate": 5.805100875523411e-07, + "loss": 1.2778, + "step": 2631 + }, + { + "epoch": 2.7036466358500255, + "grad_norm": 2.630551338195801, + "learning_rate": 5.786067757898745e-07, + "loss": 1.2549, + "step": 2632 + }, + { + "epoch": 2.70467385721623, + "grad_norm": 2.585043430328369, + "learning_rate": 5.767034640274078e-07, + "loss": 1.2639, + "step": 2633 + }, + { + "epoch": 2.7057010785824342, + "grad_norm": 2.8497092723846436, + "learning_rate": 5.74800152264941e-07, + "loss": 1.3636, + "step": 2634 + }, + { + "epoch": 2.706728299948639, + "grad_norm": 2.7136929035186768, + "learning_rate": 5.728968405024744e-07, + "loss": 1.2788, + "step": 2635 + }, + { + "epoch": 2.7077555213148434, + "grad_norm": 2.7176077365875244, + "learning_rate": 5.709935287400076e-07, + "loss": 1.3561, + "step": 2636 + }, + { + "epoch": 2.7087827426810476, + "grad_norm": 2.754312515258789, + "learning_rate": 5.69090216977541e-07, + "loss": 1.3105, + "step": 2637 + }, + { + "epoch": 2.709809964047252, + "grad_norm": 2.6190314292907715, + "learning_rate": 5.671869052150742e-07, + "loss": 1.2946, + "step": 2638 + }, + { + "epoch": 2.7108371854134568, + "grad_norm": 2.5728366374969482, + "learning_rate": 5.652835934526075e-07, + "loss": 1.2989, + "step": 2639 + }, + { + "epoch": 2.711864406779661, + "grad_norm": 2.5806870460510254, + "learning_rate": 5.633802816901409e-07, + "loss": 1.3021, + "step": 2640 + }, + { + "epoch": 2.711864406779661, + "eval_loss": 1.7884520292282104, + "eval_runtime": 38.0067, + "eval_samples_per_second": 26.311, + "eval_steps_per_second": 1.105, + "step": 2640 + }, + { + "epoch": 2.7128916281458655, + "grad_norm": 2.5025806427001953, + "learning_rate": 5.614769699276742e-07, + "loss": 1.2784, + "step": 2641 + }, + { + "epoch": 2.71391884951207, + "grad_norm": 2.7000210285186768, + "learning_rate": 5.595736581652075e-07, + "loss": 1.2955, + "step": 2642 + }, + { + "epoch": 2.714946070878274, + "grad_norm": 2.5144412517547607, + "learning_rate": 5.576703464027408e-07, + "loss": 1.3036, + "step": 2643 + }, + { + "epoch": 2.715973292244479, + "grad_norm": 2.511216878890991, + "learning_rate": 5.557670346402741e-07, + "loss": 1.3266, + "step": 2644 + }, + { + "epoch": 2.717000513610683, + "grad_norm": 2.7001969814300537, + "learning_rate": 5.538637228778075e-07, + "loss": 1.3082, + "step": 2645 + }, + { + "epoch": 2.7180277349768875, + "grad_norm": 2.7682130336761475, + "learning_rate": 5.519604111153407e-07, + "loss": 1.3354, + "step": 2646 + }, + { + "epoch": 2.7190549563430917, + "grad_norm": 2.7113163471221924, + "learning_rate": 5.500570993528741e-07, + "loss": 1.2606, + "step": 2647 + }, + { + "epoch": 2.7200821777092963, + "grad_norm": 2.772627592086792, + "learning_rate": 5.481537875904073e-07, + "loss": 1.3362, + "step": 2648 + }, + { + "epoch": 2.721109399075501, + "grad_norm": 2.7985000610351562, + "learning_rate": 5.462504758279406e-07, + "loss": 1.3503, + "step": 2649 + }, + { + "epoch": 2.722136620441705, + "grad_norm": 2.6800785064697266, + "learning_rate": 5.44347164065474e-07, + "loss": 1.2854, + "step": 2650 + }, + { + "epoch": 2.722136620441705, + "eval_loss": 1.7845864295959473, + "eval_runtime": 38.0453, + "eval_samples_per_second": 26.284, + "eval_steps_per_second": 1.104, + "step": 2650 + }, + { + "epoch": 2.7231638418079096, + "grad_norm": 2.610942840576172, + "learning_rate": 5.424438523030072e-07, + "loss": 1.3161, + "step": 2651 + }, + { + "epoch": 2.724191063174114, + "grad_norm": 2.7399675846099854, + "learning_rate": 5.405405405405406e-07, + "loss": 1.357, + "step": 2652 + }, + { + "epoch": 2.7252182845403183, + "grad_norm": 2.6037936210632324, + "learning_rate": 5.386372287780739e-07, + "loss": 1.3198, + "step": 2653 + }, + { + "epoch": 2.726245505906523, + "grad_norm": 2.5747087001800537, + "learning_rate": 5.367339170156072e-07, + "loss": 1.3382, + "step": 2654 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 2.8241353034973145, + "learning_rate": 5.348306052531406e-07, + "loss": 1.3049, + "step": 2655 + }, + { + "epoch": 2.7282999486389317, + "grad_norm": 2.624354600906372, + "learning_rate": 5.329272934906738e-07, + "loss": 1.3272, + "step": 2656 + }, + { + "epoch": 2.7293271700051362, + "grad_norm": 2.5246741771698, + "learning_rate": 5.310239817282072e-07, + "loss": 1.2704, + "step": 2657 + }, + { + "epoch": 2.7303543913713404, + "grad_norm": 2.817598342895508, + "learning_rate": 5.291206699657404e-07, + "loss": 1.3404, + "step": 2658 + }, + { + "epoch": 2.731381612737545, + "grad_norm": 2.4856812953948975, + "learning_rate": 5.272173582032737e-07, + "loss": 1.2904, + "step": 2659 + }, + { + "epoch": 2.732408834103749, + "grad_norm": 2.7321617603302, + "learning_rate": 5.253140464408071e-07, + "loss": 1.3131, + "step": 2660 + }, + { + "epoch": 2.732408834103749, + "eval_loss": 1.784244418144226, + "eval_runtime": 38.0266, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 2660 + }, + { + "epoch": 2.7334360554699537, + "grad_norm": 2.5358948707580566, + "learning_rate": 5.234107346783403e-07, + "loss": 1.2882, + "step": 2661 + }, + { + "epoch": 2.7344632768361583, + "grad_norm": 2.5920722484588623, + "learning_rate": 5.215074229158737e-07, + "loss": 1.2577, + "step": 2662 + }, + { + "epoch": 2.7354904982023625, + "grad_norm": 2.589002847671509, + "learning_rate": 5.196041111534069e-07, + "loss": 1.2934, + "step": 2663 + }, + { + "epoch": 2.736517719568567, + "grad_norm": 2.5873234272003174, + "learning_rate": 5.177007993909403e-07, + "loss": 1.3088, + "step": 2664 + }, + { + "epoch": 2.7375449409347716, + "grad_norm": 2.7201805114746094, + "learning_rate": 5.157974876284736e-07, + "loss": 1.3226, + "step": 2665 + }, + { + "epoch": 2.7385721623009758, + "grad_norm": 2.5658745765686035, + "learning_rate": 5.138941758660069e-07, + "loss": 1.3197, + "step": 2666 + }, + { + "epoch": 2.7395993836671804, + "grad_norm": 2.6131479740142822, + "learning_rate": 5.119908641035403e-07, + "loss": 1.3035, + "step": 2667 + }, + { + "epoch": 2.740626605033385, + "grad_norm": 2.5535855293273926, + "learning_rate": 5.100875523410735e-07, + "loss": 1.2709, + "step": 2668 + }, + { + "epoch": 2.741653826399589, + "grad_norm": 2.4551708698272705, + "learning_rate": 5.081842405786068e-07, + "loss": 1.2773, + "step": 2669 + }, + { + "epoch": 2.7426810477657937, + "grad_norm": 2.6793084144592285, + "learning_rate": 5.062809288161402e-07, + "loss": 1.2932, + "step": 2670 + }, + { + "epoch": 2.7426810477657937, + "eval_loss": 1.7706480026245117, + "eval_runtime": 38.033, + "eval_samples_per_second": 26.293, + "eval_steps_per_second": 1.104, + "step": 2670 + }, + { + "epoch": 2.743708269131998, + "grad_norm": 2.6757824420928955, + "learning_rate": 5.043776170536734e-07, + "loss": 1.3053, + "step": 2671 + }, + { + "epoch": 2.7447354904982024, + "grad_norm": 2.7082114219665527, + "learning_rate": 5.024743052912067e-07, + "loss": 1.3196, + "step": 2672 + }, + { + "epoch": 2.7457627118644066, + "grad_norm": 2.6333837509155273, + "learning_rate": 5.0057099352874e-07, + "loss": 1.3112, + "step": 2673 + }, + { + "epoch": 2.746789933230611, + "grad_norm": 2.7922589778900146, + "learning_rate": 4.986676817662733e-07, + "loss": 1.3421, + "step": 2674 + }, + { + "epoch": 2.7478171545968157, + "grad_norm": 2.5689852237701416, + "learning_rate": 4.967643700038067e-07, + "loss": 1.3001, + "step": 2675 + }, + { + "epoch": 2.74884437596302, + "grad_norm": 2.6763274669647217, + "learning_rate": 4.9486105824134e-07, + "loss": 1.3132, + "step": 2676 + }, + { + "epoch": 2.7498715973292245, + "grad_norm": 2.778526544570923, + "learning_rate": 4.929577464788733e-07, + "loss": 1.3548, + "step": 2677 + }, + { + "epoch": 2.750898818695429, + "grad_norm": 2.6535632610321045, + "learning_rate": 4.910544347164066e-07, + "loss": 1.3193, + "step": 2678 + }, + { + "epoch": 2.751926040061633, + "grad_norm": 2.5167694091796875, + "learning_rate": 4.891511229539399e-07, + "loss": 1.2681, + "step": 2679 + }, + { + "epoch": 2.752953261427838, + "grad_norm": 2.6354637145996094, + "learning_rate": 4.872478111914733e-07, + "loss": 1.3046, + "step": 2680 + }, + { + "epoch": 2.752953261427838, + "eval_loss": 1.7752299308776855, + "eval_runtime": 38.0209, + "eval_samples_per_second": 26.301, + "eval_steps_per_second": 1.105, + "step": 2680 + }, + { + "epoch": 2.753980482794042, + "grad_norm": 2.5887973308563232, + "learning_rate": 4.853444994290065e-07, + "loss": 1.2766, + "step": 2681 + }, + { + "epoch": 2.7550077041602465, + "grad_norm": 2.693185567855835, + "learning_rate": 4.834411876665398e-07, + "loss": 1.3078, + "step": 2682 + }, + { + "epoch": 2.7560349255264507, + "grad_norm": 2.7474021911621094, + "learning_rate": 4.815378759040731e-07, + "loss": 1.286, + "step": 2683 + }, + { + "epoch": 2.7570621468926553, + "grad_norm": 2.7882676124572754, + "learning_rate": 4.796345641416064e-07, + "loss": 1.3296, + "step": 2684 + }, + { + "epoch": 2.75808936825886, + "grad_norm": 2.633437156677246, + "learning_rate": 4.777312523791397e-07, + "loss": 1.29, + "step": 2685 + }, + { + "epoch": 2.759116589625064, + "grad_norm": 2.5937581062316895, + "learning_rate": 4.7582794061667306e-07, + "loss": 1.3181, + "step": 2686 + }, + { + "epoch": 2.7601438109912686, + "grad_norm": 2.710019588470459, + "learning_rate": 4.739246288542063e-07, + "loss": 1.3347, + "step": 2687 + }, + { + "epoch": 2.761171032357473, + "grad_norm": 2.728793144226074, + "learning_rate": 4.720213170917397e-07, + "loss": 1.2981, + "step": 2688 + }, + { + "epoch": 2.7621982537236773, + "grad_norm": 2.6515908241271973, + "learning_rate": 4.7011800532927293e-07, + "loss": 1.3758, + "step": 2689 + }, + { + "epoch": 2.763225475089882, + "grad_norm": 2.585145950317383, + "learning_rate": 4.682146935668063e-07, + "loss": 1.2745, + "step": 2690 + }, + { + "epoch": 2.763225475089882, + "eval_loss": 1.7617937326431274, + "eval_runtime": 38.0261, + "eval_samples_per_second": 26.298, + "eval_steps_per_second": 1.105, + "step": 2690 + }, + { + "epoch": 2.7642526964560865, + "grad_norm": 2.5894737243652344, + "learning_rate": 4.663113818043396e-07, + "loss": 1.3174, + "step": 2691 + }, + { + "epoch": 2.7652799178222907, + "grad_norm": 2.718618392944336, + "learning_rate": 4.6440807004187286e-07, + "loss": 1.3024, + "step": 2692 + }, + { + "epoch": 2.7663071391884952, + "grad_norm": 2.6753177642822266, + "learning_rate": 4.625047582794062e-07, + "loss": 1.3204, + "step": 2693 + }, + { + "epoch": 2.7673343605546994, + "grad_norm": 2.5795040130615234, + "learning_rate": 4.606014465169395e-07, + "loss": 1.2935, + "step": 2694 + }, + { + "epoch": 2.768361581920904, + "grad_norm": 2.5900628566741943, + "learning_rate": 4.5869813475447284e-07, + "loss": 1.2988, + "step": 2695 + }, + { + "epoch": 2.769388803287108, + "grad_norm": 2.565168619155884, + "learning_rate": 4.5679482299200615e-07, + "loss": 1.3121, + "step": 2696 + }, + { + "epoch": 2.7704160246533127, + "grad_norm": 2.666557788848877, + "learning_rate": 4.548915112295394e-07, + "loss": 1.3062, + "step": 2697 + }, + { + "epoch": 2.7714432460195173, + "grad_norm": 2.5423285961151123, + "learning_rate": 4.529881994670728e-07, + "loss": 1.2633, + "step": 2698 + }, + { + "epoch": 2.7724704673857214, + "grad_norm": 2.5645787715911865, + "learning_rate": 4.5108488770460603e-07, + "loss": 1.3067, + "step": 2699 + }, + { + "epoch": 2.773497688751926, + "grad_norm": 2.4956552982330322, + "learning_rate": 4.491815759421394e-07, + "loss": 1.3153, + "step": 2700 + }, + { + "epoch": 2.773497688751926, + "eval_loss": 1.7727458477020264, + "eval_runtime": 38.0396, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 2700 + } + ], + "logging_steps": 1, + "max_steps": 2919, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 8.416735405277184e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/zero_to_fp32.py b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2700/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/generation_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/latest b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/latest new file mode 100644 index 0000000000000000000000000000000000000000..97df0619d2d32c8aa7c5bb67e83bc29e11b103ba --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/latest @@ -0,0 +1 @@ +global_step2919 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/model.safetensors.index.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..45a8ede868fac67b0ed59ee65981e4073904e47f --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ee2439f84cae8e963031210af4a1bfe8b2f6409a8ddd3777cce58aa6093cfb4c +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..d687dd7fb6e5b5e688b9d28cf0c9771cb4f2a263 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d346e32d0b8bd9a890e601c3e7c937ef265318cbd21636f057f7029d9a3c877 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..a8c1db00306e15bc131c5ac15dfe155b1e524cac --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ad194513c2322339c1316c66c07c8ad570c978d5f11c7e90ba00d0122d915c7c +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..0b90f306fe66fbc13102c20b66b637b3ceffc696 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f20c9a33cde38b42c51433fbd8e28efd905aa489c3e371738e7e91bafa139af1 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/special_tokens_map.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/tokenizer_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0b782d4602b361f7577abb52b710a8c0e088588d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/tokenizer_config.json @@ -0,0 +1,2078 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128257": { + "content": "🅁", + "lstrip": true, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/trainer_state.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..74df34f9528ff8afbc5de3dbcf3161a13cf49625 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/trainer_state.json @@ -0,0 +1,22794 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.9984591679506933, + "eval_steps": 10, + "global_step": 2919, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7525, + "step": 1 + }, + { + "epoch": 0.002054442732408834, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7605, + "step": 2 + }, + { + "epoch": 0.0030816640986132513, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8008, + "step": 3 + }, + { + "epoch": 0.004108885464817668, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.809, + "step": 4 + }, + { + "epoch": 0.005136106831022085, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7841, + "step": 5 + }, + { + "epoch": 0.0061633281972265025, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7924, + "step": 6 + }, + { + "epoch": 0.007190549563430919, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7732, + "step": 7 + }, + { + "epoch": 0.008217770929635337, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7953, + "step": 8 + }, + { + "epoch": 0.009244992295839754, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7747, + "step": 9 + }, + { + "epoch": 0.01027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7909, + "step": 10 + }, + { + "epoch": 0.01027221366204417, + "eval_loss": 2.813098192214966, + "eval_runtime": 38.1187, + "eval_samples_per_second": 26.234, + "eval_steps_per_second": 1.102, + "step": 10 + }, + { + "epoch": 0.011299435028248588, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7592, + "step": 11 + }, + { + "epoch": 0.012326656394453005, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7394, + "step": 12 + }, + { + "epoch": 0.01335387776065742, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8111, + "step": 13 + }, + { + "epoch": 0.014381099126861838, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7786, + "step": 14 + }, + { + "epoch": 0.015408320493066256, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7711, + "step": 15 + }, + { + "epoch": 0.016435541859270673, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7622, + "step": 16 + }, + { + "epoch": 0.01746276322547509, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7291, + "step": 17 + }, + { + "epoch": 0.01848998459167951, + "grad_norm": 6.8249125480651855, + "learning_rate": 3.424657534246575e-08, + "loss": 2.7995, + "step": 18 + }, + { + "epoch": 0.019517205957883924, + "grad_norm": 5.581949710845947, + "learning_rate": 5.136986301369863e-08, + "loss": 2.5998, + "step": 19 + }, + { + "epoch": 0.02054442732408834, + "grad_norm": 3.0888617038726807, + "learning_rate": 6.84931506849315e-08, + "loss": 2.6031, + "step": 20 + }, + { + "epoch": 0.02054442732408834, + "eval_loss": 2.60642147064209, + "eval_runtime": 38.2226, + "eval_samples_per_second": 26.163, + "eval_steps_per_second": 1.099, + "step": 20 + }, + { + "epoch": 0.02157164869029276, + "grad_norm": 4.42941427230835, + "learning_rate": 8.561643835616439e-08, + "loss": 2.5816, + "step": 21 + }, + { + "epoch": 0.022598870056497175, + "grad_norm": 2.359813690185547, + "learning_rate": 1.0273972602739726e-07, + "loss": 2.5352, + "step": 22 + }, + { + "epoch": 0.02362609142270159, + "grad_norm": 2.9022927284240723, + "learning_rate": 1.1986301369863014e-07, + "loss": 2.4641, + "step": 23 + }, + { + "epoch": 0.02465331278890601, + "grad_norm": 1.6895997524261475, + "learning_rate": 1.36986301369863e-07, + "loss": 2.4582, + "step": 24 + }, + { + "epoch": 0.025680534155110426, + "grad_norm": 2.4555776119232178, + "learning_rate": 1.541095890410959e-07, + "loss": 2.4171, + "step": 25 + }, + { + "epoch": 0.02670775552131484, + "grad_norm": 1.9735779762268066, + "learning_rate": 1.7123287671232878e-07, + "loss": 2.4249, + "step": 26 + }, + { + "epoch": 0.02773497688751926, + "grad_norm": 2.71161150932312, + "learning_rate": 1.8835616438356165e-07, + "loss": 2.3791, + "step": 27 + }, + { + "epoch": 0.028762198253723677, + "grad_norm": 1.8482733964920044, + "learning_rate": 2.0547945205479452e-07, + "loss": 2.374, + "step": 28 + }, + { + "epoch": 0.029789419619928096, + "grad_norm": 2.191246747970581, + "learning_rate": 2.226027397260274e-07, + "loss": 2.3557, + "step": 29 + }, + { + "epoch": 0.030816640986132512, + "grad_norm": 2.0575971603393555, + "learning_rate": 2.397260273972603e-07, + "loss": 2.3315, + "step": 30 + }, + { + "epoch": 0.030816640986132512, + "eval_loss": 2.399639844894409, + "eval_runtime": 37.737, + "eval_samples_per_second": 26.499, + "eval_steps_per_second": 1.113, + "step": 30 + }, + { + "epoch": 0.03184386235233693, + "grad_norm": 2.0346758365631104, + "learning_rate": 2.568493150684932e-07, + "loss": 2.2791, + "step": 31 + }, + { + "epoch": 0.03287108371854135, + "grad_norm": 1.8710602521896362, + "learning_rate": 2.73972602739726e-07, + "loss": 2.3219, + "step": 32 + }, + { + "epoch": 0.03389830508474576, + "grad_norm": 1.9982675313949585, + "learning_rate": 2.910958904109589e-07, + "loss": 2.273, + "step": 33 + }, + { + "epoch": 0.03492552645095018, + "grad_norm": 2.092076301574707, + "learning_rate": 3.082191780821918e-07, + "loss": 2.2555, + "step": 34 + }, + { + "epoch": 0.035952747817154594, + "grad_norm": 1.870373249053955, + "learning_rate": 3.2534246575342466e-07, + "loss": 2.2314, + "step": 35 + }, + { + "epoch": 0.03697996918335902, + "grad_norm": 1.8378052711486816, + "learning_rate": 3.4246575342465755e-07, + "loss": 2.1244, + "step": 36 + }, + { + "epoch": 0.03800719054956343, + "grad_norm": 2.0156054496765137, + "learning_rate": 3.595890410958904e-07, + "loss": 2.2502, + "step": 37 + }, + { + "epoch": 0.03903441191576785, + "grad_norm": 1.6175312995910645, + "learning_rate": 3.767123287671233e-07, + "loss": 2.1566, + "step": 38 + }, + { + "epoch": 0.040061633281972264, + "grad_norm": 1.9051450490951538, + "learning_rate": 3.938356164383562e-07, + "loss": 2.167, + "step": 39 + }, + { + "epoch": 0.04108885464817668, + "grad_norm": 2.383063793182373, + "learning_rate": 4.1095890410958903e-07, + "loss": 2.2099, + "step": 40 + }, + { + "epoch": 0.04108885464817668, + "eval_loss": 2.2763354778289795, + "eval_runtime": 37.8085, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 40 + }, + { + "epoch": 0.042116076014381096, + "grad_norm": 2.1310720443725586, + "learning_rate": 4.2808219178082193e-07, + "loss": 2.1714, + "step": 41 + }, + { + "epoch": 0.04314329738058552, + "grad_norm": 1.771941900253296, + "learning_rate": 4.452054794520548e-07, + "loss": 2.0895, + "step": 42 + }, + { + "epoch": 0.044170518746789934, + "grad_norm": 2.3784172534942627, + "learning_rate": 4.6232876712328767e-07, + "loss": 2.1555, + "step": 43 + }, + { + "epoch": 0.04519774011299435, + "grad_norm": 1.6692813634872437, + "learning_rate": 4.794520547945206e-07, + "loss": 2.11, + "step": 44 + }, + { + "epoch": 0.046224961479198766, + "grad_norm": 2.1780643463134766, + "learning_rate": 4.965753424657534e-07, + "loss": 2.1256, + "step": 45 + }, + { + "epoch": 0.04725218284540318, + "grad_norm": 1.9390568733215332, + "learning_rate": 5.136986301369864e-07, + "loss": 2.0971, + "step": 46 + }, + { + "epoch": 0.048279404211607604, + "grad_norm": 2.276216506958008, + "learning_rate": 5.308219178082192e-07, + "loss": 2.1479, + "step": 47 + }, + { + "epoch": 0.04930662557781202, + "grad_norm": 2.148573160171509, + "learning_rate": 5.47945205479452e-07, + "loss": 2.0808, + "step": 48 + }, + { + "epoch": 0.050333846944016436, + "grad_norm": 2.259422540664673, + "learning_rate": 5.65068493150685e-07, + "loss": 2.0823, + "step": 49 + }, + { + "epoch": 0.05136106831022085, + "grad_norm": 1.7703359127044678, + "learning_rate": 5.821917808219178e-07, + "loss": 2.0236, + "step": 50 + }, + { + "epoch": 0.05136106831022085, + "eval_loss": 2.2037274837493896, + "eval_runtime": 37.8385, + "eval_samples_per_second": 26.428, + "eval_steps_per_second": 1.11, + "step": 50 + }, + { + "epoch": 0.05238828967642527, + "grad_norm": 2.56268310546875, + "learning_rate": 5.993150684931507e-07, + "loss": 2.0228, + "step": 51 + }, + { + "epoch": 0.05341551104262968, + "grad_norm": 2.5256848335266113, + "learning_rate": 6.164383561643836e-07, + "loss": 2.1401, + "step": 52 + }, + { + "epoch": 0.054442732408834106, + "grad_norm": 2.6329638957977295, + "learning_rate": 6.335616438356165e-07, + "loss": 2.0107, + "step": 53 + }, + { + "epoch": 0.05546995377503852, + "grad_norm": 2.477762222290039, + "learning_rate": 6.506849315068493e-07, + "loss": 2.0648, + "step": 54 + }, + { + "epoch": 0.05649717514124294, + "grad_norm": 1.8869198560714722, + "learning_rate": 6.678082191780823e-07, + "loss": 2.0214, + "step": 55 + }, + { + "epoch": 0.05752439650744735, + "grad_norm": 2.707345485687256, + "learning_rate": 6.849315068493151e-07, + "loss": 1.9893, + "step": 56 + }, + { + "epoch": 0.05855161787365177, + "grad_norm": 2.067230701446533, + "learning_rate": 7.020547945205481e-07, + "loss": 2.0308, + "step": 57 + }, + { + "epoch": 0.05957883923985619, + "grad_norm": 2.1961982250213623, + "learning_rate": 7.191780821917808e-07, + "loss": 1.948, + "step": 58 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 2.1032280921936035, + "learning_rate": 7.363013698630137e-07, + "loss": 2.0098, + "step": 59 + }, + { + "epoch": 0.061633281972265024, + "grad_norm": 2.5191144943237305, + "learning_rate": 7.534246575342466e-07, + "loss": 2.0183, + "step": 60 + }, + { + "epoch": 0.061633281972265024, + "eval_loss": 2.1474344730377197, + "eval_runtime": 37.8631, + "eval_samples_per_second": 26.411, + "eval_steps_per_second": 1.109, + "step": 60 + }, + { + "epoch": 0.06266050333846944, + "grad_norm": 2.177739143371582, + "learning_rate": 7.705479452054795e-07, + "loss": 1.9702, + "step": 61 + }, + { + "epoch": 0.06368772470467386, + "grad_norm": 2.499784231185913, + "learning_rate": 7.876712328767124e-07, + "loss": 1.9825, + "step": 62 + }, + { + "epoch": 0.06471494607087827, + "grad_norm": 2.181591749191284, + "learning_rate": 8.047945205479453e-07, + "loss": 2.0414, + "step": 63 + }, + { + "epoch": 0.0657421674370827, + "grad_norm": 1.9575999975204468, + "learning_rate": 8.219178082191781e-07, + "loss": 2.0115, + "step": 64 + }, + { + "epoch": 0.0667693888032871, + "grad_norm": 2.3431336879730225, + "learning_rate": 8.39041095890411e-07, + "loss": 1.8781, + "step": 65 + }, + { + "epoch": 0.06779661016949153, + "grad_norm": 2.042747974395752, + "learning_rate": 8.561643835616439e-07, + "loss": 1.9798, + "step": 66 + }, + { + "epoch": 0.06882383153569595, + "grad_norm": 2.784104585647583, + "learning_rate": 8.732876712328768e-07, + "loss": 1.9274, + "step": 67 + }, + { + "epoch": 0.06985105290190036, + "grad_norm": 1.9781445264816284, + "learning_rate": 8.904109589041097e-07, + "loss": 2.0036, + "step": 68 + }, + { + "epoch": 0.07087827426810478, + "grad_norm": 2.631537437438965, + "learning_rate": 9.075342465753426e-07, + "loss": 1.938, + "step": 69 + }, + { + "epoch": 0.07190549563430919, + "grad_norm": 1.902701497077942, + "learning_rate": 9.246575342465753e-07, + "loss": 2.0545, + "step": 70 + }, + { + "epoch": 0.07190549563430919, + "eval_loss": 2.115269422531128, + "eval_runtime": 37.887, + "eval_samples_per_second": 26.394, + "eval_steps_per_second": 1.109, + "step": 70 + }, + { + "epoch": 0.07293271700051361, + "grad_norm": 1.9323904514312744, + "learning_rate": 9.417808219178083e-07, + "loss": 1.9463, + "step": 71 + }, + { + "epoch": 0.07395993836671803, + "grad_norm": 2.1631956100463867, + "learning_rate": 9.589041095890411e-07, + "loss": 1.8982, + "step": 72 + }, + { + "epoch": 0.07498715973292244, + "grad_norm": 1.959518313407898, + "learning_rate": 9.76027397260274e-07, + "loss": 1.9538, + "step": 73 + }, + { + "epoch": 0.07601438109912687, + "grad_norm": 1.9834569692611694, + "learning_rate": 9.931506849315068e-07, + "loss": 1.9257, + "step": 74 + }, + { + "epoch": 0.07704160246533127, + "grad_norm": 2.2818732261657715, + "learning_rate": 1.0102739726027399e-06, + "loss": 1.9318, + "step": 75 + }, + { + "epoch": 0.0780688238315357, + "grad_norm": 1.7696406841278076, + "learning_rate": 1.0273972602739727e-06, + "loss": 1.912, + "step": 76 + }, + { + "epoch": 0.07909604519774012, + "grad_norm": 2.396303176879883, + "learning_rate": 1.0445205479452056e-06, + "loss": 1.9482, + "step": 77 + }, + { + "epoch": 0.08012326656394453, + "grad_norm": 2.2371881008148193, + "learning_rate": 1.0616438356164384e-06, + "loss": 1.9448, + "step": 78 + }, + { + "epoch": 0.08115048793014895, + "grad_norm": 2.40077543258667, + "learning_rate": 1.0787671232876712e-06, + "loss": 1.8764, + "step": 79 + }, + { + "epoch": 0.08217770929635336, + "grad_norm": 2.267916679382324, + "learning_rate": 1.095890410958904e-06, + "loss": 1.9478, + "step": 80 + }, + { + "epoch": 0.08217770929635336, + "eval_loss": 2.0851893424987793, + "eval_runtime": 37.9176, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 80 + }, + { + "epoch": 0.08320493066255778, + "grad_norm": 2.286259651184082, + "learning_rate": 1.1130136986301371e-06, + "loss": 1.8856, + "step": 81 + }, + { + "epoch": 0.08423215202876219, + "grad_norm": 2.160848379135132, + "learning_rate": 1.13013698630137e-06, + "loss": 1.9196, + "step": 82 + }, + { + "epoch": 0.08525937339496661, + "grad_norm": 2.3618931770324707, + "learning_rate": 1.1472602739726028e-06, + "loss": 1.9229, + "step": 83 + }, + { + "epoch": 0.08628659476117104, + "grad_norm": 2.1963961124420166, + "learning_rate": 1.1643835616438357e-06, + "loss": 1.8998, + "step": 84 + }, + { + "epoch": 0.08731381612737545, + "grad_norm": 2.197981357574463, + "learning_rate": 1.1815068493150685e-06, + "loss": 1.8716, + "step": 85 + }, + { + "epoch": 0.08834103749357987, + "grad_norm": 2.3562612533569336, + "learning_rate": 1.1986301369863014e-06, + "loss": 1.9336, + "step": 86 + }, + { + "epoch": 0.08936825885978428, + "grad_norm": 2.49712872505188, + "learning_rate": 1.2157534246575344e-06, + "loss": 1.8827, + "step": 87 + }, + { + "epoch": 0.0903954802259887, + "grad_norm": 2.315244436264038, + "learning_rate": 1.2328767123287673e-06, + "loss": 1.8801, + "step": 88 + }, + { + "epoch": 0.09142270159219312, + "grad_norm": 2.774587869644165, + "learning_rate": 1.25e-06, + "loss": 1.8972, + "step": 89 + }, + { + "epoch": 0.09244992295839753, + "grad_norm": 2.262364387512207, + "learning_rate": 1.267123287671233e-06, + "loss": 1.8426, + "step": 90 + }, + { + "epoch": 0.09244992295839753, + "eval_loss": 2.069462299346924, + "eval_runtime": 37.9319, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 90 + }, + { + "epoch": 0.09347714432460195, + "grad_norm": 2.6195597648620605, + "learning_rate": 1.284246575342466e-06, + "loss": 1.871, + "step": 91 + }, + { + "epoch": 0.09450436569080636, + "grad_norm": 2.051847219467163, + "learning_rate": 1.3013698630136986e-06, + "loss": 1.8922, + "step": 92 + }, + { + "epoch": 0.09553158705701079, + "grad_norm": 2.5930886268615723, + "learning_rate": 1.3184931506849317e-06, + "loss": 1.8818, + "step": 93 + }, + { + "epoch": 0.09655880842321521, + "grad_norm": 2.304164171218872, + "learning_rate": 1.3356164383561645e-06, + "loss": 1.8779, + "step": 94 + }, + { + "epoch": 0.09758602978941962, + "grad_norm": 2.187354564666748, + "learning_rate": 1.3527397260273976e-06, + "loss": 1.9272, + "step": 95 + }, + { + "epoch": 0.09861325115562404, + "grad_norm": 2.2492458820343018, + "learning_rate": 1.3698630136986302e-06, + "loss": 1.805, + "step": 96 + }, + { + "epoch": 0.09964047252182845, + "grad_norm": 2.3325352668762207, + "learning_rate": 1.386986301369863e-06, + "loss": 1.8817, + "step": 97 + }, + { + "epoch": 0.10066769388803287, + "grad_norm": 2.1708617210388184, + "learning_rate": 1.4041095890410961e-06, + "loss": 1.7956, + "step": 98 + }, + { + "epoch": 0.1016949152542373, + "grad_norm": 2.4267706871032715, + "learning_rate": 1.421232876712329e-06, + "loss": 1.8116, + "step": 99 + }, + { + "epoch": 0.1027221366204417, + "grad_norm": 2.3286805152893066, + "learning_rate": 1.4383561643835616e-06, + "loss": 1.8299, + "step": 100 + }, + { + "epoch": 0.1027221366204417, + "eval_loss": 2.0549135208129883, + "eval_runtime": 37.935, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 100 + }, + { + "epoch": 0.10374935798664613, + "grad_norm": 2.5391643047332764, + "learning_rate": 1.4554794520547946e-06, + "loss": 1.8521, + "step": 101 + }, + { + "epoch": 0.10477657935285054, + "grad_norm": 2.018380880355835, + "learning_rate": 1.4726027397260275e-06, + "loss": 1.8209, + "step": 102 + }, + { + "epoch": 0.10580380071905496, + "grad_norm": 2.3706746101379395, + "learning_rate": 1.4897260273972605e-06, + "loss": 1.79, + "step": 103 + }, + { + "epoch": 0.10683102208525937, + "grad_norm": 2.1290764808654785, + "learning_rate": 1.5068493150684932e-06, + "loss": 1.8141, + "step": 104 + }, + { + "epoch": 0.10785824345146379, + "grad_norm": 2.2768120765686035, + "learning_rate": 1.5239726027397262e-06, + "loss": 1.8826, + "step": 105 + }, + { + "epoch": 0.10888546481766821, + "grad_norm": 2.316584348678589, + "learning_rate": 1.541095890410959e-06, + "loss": 1.8604, + "step": 106 + }, + { + "epoch": 0.10991268618387262, + "grad_norm": 2.14679217338562, + "learning_rate": 1.5582191780821921e-06, + "loss": 1.865, + "step": 107 + }, + { + "epoch": 0.11093990755007704, + "grad_norm": 2.3288021087646484, + "learning_rate": 1.5753424657534248e-06, + "loss": 1.8512, + "step": 108 + }, + { + "epoch": 0.11196712891628145, + "grad_norm": 2.300482988357544, + "learning_rate": 1.5924657534246576e-06, + "loss": 1.7455, + "step": 109 + }, + { + "epoch": 0.11299435028248588, + "grad_norm": 2.4226229190826416, + "learning_rate": 1.6095890410958907e-06, + "loss": 1.8058, + "step": 110 + }, + { + "epoch": 0.11299435028248588, + "eval_loss": 2.025261640548706, + "eval_runtime": 37.6779, + "eval_samples_per_second": 26.541, + "eval_steps_per_second": 1.115, + "step": 110 + }, + { + "epoch": 0.1140215716486903, + "grad_norm": 2.0930111408233643, + "learning_rate": 1.6267123287671235e-06, + "loss": 1.7375, + "step": 111 + }, + { + "epoch": 0.1150487930148947, + "grad_norm": 2.4973645210266113, + "learning_rate": 1.6438356164383561e-06, + "loss": 1.8268, + "step": 112 + }, + { + "epoch": 0.11607601438109913, + "grad_norm": 2.3295114040374756, + "learning_rate": 1.6609589041095892e-06, + "loss": 1.7882, + "step": 113 + }, + { + "epoch": 0.11710323574730354, + "grad_norm": 2.3693318367004395, + "learning_rate": 1.678082191780822e-06, + "loss": 1.8192, + "step": 114 + }, + { + "epoch": 0.11813045711350796, + "grad_norm": 2.553637742996216, + "learning_rate": 1.695205479452055e-06, + "loss": 1.7491, + "step": 115 + }, + { + "epoch": 0.11915767847971238, + "grad_norm": 2.303945541381836, + "learning_rate": 1.7123287671232877e-06, + "loss": 1.8737, + "step": 116 + }, + { + "epoch": 0.12018489984591679, + "grad_norm": 2.250356435775757, + "learning_rate": 1.7294520547945206e-06, + "loss": 1.8039, + "step": 117 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 2.3854403495788574, + "learning_rate": 1.7465753424657536e-06, + "loss": 1.8357, + "step": 118 + }, + { + "epoch": 0.12223934257832562, + "grad_norm": 2.4049572944641113, + "learning_rate": 1.7636986301369865e-06, + "loss": 1.8209, + "step": 119 + }, + { + "epoch": 0.12326656394453005, + "grad_norm": 2.313551425933838, + "learning_rate": 1.7808219178082193e-06, + "loss": 1.9026, + "step": 120 + }, + { + "epoch": 0.12326656394453005, + "eval_loss": 2.012171745300293, + "eval_runtime": 37.7908, + "eval_samples_per_second": 26.461, + "eval_steps_per_second": 1.111, + "step": 120 + }, + { + "epoch": 0.12429378531073447, + "grad_norm": 2.518320322036743, + "learning_rate": 1.7979452054794521e-06, + "loss": 1.8241, + "step": 121 + }, + { + "epoch": 0.12532100667693888, + "grad_norm": 2.6504592895507812, + "learning_rate": 1.8150684931506852e-06, + "loss": 1.8091, + "step": 122 + }, + { + "epoch": 0.1263482280431433, + "grad_norm": 2.556610107421875, + "learning_rate": 1.832191780821918e-06, + "loss": 1.7547, + "step": 123 + }, + { + "epoch": 0.12737544940934772, + "grad_norm": 2.247633695602417, + "learning_rate": 1.8493150684931507e-06, + "loss": 1.7923, + "step": 124 + }, + { + "epoch": 0.12840267077555212, + "grad_norm": 2.195596933364868, + "learning_rate": 1.8664383561643837e-06, + "loss": 1.747, + "step": 125 + }, + { + "epoch": 0.12942989214175654, + "grad_norm": 2.121396780014038, + "learning_rate": 1.8835616438356166e-06, + "loss": 1.8088, + "step": 126 + }, + { + "epoch": 0.13045711350796096, + "grad_norm": 2.4572501182556152, + "learning_rate": 1.9006849315068496e-06, + "loss": 1.7285, + "step": 127 + }, + { + "epoch": 0.1314843348741654, + "grad_norm": 2.2590177059173584, + "learning_rate": 1.9178082191780823e-06, + "loss": 1.7929, + "step": 128 + }, + { + "epoch": 0.1325115562403698, + "grad_norm": 2.5099234580993652, + "learning_rate": 1.9349315068493153e-06, + "loss": 1.8365, + "step": 129 + }, + { + "epoch": 0.1335387776065742, + "grad_norm": 2.3064465522766113, + "learning_rate": 1.952054794520548e-06, + "loss": 1.8351, + "step": 130 + }, + { + "epoch": 0.1335387776065742, + "eval_loss": 2.0043866634368896, + "eval_runtime": 37.8661, + "eval_samples_per_second": 26.409, + "eval_steps_per_second": 1.109, + "step": 130 + }, + { + "epoch": 0.13456599897277863, + "grad_norm": 2.5977718830108643, + "learning_rate": 1.969178082191781e-06, + "loss": 1.7603, + "step": 131 + }, + { + "epoch": 0.13559322033898305, + "grad_norm": 2.23221755027771, + "learning_rate": 1.9863013698630136e-06, + "loss": 1.8237, + "step": 132 + }, + { + "epoch": 0.13662044170518747, + "grad_norm": 2.247044324874878, + "learning_rate": 2.0034246575342467e-06, + "loss": 1.8006, + "step": 133 + }, + { + "epoch": 0.1376476630713919, + "grad_norm": 2.1796107292175293, + "learning_rate": 2.0205479452054797e-06, + "loss": 1.7354, + "step": 134 + }, + { + "epoch": 0.1386748844375963, + "grad_norm": 2.3269693851470947, + "learning_rate": 2.037671232876713e-06, + "loss": 1.783, + "step": 135 + }, + { + "epoch": 0.1397021058038007, + "grad_norm": 2.2903053760528564, + "learning_rate": 2.0547945205479454e-06, + "loss": 1.8072, + "step": 136 + }, + { + "epoch": 0.14072932717000514, + "grad_norm": 2.634150981903076, + "learning_rate": 2.071917808219178e-06, + "loss": 1.7251, + "step": 137 + }, + { + "epoch": 0.14175654853620956, + "grad_norm": 2.42579984664917, + "learning_rate": 2.089041095890411e-06, + "loss": 1.7997, + "step": 138 + }, + { + "epoch": 0.14278376990241398, + "grad_norm": 2.615828514099121, + "learning_rate": 2.106164383561644e-06, + "loss": 1.766, + "step": 139 + }, + { + "epoch": 0.14381099126861838, + "grad_norm": 2.312714099884033, + "learning_rate": 2.123287671232877e-06, + "loss": 1.7538, + "step": 140 + }, + { + "epoch": 0.14381099126861838, + "eval_loss": 1.9918214082717896, + "eval_runtime": 37.8698, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 140 + }, + { + "epoch": 0.1448382126348228, + "grad_norm": 2.7540030479431152, + "learning_rate": 2.14041095890411e-06, + "loss": 1.7856, + "step": 141 + }, + { + "epoch": 0.14586543400102722, + "grad_norm": 2.289191246032715, + "learning_rate": 2.1575342465753425e-06, + "loss": 1.7288, + "step": 142 + }, + { + "epoch": 0.14689265536723164, + "grad_norm": 2.788001775741577, + "learning_rate": 2.1746575342465755e-06, + "loss": 1.8703, + "step": 143 + }, + { + "epoch": 0.14791987673343607, + "grad_norm": 2.4128854274749756, + "learning_rate": 2.191780821917808e-06, + "loss": 1.7461, + "step": 144 + }, + { + "epoch": 0.14894709809964046, + "grad_norm": 2.4108099937438965, + "learning_rate": 2.2089041095890412e-06, + "loss": 1.73, + "step": 145 + }, + { + "epoch": 0.14997431946584489, + "grad_norm": 2.4315948486328125, + "learning_rate": 2.2260273972602743e-06, + "loss": 1.7303, + "step": 146 + }, + { + "epoch": 0.1510015408320493, + "grad_norm": 2.494056224822998, + "learning_rate": 2.243150684931507e-06, + "loss": 1.7616, + "step": 147 + }, + { + "epoch": 0.15202876219825373, + "grad_norm": 2.4557394981384277, + "learning_rate": 2.26027397260274e-06, + "loss": 1.7655, + "step": 148 + }, + { + "epoch": 0.15305598356445815, + "grad_norm": 2.4114091396331787, + "learning_rate": 2.2773972602739726e-06, + "loss": 1.8538, + "step": 149 + }, + { + "epoch": 0.15408320493066255, + "grad_norm": 2.609219789505005, + "learning_rate": 2.2945205479452057e-06, + "loss": 1.7767, + "step": 150 + }, + { + "epoch": 0.15408320493066255, + "eval_loss": 1.981367588043213, + "eval_runtime": 37.9562, + "eval_samples_per_second": 26.346, + "eval_steps_per_second": 1.107, + "step": 150 + }, + { + "epoch": 0.15511042629686697, + "grad_norm": 2.213622808456421, + "learning_rate": 2.3116438356164387e-06, + "loss": 1.687, + "step": 151 + }, + { + "epoch": 0.1561376476630714, + "grad_norm": 2.6517975330352783, + "learning_rate": 2.3287671232876713e-06, + "loss": 1.7311, + "step": 152 + }, + { + "epoch": 0.15716486902927582, + "grad_norm": 2.5174458026885986, + "learning_rate": 2.3458904109589044e-06, + "loss": 1.7835, + "step": 153 + }, + { + "epoch": 0.15819209039548024, + "grad_norm": 2.485480308532715, + "learning_rate": 2.363013698630137e-06, + "loss": 1.7749, + "step": 154 + }, + { + "epoch": 0.15921931176168463, + "grad_norm": 2.5386061668395996, + "learning_rate": 2.38013698630137e-06, + "loss": 1.7223, + "step": 155 + }, + { + "epoch": 0.16024653312788906, + "grad_norm": 2.2895941734313965, + "learning_rate": 2.3972602739726027e-06, + "loss": 1.6845, + "step": 156 + }, + { + "epoch": 0.16127375449409348, + "grad_norm": 2.5453789234161377, + "learning_rate": 2.4143835616438358e-06, + "loss": 1.7627, + "step": 157 + }, + { + "epoch": 0.1623009758602979, + "grad_norm": 2.5322303771972656, + "learning_rate": 2.431506849315069e-06, + "loss": 1.764, + "step": 158 + }, + { + "epoch": 0.1633281972265023, + "grad_norm": 2.2669899463653564, + "learning_rate": 2.4486301369863015e-06, + "loss": 1.7487, + "step": 159 + }, + { + "epoch": 0.16435541859270672, + "grad_norm": 2.4577853679656982, + "learning_rate": 2.4657534246575345e-06, + "loss": 1.7635, + "step": 160 + }, + { + "epoch": 0.16435541859270672, + "eval_loss": 1.9793355464935303, + "eval_runtime": 37.8576, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 160 + }, + { + "epoch": 0.16538263995891114, + "grad_norm": 2.354893922805786, + "learning_rate": 2.482876712328767e-06, + "loss": 1.7538, + "step": 161 + }, + { + "epoch": 0.16640986132511557, + "grad_norm": 2.3255856037139893, + "learning_rate": 2.5e-06, + "loss": 1.7277, + "step": 162 + }, + { + "epoch": 0.16743708269132, + "grad_norm": 2.3501994609832764, + "learning_rate": 2.5171232876712333e-06, + "loss": 1.6985, + "step": 163 + }, + { + "epoch": 0.16846430405752438, + "grad_norm": 2.386627674102783, + "learning_rate": 2.534246575342466e-06, + "loss": 1.7552, + "step": 164 + }, + { + "epoch": 0.1694915254237288, + "grad_norm": 2.2238457202911377, + "learning_rate": 2.551369863013699e-06, + "loss": 1.7805, + "step": 165 + }, + { + "epoch": 0.17051874678993323, + "grad_norm": 2.2653403282165527, + "learning_rate": 2.568493150684932e-06, + "loss": 1.7656, + "step": 166 + }, + { + "epoch": 0.17154596815613765, + "grad_norm": 2.233132839202881, + "learning_rate": 2.585616438356164e-06, + "loss": 1.6997, + "step": 167 + }, + { + "epoch": 0.17257318952234207, + "grad_norm": 2.1696043014526367, + "learning_rate": 2.6027397260273973e-06, + "loss": 1.7487, + "step": 168 + }, + { + "epoch": 0.17360041088854647, + "grad_norm": 2.4479899406433105, + "learning_rate": 2.6198630136986303e-06, + "loss": 1.7957, + "step": 169 + }, + { + "epoch": 0.1746276322547509, + "grad_norm": 2.3054702281951904, + "learning_rate": 2.6369863013698634e-06, + "loss": 1.766, + "step": 170 + }, + { + "epoch": 0.1746276322547509, + "eval_loss": 1.9543790817260742, + "eval_runtime": 37.8837, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 170 + }, + { + "epoch": 0.17565485362095531, + "grad_norm": 2.3591110706329346, + "learning_rate": 2.654109589041096e-06, + "loss": 1.7149, + "step": 171 + }, + { + "epoch": 0.17668207498715974, + "grad_norm": 2.450526237487793, + "learning_rate": 2.671232876712329e-06, + "loss": 1.7074, + "step": 172 + }, + { + "epoch": 0.17770929635336416, + "grad_norm": 2.1511754989624023, + "learning_rate": 2.688356164383562e-06, + "loss": 1.7036, + "step": 173 + }, + { + "epoch": 0.17873651771956856, + "grad_norm": 2.5944275856018066, + "learning_rate": 2.705479452054795e-06, + "loss": 1.6566, + "step": 174 + }, + { + "epoch": 0.17976373908577298, + "grad_norm": 2.4301369190216064, + "learning_rate": 2.7226027397260274e-06, + "loss": 1.7977, + "step": 175 + }, + { + "epoch": 0.1807909604519774, + "grad_norm": 2.287665367126465, + "learning_rate": 2.7397260273972604e-06, + "loss": 1.7526, + "step": 176 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.380032777786255, + "learning_rate": 2.7568493150684935e-06, + "loss": 1.6753, + "step": 177 + }, + { + "epoch": 0.18284540318438625, + "grad_norm": 2.4040510654449463, + "learning_rate": 2.773972602739726e-06, + "loss": 1.758, + "step": 178 + }, + { + "epoch": 0.18387262455059064, + "grad_norm": 2.5074045658111572, + "learning_rate": 2.791095890410959e-06, + "loss": 1.6498, + "step": 179 + }, + { + "epoch": 0.18489984591679506, + "grad_norm": 2.3926949501037598, + "learning_rate": 2.8082191780821922e-06, + "loss": 1.8006, + "step": 180 + }, + { + "epoch": 0.18489984591679506, + "eval_loss": 1.9446762800216675, + "eval_runtime": 37.8572, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 180 + }, + { + "epoch": 0.1859270672829995, + "grad_norm": 2.4991633892059326, + "learning_rate": 2.8253424657534253e-06, + "loss": 1.7649, + "step": 181 + }, + { + "epoch": 0.1869542886492039, + "grad_norm": 2.2558324337005615, + "learning_rate": 2.842465753424658e-06, + "loss": 1.7743, + "step": 182 + }, + { + "epoch": 0.18798151001540833, + "grad_norm": 2.219717025756836, + "learning_rate": 2.8595890410958905e-06, + "loss": 1.7807, + "step": 183 + }, + { + "epoch": 0.18900873138161273, + "grad_norm": 2.406562566757202, + "learning_rate": 2.876712328767123e-06, + "loss": 1.6289, + "step": 184 + }, + { + "epoch": 0.19003595274781715, + "grad_norm": 2.356579303741455, + "learning_rate": 2.8938356164383562e-06, + "loss": 1.6747, + "step": 185 + }, + { + "epoch": 0.19106317411402157, + "grad_norm": 2.4844322204589844, + "learning_rate": 2.9109589041095893e-06, + "loss": 1.6731, + "step": 186 + }, + { + "epoch": 0.192090395480226, + "grad_norm": 2.8179149627685547, + "learning_rate": 2.9280821917808223e-06, + "loss": 1.6786, + "step": 187 + }, + { + "epoch": 0.19311761684643042, + "grad_norm": 2.395416259765625, + "learning_rate": 2.945205479452055e-06, + "loss": 1.7313, + "step": 188 + }, + { + "epoch": 0.1941448382126348, + "grad_norm": 2.848740816116333, + "learning_rate": 2.962328767123288e-06, + "loss": 1.7258, + "step": 189 + }, + { + "epoch": 0.19517205957883924, + "grad_norm": 2.427395820617676, + "learning_rate": 2.979452054794521e-06, + "loss": 1.6633, + "step": 190 + }, + { + "epoch": 0.19517205957883924, + "eval_loss": 1.9443999528884888, + "eval_runtime": 37.8579, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 190 + }, + { + "epoch": 0.19619928094504366, + "grad_norm": 2.5347256660461426, + "learning_rate": 2.9965753424657533e-06, + "loss": 1.7342, + "step": 191 + }, + { + "epoch": 0.19722650231124808, + "grad_norm": 2.8696866035461426, + "learning_rate": 3.0136986301369864e-06, + "loss": 1.7037, + "step": 192 + }, + { + "epoch": 0.1982537236774525, + "grad_norm": 2.427468776702881, + "learning_rate": 3.0308219178082194e-06, + "loss": 1.672, + "step": 193 + }, + { + "epoch": 0.1992809450436569, + "grad_norm": 2.4253737926483154, + "learning_rate": 3.0479452054794525e-06, + "loss": 1.7252, + "step": 194 + }, + { + "epoch": 0.20030816640986132, + "grad_norm": 2.801668643951416, + "learning_rate": 3.065068493150685e-06, + "loss": 1.6694, + "step": 195 + }, + { + "epoch": 0.20133538777606574, + "grad_norm": 2.442263603210449, + "learning_rate": 3.082191780821918e-06, + "loss": 1.6866, + "step": 196 + }, + { + "epoch": 0.20236260914227017, + "grad_norm": 2.777463674545288, + "learning_rate": 3.099315068493151e-06, + "loss": 1.686, + "step": 197 + }, + { + "epoch": 0.2033898305084746, + "grad_norm": 2.491227388381958, + "learning_rate": 3.1164383561643843e-06, + "loss": 1.6414, + "step": 198 + }, + { + "epoch": 0.20441705187467898, + "grad_norm": 2.33585262298584, + "learning_rate": 3.1335616438356165e-06, + "loss": 1.7808, + "step": 199 + }, + { + "epoch": 0.2054442732408834, + "grad_norm": 2.518160581588745, + "learning_rate": 3.1506849315068495e-06, + "loss": 1.6592, + "step": 200 + }, + { + "epoch": 0.2054442732408834, + "eval_loss": 1.9371097087860107, + "eval_runtime": 37.8602, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 200 + }, + { + "epoch": 0.20647149460708783, + "grad_norm": 2.460204839706421, + "learning_rate": 3.167808219178082e-06, + "loss": 1.6668, + "step": 201 + }, + { + "epoch": 0.20749871597329225, + "grad_norm": 2.377824068069458, + "learning_rate": 3.184931506849315e-06, + "loss": 1.6462, + "step": 202 + }, + { + "epoch": 0.20852593733949668, + "grad_norm": 2.426177740097046, + "learning_rate": 3.2020547945205483e-06, + "loss": 1.6272, + "step": 203 + }, + { + "epoch": 0.20955315870570107, + "grad_norm": 2.4532034397125244, + "learning_rate": 3.2191780821917813e-06, + "loss": 1.6169, + "step": 204 + }, + { + "epoch": 0.2105803800719055, + "grad_norm": 2.517059564590454, + "learning_rate": 3.236301369863014e-06, + "loss": 1.5991, + "step": 205 + }, + { + "epoch": 0.21160760143810992, + "grad_norm": 2.22658109664917, + "learning_rate": 3.253424657534247e-06, + "loss": 1.6754, + "step": 206 + }, + { + "epoch": 0.21263482280431434, + "grad_norm": 2.344799518585205, + "learning_rate": 3.2705479452054796e-06, + "loss": 1.6648, + "step": 207 + }, + { + "epoch": 0.21366204417051873, + "grad_norm": 2.331875801086426, + "learning_rate": 3.2876712328767123e-06, + "loss": 1.6112, + "step": 208 + }, + { + "epoch": 0.21468926553672316, + "grad_norm": 2.262343645095825, + "learning_rate": 3.3047945205479453e-06, + "loss": 1.6479, + "step": 209 + }, + { + "epoch": 0.21571648690292758, + "grad_norm": 2.282278537750244, + "learning_rate": 3.3219178082191784e-06, + "loss": 1.6753, + "step": 210 + }, + { + "epoch": 0.21571648690292758, + "eval_loss": 1.9338418245315552, + "eval_runtime": 37.6083, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 210 + }, + { + "epoch": 0.216743708269132, + "grad_norm": 2.578476905822754, + "learning_rate": 3.3390410958904114e-06, + "loss": 1.7043, + "step": 211 + }, + { + "epoch": 0.21777092963533642, + "grad_norm": 2.3118813037872314, + "learning_rate": 3.356164383561644e-06, + "loss": 1.7541, + "step": 212 + }, + { + "epoch": 0.21879815100154082, + "grad_norm": 2.2418692111968994, + "learning_rate": 3.373287671232877e-06, + "loss": 1.6886, + "step": 213 + }, + { + "epoch": 0.21982537236774524, + "grad_norm": 2.3583438396453857, + "learning_rate": 3.39041095890411e-06, + "loss": 1.7241, + "step": 214 + }, + { + "epoch": 0.22085259373394966, + "grad_norm": 2.345090866088867, + "learning_rate": 3.4075342465753424e-06, + "loss": 1.6194, + "step": 215 + }, + { + "epoch": 0.2218798151001541, + "grad_norm": 2.3996200561523438, + "learning_rate": 3.4246575342465754e-06, + "loss": 1.6889, + "step": 216 + }, + { + "epoch": 0.2229070364663585, + "grad_norm": 2.3552374839782715, + "learning_rate": 3.4417808219178085e-06, + "loss": 1.6713, + "step": 217 + }, + { + "epoch": 0.2239342578325629, + "grad_norm": 2.2653744220733643, + "learning_rate": 3.458904109589041e-06, + "loss": 1.673, + "step": 218 + }, + { + "epoch": 0.22496147919876733, + "grad_norm": 2.263288974761963, + "learning_rate": 3.476027397260274e-06, + "loss": 1.673, + "step": 219 + }, + { + "epoch": 0.22598870056497175, + "grad_norm": 2.3173880577087402, + "learning_rate": 3.4931506849315072e-06, + "loss": 1.6803, + "step": 220 + }, + { + "epoch": 0.22598870056497175, + "eval_loss": 1.9166476726531982, + "eval_runtime": 37.7659, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 220 + }, + { + "epoch": 0.22701592193117617, + "grad_norm": 2.4005839824676514, + "learning_rate": 3.5102739726027403e-06, + "loss": 1.6809, + "step": 221 + }, + { + "epoch": 0.2280431432973806, + "grad_norm": 2.5032525062561035, + "learning_rate": 3.527397260273973e-06, + "loss": 1.6411, + "step": 222 + }, + { + "epoch": 0.229070364663585, + "grad_norm": 2.533583879470825, + "learning_rate": 3.5445205479452056e-06, + "loss": 1.6608, + "step": 223 + }, + { + "epoch": 0.2300975860297894, + "grad_norm": 2.3575572967529297, + "learning_rate": 3.5616438356164386e-06, + "loss": 1.7236, + "step": 224 + }, + { + "epoch": 0.23112480739599384, + "grad_norm": 2.3408942222595215, + "learning_rate": 3.5787671232876712e-06, + "loss": 1.7101, + "step": 225 + }, + { + "epoch": 0.23215202876219826, + "grad_norm": 2.3819596767425537, + "learning_rate": 3.5958904109589043e-06, + "loss": 1.6414, + "step": 226 + }, + { + "epoch": 0.23317925012840268, + "grad_norm": 2.2874338626861572, + "learning_rate": 3.6130136986301373e-06, + "loss": 1.6767, + "step": 227 + }, + { + "epoch": 0.23420647149460708, + "grad_norm": 2.3523948192596436, + "learning_rate": 3.6301369863013704e-06, + "loss": 1.6932, + "step": 228 + }, + { + "epoch": 0.2352336928608115, + "grad_norm": 2.476128578186035, + "learning_rate": 3.647260273972603e-06, + "loss": 1.7125, + "step": 229 + }, + { + "epoch": 0.23626091422701592, + "grad_norm": 2.3841593265533447, + "learning_rate": 3.664383561643836e-06, + "loss": 1.6469, + "step": 230 + }, + { + "epoch": 0.23626091422701592, + "eval_loss": 1.925522804260254, + "eval_runtime": 37.8581, + "eval_samples_per_second": 26.414, + "eval_steps_per_second": 1.109, + "step": 230 + }, + { + "epoch": 0.23728813559322035, + "grad_norm": 2.4401519298553467, + "learning_rate": 3.6815068493150687e-06, + "loss": 1.6726, + "step": 231 + }, + { + "epoch": 0.23831535695942477, + "grad_norm": 2.329674005508423, + "learning_rate": 3.6986301369863014e-06, + "loss": 1.6349, + "step": 232 + }, + { + "epoch": 0.23934257832562916, + "grad_norm": 2.654290199279785, + "learning_rate": 3.7157534246575344e-06, + "loss": 1.7309, + "step": 233 + }, + { + "epoch": 0.24036979969183359, + "grad_norm": 2.1599373817443848, + "learning_rate": 3.7328767123287675e-06, + "loss": 1.6437, + "step": 234 + }, + { + "epoch": 0.241397021058038, + "grad_norm": 2.6170263290405273, + "learning_rate": 3.7500000000000005e-06, + "loss": 1.6862, + "step": 235 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 2.2836415767669678, + "learning_rate": 3.767123287671233e-06, + "loss": 1.7073, + "step": 236 + }, + { + "epoch": 0.24345146379044685, + "grad_norm": 2.3001456260681152, + "learning_rate": 3.784246575342466e-06, + "loss": 1.6455, + "step": 237 + }, + { + "epoch": 0.24447868515665125, + "grad_norm": 2.4549131393432617, + "learning_rate": 3.8013698630136993e-06, + "loss": 1.6566, + "step": 238 + }, + { + "epoch": 0.24550590652285567, + "grad_norm": 2.4387831687927246, + "learning_rate": 3.818493150684932e-06, + "loss": 1.627, + "step": 239 + }, + { + "epoch": 0.2465331278890601, + "grad_norm": 2.136474847793579, + "learning_rate": 3.8356164383561645e-06, + "loss": 1.7286, + "step": 240 + }, + { + "epoch": 0.2465331278890601, + "eval_loss": 1.9188542366027832, + "eval_runtime": 37.9079, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 240 + }, + { + "epoch": 0.24756034925526452, + "grad_norm": 2.430297374725342, + "learning_rate": 3.852739726027397e-06, + "loss": 1.7014, + "step": 241 + }, + { + "epoch": 0.24858757062146894, + "grad_norm": 2.302480697631836, + "learning_rate": 3.869863013698631e-06, + "loss": 1.6412, + "step": 242 + }, + { + "epoch": 0.24961479198767333, + "grad_norm": 2.2905426025390625, + "learning_rate": 3.886986301369863e-06, + "loss": 1.6445, + "step": 243 + }, + { + "epoch": 0.25064201335387776, + "grad_norm": 2.55385684967041, + "learning_rate": 3.904109589041096e-06, + "loss": 1.6265, + "step": 244 + }, + { + "epoch": 0.2516692347200822, + "grad_norm": 2.3053576946258545, + "learning_rate": 3.921232876712329e-06, + "loss": 1.6338, + "step": 245 + }, + { + "epoch": 0.2526964560862866, + "grad_norm": 2.4894707202911377, + "learning_rate": 3.938356164383562e-06, + "loss": 1.5966, + "step": 246 + }, + { + "epoch": 0.253723677452491, + "grad_norm": 2.6198930740356445, + "learning_rate": 3.9554794520547955e-06, + "loss": 1.6935, + "step": 247 + }, + { + "epoch": 0.25475089881869545, + "grad_norm": 2.3548998832702637, + "learning_rate": 3.972602739726027e-06, + "loss": 1.6227, + "step": 248 + }, + { + "epoch": 0.25577812018489987, + "grad_norm": 2.535460948944092, + "learning_rate": 3.989726027397261e-06, + "loss": 1.6699, + "step": 249 + }, + { + "epoch": 0.25680534155110424, + "grad_norm": 2.3680567741394043, + "learning_rate": 4.006849315068493e-06, + "loss": 1.6666, + "step": 250 + }, + { + "epoch": 0.25680534155110424, + "eval_loss": 1.9149967432022095, + "eval_runtime": 37.9423, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 250 + }, + { + "epoch": 0.25783256291730866, + "grad_norm": 2.3914358615875244, + "learning_rate": 4.023972602739726e-06, + "loss": 1.6633, + "step": 251 + }, + { + "epoch": 0.2588597842835131, + "grad_norm": 2.4293038845062256, + "learning_rate": 4.0410958904109595e-06, + "loss": 1.6573, + "step": 252 + }, + { + "epoch": 0.2598870056497175, + "grad_norm": 2.3409900665283203, + "learning_rate": 4.058219178082192e-06, + "loss": 1.6815, + "step": 253 + }, + { + "epoch": 0.26091422701592193, + "grad_norm": 2.5807456970214844, + "learning_rate": 4.075342465753426e-06, + "loss": 1.6412, + "step": 254 + }, + { + "epoch": 0.26194144838212635, + "grad_norm": 2.317861795425415, + "learning_rate": 4.092465753424658e-06, + "loss": 1.6311, + "step": 255 + }, + { + "epoch": 0.2629686697483308, + "grad_norm": 2.231682777404785, + "learning_rate": 4.109589041095891e-06, + "loss": 1.7021, + "step": 256 + }, + { + "epoch": 0.2639958911145352, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.1267123287671235e-06, + "loss": 1.6446, + "step": 257 + }, + { + "epoch": 0.2650231124807396, + "grad_norm": 2.2619071006774902, + "learning_rate": 4.143835616438356e-06, + "loss": 1.7101, + "step": 258 + }, + { + "epoch": 0.26605033384694404, + "grad_norm": 2.461945056915283, + "learning_rate": 4.16095890410959e-06, + "loss": 1.6901, + "step": 259 + }, + { + "epoch": 0.2670775552131484, + "grad_norm": 2.516348123550415, + "learning_rate": 4.178082191780822e-06, + "loss": 1.644, + "step": 260 + }, + { + "epoch": 0.2670775552131484, + "eval_loss": 1.904897928237915, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 260 + }, + { + "epoch": 0.26810477657935283, + "grad_norm": 2.2311813831329346, + "learning_rate": 4.195205479452055e-06, + "loss": 1.6301, + "step": 261 + }, + { + "epoch": 0.26913199794555726, + "grad_norm": 2.605109214782715, + "learning_rate": 4.212328767123288e-06, + "loss": 1.6393, + "step": 262 + }, + { + "epoch": 0.2701592193117617, + "grad_norm": 2.6103134155273438, + "learning_rate": 4.229452054794521e-06, + "loss": 1.6468, + "step": 263 + }, + { + "epoch": 0.2711864406779661, + "grad_norm": 2.5538523197174072, + "learning_rate": 4.246575342465754e-06, + "loss": 1.6294, + "step": 264 + }, + { + "epoch": 0.2722136620441705, + "grad_norm": 2.5841965675354004, + "learning_rate": 4.263698630136986e-06, + "loss": 1.5837, + "step": 265 + }, + { + "epoch": 0.27324088341037495, + "grad_norm": 2.4604523181915283, + "learning_rate": 4.28082191780822e-06, + "loss": 1.7473, + "step": 266 + }, + { + "epoch": 0.27426810477657937, + "grad_norm": 2.6701765060424805, + "learning_rate": 4.297945205479452e-06, + "loss": 1.6539, + "step": 267 + }, + { + "epoch": 0.2752953261427838, + "grad_norm": 2.360069990158081, + "learning_rate": 4.315068493150685e-06, + "loss": 1.6385, + "step": 268 + }, + { + "epoch": 0.2763225475089882, + "grad_norm": 2.4922797679901123, + "learning_rate": 4.3321917808219185e-06, + "loss": 1.6586, + "step": 269 + }, + { + "epoch": 0.2773497688751926, + "grad_norm": 2.5443379878997803, + "learning_rate": 4.349315068493151e-06, + "loss": 1.6318, + "step": 270 + }, + { + "epoch": 0.2773497688751926, + "eval_loss": 1.9062609672546387, + "eval_runtime": 37.92, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 270 + }, + { + "epoch": 0.278376990241397, + "grad_norm": 2.4064481258392334, + "learning_rate": 4.3664383561643846e-06, + "loss": 1.6178, + "step": 271 + }, + { + "epoch": 0.2794042116076014, + "grad_norm": 2.681637763977051, + "learning_rate": 4.383561643835616e-06, + "loss": 1.7071, + "step": 272 + }, + { + "epoch": 0.28043143297380585, + "grad_norm": 2.5805184841156006, + "learning_rate": 4.40068493150685e-06, + "loss": 1.6553, + "step": 273 + }, + { + "epoch": 0.2814586543400103, + "grad_norm": 2.4169883728027344, + "learning_rate": 4.4178082191780825e-06, + "loss": 1.6249, + "step": 274 + }, + { + "epoch": 0.2824858757062147, + "grad_norm": 2.557570457458496, + "learning_rate": 4.434931506849315e-06, + "loss": 1.5932, + "step": 275 + }, + { + "epoch": 0.2835130970724191, + "grad_norm": 2.3808107376098633, + "learning_rate": 4.4520547945205486e-06, + "loss": 1.6377, + "step": 276 + }, + { + "epoch": 0.28454031843862354, + "grad_norm": 2.402404546737671, + "learning_rate": 4.469178082191781e-06, + "loss": 1.6285, + "step": 277 + }, + { + "epoch": 0.28556753980482796, + "grad_norm": 2.4676456451416016, + "learning_rate": 4.486301369863014e-06, + "loss": 1.6838, + "step": 278 + }, + { + "epoch": 0.28659476117103233, + "grad_norm": 2.659083366394043, + "learning_rate": 4.503424657534247e-06, + "loss": 1.5872, + "step": 279 + }, + { + "epoch": 0.28762198253723675, + "grad_norm": 2.326733350753784, + "learning_rate": 4.52054794520548e-06, + "loss": 1.6142, + "step": 280 + }, + { + "epoch": 0.28762198253723675, + "eval_loss": 1.904759407043457, + "eval_runtime": 37.9651, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 280 + }, + { + "epoch": 0.2886492039034412, + "grad_norm": 2.299504518508911, + "learning_rate": 4.537671232876713e-06, + "loss": 1.6673, + "step": 281 + }, + { + "epoch": 0.2896764252696456, + "grad_norm": 2.4932825565338135, + "learning_rate": 4.554794520547945e-06, + "loss": 1.6586, + "step": 282 + }, + { + "epoch": 0.29070364663585, + "grad_norm": 2.2599050998687744, + "learning_rate": 4.571917808219179e-06, + "loss": 1.6568, + "step": 283 + }, + { + "epoch": 0.29173086800205444, + "grad_norm": 2.435098886489868, + "learning_rate": 4.589041095890411e-06, + "loss": 1.6113, + "step": 284 + }, + { + "epoch": 0.29275808936825887, + "grad_norm": 2.539246082305908, + "learning_rate": 4.606164383561644e-06, + "loss": 1.6306, + "step": 285 + }, + { + "epoch": 0.2937853107344633, + "grad_norm": 2.596198320388794, + "learning_rate": 4.6232876712328774e-06, + "loss": 1.6473, + "step": 286 + }, + { + "epoch": 0.2948125321006677, + "grad_norm": 2.349133014678955, + "learning_rate": 4.64041095890411e-06, + "loss": 1.6042, + "step": 287 + }, + { + "epoch": 0.29583975346687214, + "grad_norm": 2.3572661876678467, + "learning_rate": 4.657534246575343e-06, + "loss": 1.6635, + "step": 288 + }, + { + "epoch": 0.2968669748330765, + "grad_norm": 2.4210307598114014, + "learning_rate": 4.674657534246575e-06, + "loss": 1.624, + "step": 289 + }, + { + "epoch": 0.2978941961992809, + "grad_norm": 2.427403688430786, + "learning_rate": 4.691780821917809e-06, + "loss": 1.6824, + "step": 290 + }, + { + "epoch": 0.2978941961992809, + "eval_loss": 1.8888994455337524, + "eval_runtime": 37.9777, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 290 + }, + { + "epoch": 0.29892141756548535, + "grad_norm": 2.534029722213745, + "learning_rate": 4.7089041095890414e-06, + "loss": 1.6456, + "step": 291 + }, + { + "epoch": 0.29994863893168977, + "grad_norm": 2.329305648803711, + "learning_rate": 4.726027397260274e-06, + "loss": 1.646, + "step": 292 + }, + { + "epoch": 0.3009758602978942, + "grad_norm": 2.561548948287964, + "learning_rate": 4.7431506849315075e-06, + "loss": 1.6511, + "step": 293 + }, + { + "epoch": 0.3020030816640986, + "grad_norm": 2.707712411880493, + "learning_rate": 4.76027397260274e-06, + "loss": 1.5984, + "step": 294 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 2.403089761734009, + "learning_rate": 4.777397260273973e-06, + "loss": 1.6164, + "step": 295 + }, + { + "epoch": 0.30405752439650746, + "grad_norm": 2.424910306930542, + "learning_rate": 4.7945205479452054e-06, + "loss": 1.6291, + "step": 296 + }, + { + "epoch": 0.3050847457627119, + "grad_norm": 2.3474302291870117, + "learning_rate": 4.811643835616439e-06, + "loss": 1.553, + "step": 297 + }, + { + "epoch": 0.3061119671289163, + "grad_norm": 2.2211227416992188, + "learning_rate": 4.8287671232876716e-06, + "loss": 1.6108, + "step": 298 + }, + { + "epoch": 0.3071391884951207, + "grad_norm": 2.495983600616455, + "learning_rate": 4.845890410958904e-06, + "loss": 1.6135, + "step": 299 + }, + { + "epoch": 0.3081664098613251, + "grad_norm": 2.525495767593384, + "learning_rate": 4.863013698630138e-06, + "loss": 1.6061, + "step": 300 + }, + { + "epoch": 0.3081664098613251, + "eval_loss": 1.8982168436050415, + "eval_runtime": 37.9848, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 300 + }, + { + "epoch": 0.3091936312275295, + "grad_norm": 2.449064254760742, + "learning_rate": 4.88013698630137e-06, + "loss": 1.5827, + "step": 301 + }, + { + "epoch": 0.31022085259373394, + "grad_norm": 2.288095712661743, + "learning_rate": 4.897260273972603e-06, + "loss": 1.704, + "step": 302 + }, + { + "epoch": 0.31124807395993837, + "grad_norm": 2.5301458835601807, + "learning_rate": 4.914383561643836e-06, + "loss": 1.6342, + "step": 303 + }, + { + "epoch": 0.3122752953261428, + "grad_norm": 2.2793986797332764, + "learning_rate": 4.931506849315069e-06, + "loss": 1.6688, + "step": 304 + }, + { + "epoch": 0.3133025166923472, + "grad_norm": 2.38783597946167, + "learning_rate": 4.948630136986302e-06, + "loss": 1.7014, + "step": 305 + }, + { + "epoch": 0.31432973805855163, + "grad_norm": 2.5224838256835938, + "learning_rate": 4.965753424657534e-06, + "loss": 1.5812, + "step": 306 + }, + { + "epoch": 0.31535695942475606, + "grad_norm": 2.434432029724121, + "learning_rate": 4.982876712328768e-06, + "loss": 1.6186, + "step": 307 + }, + { + "epoch": 0.3163841807909605, + "grad_norm": 2.3708460330963135, + "learning_rate": 5e-06, + "loss": 1.5669, + "step": 308 + }, + { + "epoch": 0.31741140215716485, + "grad_norm": 2.5524990558624268, + "learning_rate": 4.998096688237534e-06, + "loss": 1.6688, + "step": 309 + }, + { + "epoch": 0.31843862352336927, + "grad_norm": 2.322441577911377, + "learning_rate": 4.996193376475067e-06, + "loss": 1.5571, + "step": 310 + }, + { + "epoch": 0.31843862352336927, + "eval_loss": 1.8914265632629395, + "eval_runtime": 37.6505, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 310 + }, + { + "epoch": 0.3194658448895737, + "grad_norm": 2.5248687267303467, + "learning_rate": 4.9942900647126005e-06, + "loss": 1.5585, + "step": 311 + }, + { + "epoch": 0.3204930662557781, + "grad_norm": 2.2464914321899414, + "learning_rate": 4.992386752950134e-06, + "loss": 1.5921, + "step": 312 + }, + { + "epoch": 0.32152028762198254, + "grad_norm": 2.5355331897735596, + "learning_rate": 4.990483441187667e-06, + "loss": 1.6263, + "step": 313 + }, + { + "epoch": 0.32254750898818696, + "grad_norm": 2.5611143112182617, + "learning_rate": 4.9885801294252e-06, + "loss": 1.6267, + "step": 314 + }, + { + "epoch": 0.3235747303543914, + "grad_norm": 2.364356517791748, + "learning_rate": 4.986676817662733e-06, + "loss": 1.5847, + "step": 315 + }, + { + "epoch": 0.3246019517205958, + "grad_norm": 2.455165147781372, + "learning_rate": 4.9847735059002665e-06, + "loss": 1.661, + "step": 316 + }, + { + "epoch": 0.3256291730868002, + "grad_norm": 2.4312195777893066, + "learning_rate": 4.9828701941378e-06, + "loss": 1.5995, + "step": 317 + }, + { + "epoch": 0.3266563944530046, + "grad_norm": 2.4081499576568604, + "learning_rate": 4.980966882375333e-06, + "loss": 1.5756, + "step": 318 + }, + { + "epoch": 0.327683615819209, + "grad_norm": 2.2608511447906494, + "learning_rate": 4.979063570612867e-06, + "loss": 1.543, + "step": 319 + }, + { + "epoch": 0.32871083718541344, + "grad_norm": 2.6551144123077393, + "learning_rate": 4.9771602588504e-06, + "loss": 1.6467, + "step": 320 + }, + { + "epoch": 0.32871083718541344, + "eval_loss": 1.89058518409729, + "eval_runtime": 37.8192, + "eval_samples_per_second": 26.442, + "eval_steps_per_second": 1.111, + "step": 320 + }, + { + "epoch": 0.32973805855161786, + "grad_norm": 2.3897273540496826, + "learning_rate": 4.975256947087933e-06, + "loss": 1.5607, + "step": 321 + }, + { + "epoch": 0.3307652799178223, + "grad_norm": 2.3869469165802, + "learning_rate": 4.973353635325467e-06, + "loss": 1.6079, + "step": 322 + }, + { + "epoch": 0.3317925012840267, + "grad_norm": 2.159193515777588, + "learning_rate": 4.971450323563e-06, + "loss": 1.5475, + "step": 323 + }, + { + "epoch": 0.33281972265023113, + "grad_norm": 2.3069653511047363, + "learning_rate": 4.969547011800533e-06, + "loss": 1.616, + "step": 324 + }, + { + "epoch": 0.33384694401643555, + "grad_norm": 2.3131422996520996, + "learning_rate": 4.967643700038067e-06, + "loss": 1.6106, + "step": 325 + }, + { + "epoch": 0.33487416538264, + "grad_norm": 2.4398200511932373, + "learning_rate": 4.9657403882756e-06, + "loss": 1.5779, + "step": 326 + }, + { + "epoch": 0.3359013867488444, + "grad_norm": 2.278024435043335, + "learning_rate": 4.9638370765131335e-06, + "loss": 1.5724, + "step": 327 + }, + { + "epoch": 0.33692860811504877, + "grad_norm": 2.4989173412323, + "learning_rate": 4.961933764750667e-06, + "loss": 1.6337, + "step": 328 + }, + { + "epoch": 0.3379558294812532, + "grad_norm": 2.46319580078125, + "learning_rate": 4.9600304529882e-06, + "loss": 1.6519, + "step": 329 + }, + { + "epoch": 0.3389830508474576, + "grad_norm": 2.2703049182891846, + "learning_rate": 4.9581271412257335e-06, + "loss": 1.6081, + "step": 330 + }, + { + "epoch": 0.3389830508474576, + "eval_loss": 1.8755545616149902, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 330 + }, + { + "epoch": 0.34001027221366203, + "grad_norm": 2.502196788787842, + "learning_rate": 4.956223829463266e-06, + "loss": 1.5221, + "step": 331 + }, + { + "epoch": 0.34103749357986646, + "grad_norm": 2.284280776977539, + "learning_rate": 4.954320517700799e-06, + "loss": 1.6268, + "step": 332 + }, + { + "epoch": 0.3420647149460709, + "grad_norm": 2.3551881313323975, + "learning_rate": 4.952417205938333e-06, + "loss": 1.6078, + "step": 333 + }, + { + "epoch": 0.3430919363122753, + "grad_norm": 2.343780755996704, + "learning_rate": 4.950513894175866e-06, + "loss": 1.591, + "step": 334 + }, + { + "epoch": 0.3441191576784797, + "grad_norm": 2.3524727821350098, + "learning_rate": 4.9486105824133995e-06, + "loss": 1.605, + "step": 335 + }, + { + "epoch": 0.34514637904468415, + "grad_norm": 2.349896192550659, + "learning_rate": 4.946707270650933e-06, + "loss": 1.6106, + "step": 336 + }, + { + "epoch": 0.34617360041088857, + "grad_norm": 2.5154407024383545, + "learning_rate": 4.944803958888466e-06, + "loss": 1.6429, + "step": 337 + }, + { + "epoch": 0.34720082177709294, + "grad_norm": 2.2692947387695312, + "learning_rate": 4.942900647126e-06, + "loss": 1.5865, + "step": 338 + }, + { + "epoch": 0.34822804314329736, + "grad_norm": 2.2672805786132812, + "learning_rate": 4.940997335363533e-06, + "loss": 1.5464, + "step": 339 + }, + { + "epoch": 0.3492552645095018, + "grad_norm": 2.5147039890289307, + "learning_rate": 4.939094023601066e-06, + "loss": 1.5837, + "step": 340 + }, + { + "epoch": 0.3492552645095018, + "eval_loss": 1.8672455549240112, + "eval_runtime": 37.9644, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 340 + }, + { + "epoch": 0.3502824858757062, + "grad_norm": 2.3416433334350586, + "learning_rate": 4.9371907118386e-06, + "loss": 1.5937, + "step": 341 + }, + { + "epoch": 0.35130970724191063, + "grad_norm": 2.3044960498809814, + "learning_rate": 4.935287400076133e-06, + "loss": 1.6132, + "step": 342 + }, + { + "epoch": 0.35233692860811505, + "grad_norm": 2.4360430240631104, + "learning_rate": 4.933384088313666e-06, + "loss": 1.5731, + "step": 343 + }, + { + "epoch": 0.3533641499743195, + "grad_norm": 2.392050266265869, + "learning_rate": 4.9314807765512e-06, + "loss": 1.5098, + "step": 344 + }, + { + "epoch": 0.3543913713405239, + "grad_norm": 2.3815808296203613, + "learning_rate": 4.929577464788733e-06, + "loss": 1.5827, + "step": 345 + }, + { + "epoch": 0.3554185927067283, + "grad_norm": 2.423077344894409, + "learning_rate": 4.9276741530262665e-06, + "loss": 1.5857, + "step": 346 + }, + { + "epoch": 0.35644581407293274, + "grad_norm": 2.5535898208618164, + "learning_rate": 4.9257708412638e-06, + "loss": 1.5884, + "step": 347 + }, + { + "epoch": 0.3574730354391371, + "grad_norm": 2.2665445804595947, + "learning_rate": 4.923867529501332e-06, + "loss": 1.5561, + "step": 348 + }, + { + "epoch": 0.35850025680534153, + "grad_norm": 2.3357481956481934, + "learning_rate": 4.921964217738866e-06, + "loss": 1.6444, + "step": 349 + }, + { + "epoch": 0.35952747817154596, + "grad_norm": 2.4537885189056396, + "learning_rate": 4.920060905976399e-06, + "loss": 1.6262, + "step": 350 + }, + { + "epoch": 0.35952747817154596, + "eval_loss": 1.8621584177017212, + "eval_runtime": 38.0027, + "eval_samples_per_second": 26.314, + "eval_steps_per_second": 1.105, + "step": 350 + }, + { + "epoch": 0.3605546995377504, + "grad_norm": 2.379754066467285, + "learning_rate": 4.9181575942139325e-06, + "loss": 1.6227, + "step": 351 + }, + { + "epoch": 0.3615819209039548, + "grad_norm": 2.432905435562134, + "learning_rate": 4.916254282451466e-06, + "loss": 1.6554, + "step": 352 + }, + { + "epoch": 0.3626091422701592, + "grad_norm": 2.34236216545105, + "learning_rate": 4.914350970688999e-06, + "loss": 1.5613, + "step": 353 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.3019444942474365, + "learning_rate": 4.9124476589265325e-06, + "loss": 1.5535, + "step": 354 + }, + { + "epoch": 0.36466358500256807, + "grad_norm": 2.3354644775390625, + "learning_rate": 4.910544347164066e-06, + "loss": 1.6384, + "step": 355 + }, + { + "epoch": 0.3656908063687725, + "grad_norm": 2.5154788494110107, + "learning_rate": 4.908641035401599e-06, + "loss": 1.6986, + "step": 356 + }, + { + "epoch": 0.3667180277349769, + "grad_norm": 2.2707178592681885, + "learning_rate": 4.906737723639133e-06, + "loss": 1.5705, + "step": 357 + }, + { + "epoch": 0.3677452491011813, + "grad_norm": 2.3058297634124756, + "learning_rate": 4.904834411876665e-06, + "loss": 1.6033, + "step": 358 + }, + { + "epoch": 0.3687724704673857, + "grad_norm": 2.3923370838165283, + "learning_rate": 4.902931100114199e-06, + "loss": 1.5333, + "step": 359 + }, + { + "epoch": 0.3697996918335901, + "grad_norm": 2.4023196697235107, + "learning_rate": 4.901027788351733e-06, + "loss": 1.5775, + "step": 360 + }, + { + "epoch": 0.3697996918335901, + "eval_loss": 1.867856740951538, + "eval_runtime": 38.0125, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 360 + }, + { + "epoch": 0.37082691319979455, + "grad_norm": 2.4903581142425537, + "learning_rate": 4.899124476589266e-06, + "loss": 1.5491, + "step": 361 + }, + { + "epoch": 0.371854134565999, + "grad_norm": 2.556683301925659, + "learning_rate": 4.8972211648267994e-06, + "loss": 1.5967, + "step": 362 + }, + { + "epoch": 0.3728813559322034, + "grad_norm": 2.3017377853393555, + "learning_rate": 4.895317853064333e-06, + "loss": 1.5661, + "step": 363 + }, + { + "epoch": 0.3739085772984078, + "grad_norm": 2.286191940307617, + "learning_rate": 4.893414541301866e-06, + "loss": 1.5679, + "step": 364 + }, + { + "epoch": 0.37493579866461224, + "grad_norm": 2.260812520980835, + "learning_rate": 4.891511229539399e-06, + "loss": 1.5965, + "step": 365 + }, + { + "epoch": 0.37596302003081666, + "grad_norm": 2.4569408893585205, + "learning_rate": 4.889607917776932e-06, + "loss": 1.5923, + "step": 366 + }, + { + "epoch": 0.37699024139702103, + "grad_norm": 2.571552038192749, + "learning_rate": 4.887704606014465e-06, + "loss": 1.6021, + "step": 367 + }, + { + "epoch": 0.37801746276322545, + "grad_norm": 2.287754774093628, + "learning_rate": 4.885801294251999e-06, + "loss": 1.5304, + "step": 368 + }, + { + "epoch": 0.3790446841294299, + "grad_norm": 2.317784070968628, + "learning_rate": 4.883897982489532e-06, + "loss": 1.6233, + "step": 369 + }, + { + "epoch": 0.3800719054956343, + "grad_norm": 2.290601968765259, + "learning_rate": 4.8819946707270655e-06, + "loss": 1.624, + "step": 370 + }, + { + "epoch": 0.3800719054956343, + "eval_loss": 1.8604528903961182, + "eval_runtime": 38.0176, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 370 + }, + { + "epoch": 0.3810991268618387, + "grad_norm": 2.28519868850708, + "learning_rate": 4.880091358964599e-06, + "loss": 1.5286, + "step": 371 + }, + { + "epoch": 0.38212634822804314, + "grad_norm": 2.271810293197632, + "learning_rate": 4.878188047202132e-06, + "loss": 1.6879, + "step": 372 + }, + { + "epoch": 0.38315356959424757, + "grad_norm": 2.391848564147949, + "learning_rate": 4.876284735439666e-06, + "loss": 1.6458, + "step": 373 + }, + { + "epoch": 0.384180790960452, + "grad_norm": 2.3787031173706055, + "learning_rate": 4.874381423677199e-06, + "loss": 1.5738, + "step": 374 + }, + { + "epoch": 0.3852080123266564, + "grad_norm": 2.404219150543213, + "learning_rate": 4.8724781119147315e-06, + "loss": 1.6184, + "step": 375 + }, + { + "epoch": 0.38623523369286084, + "grad_norm": 2.372661590576172, + "learning_rate": 4.870574800152265e-06, + "loss": 1.5316, + "step": 376 + }, + { + "epoch": 0.3872624550590652, + "grad_norm": 2.4397103786468506, + "learning_rate": 4.868671488389798e-06, + "loss": 1.5605, + "step": 377 + }, + { + "epoch": 0.3882896764252696, + "grad_norm": 2.371006727218628, + "learning_rate": 4.8667681766273316e-06, + "loss": 1.5184, + "step": 378 + }, + { + "epoch": 0.38931689779147405, + "grad_norm": 2.2621910572052, + "learning_rate": 4.864864864864866e-06, + "loss": 1.5788, + "step": 379 + }, + { + "epoch": 0.39034411915767847, + "grad_norm": 2.2888615131378174, + "learning_rate": 4.862961553102399e-06, + "loss": 1.6244, + "step": 380 + }, + { + "epoch": 0.39034411915767847, + "eval_loss": 1.8586912155151367, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 380 + }, + { + "epoch": 0.3913713405238829, + "grad_norm": 2.4439220428466797, + "learning_rate": 4.8610582413399325e-06, + "loss": 1.5749, + "step": 381 + }, + { + "epoch": 0.3923985618900873, + "grad_norm": 2.2423763275146484, + "learning_rate": 4.859154929577465e-06, + "loss": 1.5738, + "step": 382 + }, + { + "epoch": 0.39342578325629174, + "grad_norm": 2.267467737197876, + "learning_rate": 4.857251617814998e-06, + "loss": 1.6212, + "step": 383 + }, + { + "epoch": 0.39445300462249616, + "grad_norm": 2.3680620193481445, + "learning_rate": 4.855348306052532e-06, + "loss": 1.6129, + "step": 384 + }, + { + "epoch": 0.3954802259887006, + "grad_norm": 2.3782172203063965, + "learning_rate": 4.853444994290065e-06, + "loss": 1.5605, + "step": 385 + }, + { + "epoch": 0.396507447354905, + "grad_norm": 2.5745720863342285, + "learning_rate": 4.8515416825275985e-06, + "loss": 1.634, + "step": 386 + }, + { + "epoch": 0.3975346687211094, + "grad_norm": 2.38212513923645, + "learning_rate": 4.849638370765132e-06, + "loss": 1.5671, + "step": 387 + }, + { + "epoch": 0.3985618900873138, + "grad_norm": 2.3395450115203857, + "learning_rate": 4.847735059002665e-06, + "loss": 1.5088, + "step": 388 + }, + { + "epoch": 0.3995891114535182, + "grad_norm": 2.474440574645996, + "learning_rate": 4.8458317472401985e-06, + "loss": 1.5997, + "step": 389 + }, + { + "epoch": 0.40061633281972264, + "grad_norm": 2.429133176803589, + "learning_rate": 4.843928435477732e-06, + "loss": 1.5611, + "step": 390 + }, + { + "epoch": 0.40061633281972264, + "eval_loss": 1.8613603115081787, + "eval_runtime": 37.9862, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 390 + }, + { + "epoch": 0.40164355418592707, + "grad_norm": 2.403207778930664, + "learning_rate": 4.842025123715265e-06, + "loss": 1.5976, + "step": 391 + }, + { + "epoch": 0.4026707755521315, + "grad_norm": 2.383018732070923, + "learning_rate": 4.840121811952798e-06, + "loss": 1.5791, + "step": 392 + }, + { + "epoch": 0.4036979969183359, + "grad_norm": 2.5303306579589844, + "learning_rate": 4.838218500190331e-06, + "loss": 1.6035, + "step": 393 + }, + { + "epoch": 0.40472521828454033, + "grad_norm": 2.391075372695923, + "learning_rate": 4.8363151884278645e-06, + "loss": 1.5795, + "step": 394 + }, + { + "epoch": 0.40575243965074476, + "grad_norm": 2.396934747695923, + "learning_rate": 4.834411876665398e-06, + "loss": 1.5965, + "step": 395 + }, + { + "epoch": 0.4067796610169492, + "grad_norm": 2.4631423950195312, + "learning_rate": 4.832508564902931e-06, + "loss": 1.5372, + "step": 396 + }, + { + "epoch": 0.40780688238315355, + "grad_norm": 2.3457326889038086, + "learning_rate": 4.830605253140465e-06, + "loss": 1.5729, + "step": 397 + }, + { + "epoch": 0.40883410374935797, + "grad_norm": 2.3763277530670166, + "learning_rate": 4.828701941377998e-06, + "loss": 1.5309, + "step": 398 + }, + { + "epoch": 0.4098613251155624, + "grad_norm": 2.442439556121826, + "learning_rate": 4.826798629615531e-06, + "loss": 1.5465, + "step": 399 + }, + { + "epoch": 0.4108885464817668, + "grad_norm": 2.3464581966400146, + "learning_rate": 4.824895317853065e-06, + "loss": 1.5665, + "step": 400 + }, + { + "epoch": 0.4108885464817668, + "eval_loss": 1.8671027421951294, + "eval_runtime": 38.0166, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 400 + }, + { + "epoch": 0.41191576784797124, + "grad_norm": 2.5441887378692627, + "learning_rate": 4.822992006090598e-06, + "loss": 1.5232, + "step": 401 + }, + { + "epoch": 0.41294298921417566, + "grad_norm": 2.4234976768493652, + "learning_rate": 4.821088694328131e-06, + "loss": 1.5569, + "step": 402 + }, + { + "epoch": 0.4139702105803801, + "grad_norm": 2.3603928089141846, + "learning_rate": 4.819185382565665e-06, + "loss": 1.612, + "step": 403 + }, + { + "epoch": 0.4149974319465845, + "grad_norm": 2.337862491607666, + "learning_rate": 4.817282070803198e-06, + "loss": 1.5882, + "step": 404 + }, + { + "epoch": 0.41602465331278893, + "grad_norm": 2.3805742263793945, + "learning_rate": 4.8153787590407315e-06, + "loss": 1.5204, + "step": 405 + }, + { + "epoch": 0.41705187467899335, + "grad_norm": 2.3335204124450684, + "learning_rate": 4.813475447278265e-06, + "loss": 1.6041, + "step": 406 + }, + { + "epoch": 0.4180790960451977, + "grad_norm": 2.406137466430664, + "learning_rate": 4.811572135515798e-06, + "loss": 1.5348, + "step": 407 + }, + { + "epoch": 0.41910631741140214, + "grad_norm": 2.2823917865753174, + "learning_rate": 4.809668823753332e-06, + "loss": 1.5351, + "step": 408 + }, + { + "epoch": 0.42013353877760656, + "grad_norm": 2.3577091693878174, + "learning_rate": 4.807765511990864e-06, + "loss": 1.5733, + "step": 409 + }, + { + "epoch": 0.421160760143811, + "grad_norm": 2.536691665649414, + "learning_rate": 4.8058622002283975e-06, + "loss": 1.6153, + "step": 410 + }, + { + "epoch": 0.421160760143811, + "eval_loss": 1.8613901138305664, + "eval_runtime": 37.6689, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 410 + }, + { + "epoch": 0.4221879815100154, + "grad_norm": 2.4020118713378906, + "learning_rate": 4.803958888465931e-06, + "loss": 1.5722, + "step": 411 + }, + { + "epoch": 0.42321520287621983, + "grad_norm": 2.5284810066223145, + "learning_rate": 4.802055576703464e-06, + "loss": 1.6546, + "step": 412 + }, + { + "epoch": 0.42424242424242425, + "grad_norm": 2.270094156265259, + "learning_rate": 4.8001522649409976e-06, + "loss": 1.5345, + "step": 413 + }, + { + "epoch": 0.4252696456086287, + "grad_norm": 2.241417407989502, + "learning_rate": 4.798248953178531e-06, + "loss": 1.5167, + "step": 414 + }, + { + "epoch": 0.4262968669748331, + "grad_norm": 2.3733017444610596, + "learning_rate": 4.796345641416064e-06, + "loss": 1.5532, + "step": 415 + }, + { + "epoch": 0.42732408834103747, + "grad_norm": 2.452587842941284, + "learning_rate": 4.794442329653598e-06, + "loss": 1.6136, + "step": 416 + }, + { + "epoch": 0.4283513097072419, + "grad_norm": 2.417217969894409, + "learning_rate": 4.792539017891131e-06, + "loss": 1.5398, + "step": 417 + }, + { + "epoch": 0.4293785310734463, + "grad_norm": 2.4502453804016113, + "learning_rate": 4.790635706128664e-06, + "loss": 1.5741, + "step": 418 + }, + { + "epoch": 0.43040575243965074, + "grad_norm": 2.2996811866760254, + "learning_rate": 4.788732394366197e-06, + "loss": 1.511, + "step": 419 + }, + { + "epoch": 0.43143297380585516, + "grad_norm": 2.3573555946350098, + "learning_rate": 4.786829082603731e-06, + "loss": 1.5117, + "step": 420 + }, + { + "epoch": 0.43143297380585516, + "eval_loss": 1.8690649271011353, + "eval_runtime": 37.7813, + "eval_samples_per_second": 26.468, + "eval_steps_per_second": 1.112, + "step": 420 + }, + { + "epoch": 0.4324601951720596, + "grad_norm": 2.37233567237854, + "learning_rate": 4.7849257708412645e-06, + "loss": 1.5873, + "step": 421 + }, + { + "epoch": 0.433487416538264, + "grad_norm": 2.146909713745117, + "learning_rate": 4.783022459078798e-06, + "loss": 1.5291, + "step": 422 + }, + { + "epoch": 0.4345146379044684, + "grad_norm": 2.5237205028533936, + "learning_rate": 4.781119147316331e-06, + "loss": 1.5545, + "step": 423 + }, + { + "epoch": 0.43554185927067285, + "grad_norm": 2.356842279434204, + "learning_rate": 4.7792158355538645e-06, + "loss": 1.5635, + "step": 424 + }, + { + "epoch": 0.43656908063687727, + "grad_norm": 2.1963610649108887, + "learning_rate": 4.777312523791398e-06, + "loss": 1.5355, + "step": 425 + }, + { + "epoch": 0.43759630200308164, + "grad_norm": 2.6425886154174805, + "learning_rate": 4.77540921202893e-06, + "loss": 1.5803, + "step": 426 + }, + { + "epoch": 0.43862352336928606, + "grad_norm": 2.4072272777557373, + "learning_rate": 4.773505900266464e-06, + "loss": 1.5649, + "step": 427 + }, + { + "epoch": 0.4396507447354905, + "grad_norm": 2.592374801635742, + "learning_rate": 4.771602588503997e-06, + "loss": 1.5693, + "step": 428 + }, + { + "epoch": 0.4406779661016949, + "grad_norm": 2.4534225463867188, + "learning_rate": 4.7696992767415305e-06, + "loss": 1.591, + "step": 429 + }, + { + "epoch": 0.44170518746789933, + "grad_norm": 2.434457778930664, + "learning_rate": 4.767795964979064e-06, + "loss": 1.5991, + "step": 430 + }, + { + "epoch": 0.44170518746789933, + "eval_loss": 1.8584965467453003, + "eval_runtime": 37.8854, + "eval_samples_per_second": 26.395, + "eval_steps_per_second": 1.109, + "step": 430 + }, + { + "epoch": 0.44273240883410375, + "grad_norm": 2.3150746822357178, + "learning_rate": 4.765892653216597e-06, + "loss": 1.6032, + "step": 431 + }, + { + "epoch": 0.4437596302003082, + "grad_norm": 2.3040223121643066, + "learning_rate": 4.763989341454131e-06, + "loss": 1.5115, + "step": 432 + }, + { + "epoch": 0.4447868515665126, + "grad_norm": 2.413007974624634, + "learning_rate": 4.762086029691664e-06, + "loss": 1.5305, + "step": 433 + }, + { + "epoch": 0.445814072932717, + "grad_norm": 2.526289463043213, + "learning_rate": 4.760182717929197e-06, + "loss": 1.564, + "step": 434 + }, + { + "epoch": 0.44684129429892144, + "grad_norm": 2.4122776985168457, + "learning_rate": 4.758279406166731e-06, + "loss": 1.4918, + "step": 435 + }, + { + "epoch": 0.4478685156651258, + "grad_norm": 2.474274158477783, + "learning_rate": 4.756376094404263e-06, + "loss": 1.5583, + "step": 436 + }, + { + "epoch": 0.44889573703133023, + "grad_norm": 2.5322892665863037, + "learning_rate": 4.7544727826417966e-06, + "loss": 1.6081, + "step": 437 + }, + { + "epoch": 0.44992295839753466, + "grad_norm": 2.391660690307617, + "learning_rate": 4.75256947087933e-06, + "loss": 1.5628, + "step": 438 + }, + { + "epoch": 0.4509501797637391, + "grad_norm": 2.426671028137207, + "learning_rate": 4.750666159116863e-06, + "loss": 1.531, + "step": 439 + }, + { + "epoch": 0.4519774011299435, + "grad_norm": 2.7095694541931152, + "learning_rate": 4.7487628473543975e-06, + "loss": 1.4811, + "step": 440 + }, + { + "epoch": 0.4519774011299435, + "eval_loss": 1.8348888158798218, + "eval_runtime": 37.9114, + "eval_samples_per_second": 26.377, + "eval_steps_per_second": 1.108, + "step": 440 + }, + { + "epoch": 0.4530046224961479, + "grad_norm": 2.5425240993499756, + "learning_rate": 4.746859535591931e-06, + "loss": 1.5453, + "step": 441 + }, + { + "epoch": 0.45403184386235235, + "grad_norm": 2.6848669052124023, + "learning_rate": 4.744956223829464e-06, + "loss": 1.526, + "step": 442 + }, + { + "epoch": 0.45505906522855677, + "grad_norm": 2.634068489074707, + "learning_rate": 4.743052912066997e-06, + "loss": 1.5748, + "step": 443 + }, + { + "epoch": 0.4560862865947612, + "grad_norm": 2.492030382156372, + "learning_rate": 4.74114960030453e-06, + "loss": 1.5858, + "step": 444 + }, + { + "epoch": 0.4571135079609656, + "grad_norm": 2.787740707397461, + "learning_rate": 4.7392462885420635e-06, + "loss": 1.4774, + "step": 445 + }, + { + "epoch": 0.45814072932717, + "grad_norm": 2.395914077758789, + "learning_rate": 4.737342976779597e-06, + "loss": 1.558, + "step": 446 + }, + { + "epoch": 0.4591679506933744, + "grad_norm": 2.6642563343048096, + "learning_rate": 4.73543966501713e-06, + "loss": 1.5558, + "step": 447 + }, + { + "epoch": 0.4601951720595788, + "grad_norm": 2.619943857192993, + "learning_rate": 4.7335363532546636e-06, + "loss": 1.5442, + "step": 448 + }, + { + "epoch": 0.46122239342578325, + "grad_norm": 2.355133056640625, + "learning_rate": 4.731633041492197e-06, + "loss": 1.5613, + "step": 449 + }, + { + "epoch": 0.4622496147919877, + "grad_norm": 2.6572225093841553, + "learning_rate": 4.72972972972973e-06, + "loss": 1.5951, + "step": 450 + }, + { + "epoch": 0.4622496147919877, + "eval_loss": 1.83199143409729, + "eval_runtime": 37.8996, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 450 + }, + { + "epoch": 0.4632768361581921, + "grad_norm": 2.692873001098633, + "learning_rate": 4.727826417967264e-06, + "loss": 1.5531, + "step": 451 + }, + { + "epoch": 0.4643040575243965, + "grad_norm": 2.3903489112854004, + "learning_rate": 4.725923106204797e-06, + "loss": 1.5411, + "step": 452 + }, + { + "epoch": 0.46533127889060094, + "grad_norm": 2.453263521194458, + "learning_rate": 4.72401979444233e-06, + "loss": 1.5632, + "step": 453 + }, + { + "epoch": 0.46635850025680536, + "grad_norm": 2.5637760162353516, + "learning_rate": 4.722116482679863e-06, + "loss": 1.5629, + "step": 454 + }, + { + "epoch": 0.46738572162300973, + "grad_norm": 2.387033700942993, + "learning_rate": 4.720213170917396e-06, + "loss": 1.5786, + "step": 455 + }, + { + "epoch": 0.46841294298921415, + "grad_norm": 2.4563071727752686, + "learning_rate": 4.71830985915493e-06, + "loss": 1.5321, + "step": 456 + }, + { + "epoch": 0.4694401643554186, + "grad_norm": 2.607451915740967, + "learning_rate": 4.716406547392463e-06, + "loss": 1.5958, + "step": 457 + }, + { + "epoch": 0.470467385721623, + "grad_norm": 2.4846854209899902, + "learning_rate": 4.714503235629996e-06, + "loss": 1.5205, + "step": 458 + }, + { + "epoch": 0.4714946070878274, + "grad_norm": 2.377629041671753, + "learning_rate": 4.71259992386753e-06, + "loss": 1.5403, + "step": 459 + }, + { + "epoch": 0.47252182845403184, + "grad_norm": 2.4470160007476807, + "learning_rate": 4.710696612105063e-06, + "loss": 1.5456, + "step": 460 + }, + { + "epoch": 0.47252182845403184, + "eval_loss": 1.83836829662323, + "eval_runtime": 37.88, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 460 + }, + { + "epoch": 0.47354904982023627, + "grad_norm": 2.383418560028076, + "learning_rate": 4.708793300342596e-06, + "loss": 1.5468, + "step": 461 + }, + { + "epoch": 0.4745762711864407, + "grad_norm": 2.3672895431518555, + "learning_rate": 4.70688998858013e-06, + "loss": 1.5313, + "step": 462 + }, + { + "epoch": 0.4756034925526451, + "grad_norm": 2.415717124938965, + "learning_rate": 4.704986676817663e-06, + "loss": 1.5645, + "step": 463 + }, + { + "epoch": 0.47663071391884954, + "grad_norm": 2.4690067768096924, + "learning_rate": 4.7030833650551965e-06, + "loss": 1.5285, + "step": 464 + }, + { + "epoch": 0.4776579352850539, + "grad_norm": 2.3845672607421875, + "learning_rate": 4.70118005329273e-06, + "loss": 1.596, + "step": 465 + }, + { + "epoch": 0.4786851566512583, + "grad_norm": 2.2801802158355713, + "learning_rate": 4.699276741530263e-06, + "loss": 1.5777, + "step": 466 + }, + { + "epoch": 0.47971237801746275, + "grad_norm": 2.4668991565704346, + "learning_rate": 4.697373429767797e-06, + "loss": 1.5566, + "step": 467 + }, + { + "epoch": 0.48073959938366717, + "grad_norm": 2.446078062057495, + "learning_rate": 4.69547011800533e-06, + "loss": 1.5664, + "step": 468 + }, + { + "epoch": 0.4817668207498716, + "grad_norm": 2.2310681343078613, + "learning_rate": 4.693566806242863e-06, + "loss": 1.5425, + "step": 469 + }, + { + "epoch": 0.482794042116076, + "grad_norm": 2.3938229084014893, + "learning_rate": 4.691663494480397e-06, + "loss": 1.576, + "step": 470 + }, + { + "epoch": 0.482794042116076, + "eval_loss": 1.83778977394104, + "eval_runtime": 37.9276, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 470 + }, + { + "epoch": 0.48382126348228044, + "grad_norm": 2.23827862739563, + "learning_rate": 4.689760182717929e-06, + "loss": 1.4763, + "step": 471 + }, + { + "epoch": 0.48484848484848486, + "grad_norm": 2.3148930072784424, + "learning_rate": 4.6878568709554626e-06, + "loss": 1.5386, + "step": 472 + }, + { + "epoch": 0.4858757062146893, + "grad_norm": 2.407036781311035, + "learning_rate": 4.685953559192996e-06, + "loss": 1.5576, + "step": 473 + }, + { + "epoch": 0.4869029275808937, + "grad_norm": 2.4675374031066895, + "learning_rate": 4.684050247430529e-06, + "loss": 1.4931, + "step": 474 + }, + { + "epoch": 0.4879301489470981, + "grad_norm": 2.4593775272369385, + "learning_rate": 4.682146935668063e-06, + "loss": 1.5348, + "step": 475 + }, + { + "epoch": 0.4889573703133025, + "grad_norm": 2.4169375896453857, + "learning_rate": 4.680243623905596e-06, + "loss": 1.5614, + "step": 476 + }, + { + "epoch": 0.4899845916795069, + "grad_norm": 2.6177303791046143, + "learning_rate": 4.678340312143129e-06, + "loss": 1.5138, + "step": 477 + }, + { + "epoch": 0.49101181304571134, + "grad_norm": 2.6418113708496094, + "learning_rate": 4.676437000380663e-06, + "loss": 1.6377, + "step": 478 + }, + { + "epoch": 0.49203903441191577, + "grad_norm": 2.338491439819336, + "learning_rate": 4.674533688618196e-06, + "loss": 1.5682, + "step": 479 + }, + { + "epoch": 0.4930662557781202, + "grad_norm": 2.5901622772216797, + "learning_rate": 4.6726303768557295e-06, + "loss": 1.4899, + "step": 480 + }, + { + "epoch": 0.4930662557781202, + "eval_loss": 1.8238886594772339, + "eval_runtime": 37.9605, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 480 + }, + { + "epoch": 0.4940934771443246, + "grad_norm": 2.470827102661133, + "learning_rate": 4.670727065093263e-06, + "loss": 1.5016, + "step": 481 + }, + { + "epoch": 0.49512069851052903, + "grad_norm": 2.371201992034912, + "learning_rate": 4.668823753330796e-06, + "loss": 1.5399, + "step": 482 + }, + { + "epoch": 0.49614791987673346, + "grad_norm": 2.532802104949951, + "learning_rate": 4.6669204415683295e-06, + "loss": 1.5985, + "step": 483 + }, + { + "epoch": 0.4971751412429379, + "grad_norm": 2.6294727325439453, + "learning_rate": 4.665017129805863e-06, + "loss": 1.5802, + "step": 484 + }, + { + "epoch": 0.49820236260914225, + "grad_norm": 2.2164878845214844, + "learning_rate": 4.663113818043396e-06, + "loss": 1.5384, + "step": 485 + }, + { + "epoch": 0.49922958397534667, + "grad_norm": 2.311232805252075, + "learning_rate": 4.66121050628093e-06, + "loss": 1.5785, + "step": 486 + }, + { + "epoch": 0.5002568053415511, + "grad_norm": 2.2916736602783203, + "learning_rate": 4.659307194518463e-06, + "loss": 1.5401, + "step": 487 + }, + { + "epoch": 0.5012840267077555, + "grad_norm": 2.3978800773620605, + "learning_rate": 4.6574038827559955e-06, + "loss": 1.5619, + "step": 488 + }, + { + "epoch": 0.50231124807396, + "grad_norm": 2.438917398452759, + "learning_rate": 4.655500570993529e-06, + "loss": 1.5824, + "step": 489 + }, + { + "epoch": 0.5033384694401644, + "grad_norm": 2.528862237930298, + "learning_rate": 4.653597259231062e-06, + "loss": 1.4802, + "step": 490 + }, + { + "epoch": 0.5033384694401644, + "eval_loss": 1.83504056930542, + "eval_runtime": 37.9357, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 490 + }, + { + "epoch": 0.5043656908063687, + "grad_norm": 2.54522442817688, + "learning_rate": 4.651693947468596e-06, + "loss": 1.6186, + "step": 491 + }, + { + "epoch": 0.5053929121725732, + "grad_norm": 2.5037031173706055, + "learning_rate": 4.649790635706129e-06, + "loss": 1.5159, + "step": 492 + }, + { + "epoch": 0.5064201335387776, + "grad_norm": 2.4600465297698975, + "learning_rate": 4.647887323943662e-06, + "loss": 1.5378, + "step": 493 + }, + { + "epoch": 0.507447354904982, + "grad_norm": 2.393695831298828, + "learning_rate": 4.645984012181196e-06, + "loss": 1.4485, + "step": 494 + }, + { + "epoch": 0.5084745762711864, + "grad_norm": 2.488626718521118, + "learning_rate": 4.644080700418729e-06, + "loss": 1.57, + "step": 495 + }, + { + "epoch": 0.5095017976373909, + "grad_norm": 2.409858465194702, + "learning_rate": 4.642177388656262e-06, + "loss": 1.5343, + "step": 496 + }, + { + "epoch": 0.5105290190035953, + "grad_norm": 2.5607573986053467, + "learning_rate": 4.640274076893796e-06, + "loss": 1.5825, + "step": 497 + }, + { + "epoch": 0.5115562403697997, + "grad_norm": 2.5953354835510254, + "learning_rate": 4.638370765131328e-06, + "loss": 1.5129, + "step": 498 + }, + { + "epoch": 0.5125834617360041, + "grad_norm": 2.3321897983551025, + "learning_rate": 4.636467453368862e-06, + "loss": 1.4972, + "step": 499 + }, + { + "epoch": 0.5136106831022085, + "grad_norm": 2.332052707672119, + "learning_rate": 4.634564141606396e-06, + "loss": 1.567, + "step": 500 + }, + { + "epoch": 0.5136106831022085, + "eval_loss": 1.824558138847351, + "eval_runtime": 37.9157, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 500 + }, + { + "epoch": 0.514637904468413, + "grad_norm": 2.431809902191162, + "learning_rate": 4.632660829843929e-06, + "loss": 1.494, + "step": 501 + }, + { + "epoch": 0.5156651258346173, + "grad_norm": 2.7118983268737793, + "learning_rate": 4.630757518081463e-06, + "loss": 1.6289, + "step": 502 + }, + { + "epoch": 0.5166923472008218, + "grad_norm": 2.3549370765686035, + "learning_rate": 4.628854206318996e-06, + "loss": 1.5363, + "step": 503 + }, + { + "epoch": 0.5177195685670262, + "grad_norm": 2.3343000411987305, + "learning_rate": 4.626950894556529e-06, + "loss": 1.5749, + "step": 504 + }, + { + "epoch": 0.5187467899332306, + "grad_norm": 2.4281466007232666, + "learning_rate": 4.625047582794062e-06, + "loss": 1.4958, + "step": 505 + }, + { + "epoch": 0.519774011299435, + "grad_norm": 2.3858845233917236, + "learning_rate": 4.623144271031595e-06, + "loss": 1.5031, + "step": 506 + }, + { + "epoch": 0.5208012326656395, + "grad_norm": 2.3714635372161865, + "learning_rate": 4.6212409592691286e-06, + "loss": 1.5808, + "step": 507 + }, + { + "epoch": 0.5218284540318439, + "grad_norm": 2.5290520191192627, + "learning_rate": 4.619337647506662e-06, + "loss": 1.541, + "step": 508 + }, + { + "epoch": 0.5228556753980482, + "grad_norm": 2.5521368980407715, + "learning_rate": 4.617434335744195e-06, + "loss": 1.4992, + "step": 509 + }, + { + "epoch": 0.5238828967642527, + "grad_norm": 2.436588764190674, + "learning_rate": 4.615531023981729e-06, + "loss": 1.5283, + "step": 510 + }, + { + "epoch": 0.5238828967642527, + "eval_loss": 1.8220165967941284, + "eval_runtime": 37.6123, + "eval_samples_per_second": 26.587, + "eval_steps_per_second": 1.117, + "step": 510 + }, + { + "epoch": 0.5249101181304571, + "grad_norm": 2.5240023136138916, + "learning_rate": 4.613627712219262e-06, + "loss": 1.5426, + "step": 511 + }, + { + "epoch": 0.5259373394966615, + "grad_norm": 2.5292775630950928, + "learning_rate": 4.611724400456795e-06, + "loss": 1.5411, + "step": 512 + }, + { + "epoch": 0.5269645608628659, + "grad_norm": 2.3622210025787354, + "learning_rate": 4.609821088694329e-06, + "loss": 1.5083, + "step": 513 + }, + { + "epoch": 0.5279917822290704, + "grad_norm": 2.545267105102539, + "learning_rate": 4.607917776931862e-06, + "loss": 1.5541, + "step": 514 + }, + { + "epoch": 0.5290190035952748, + "grad_norm": 2.3417961597442627, + "learning_rate": 4.606014465169395e-06, + "loss": 1.5686, + "step": 515 + }, + { + "epoch": 0.5300462249614792, + "grad_norm": 2.2782580852508545, + "learning_rate": 4.604111153406928e-06, + "loss": 1.4761, + "step": 516 + }, + { + "epoch": 0.5310734463276836, + "grad_norm": 2.4437527656555176, + "learning_rate": 4.602207841644461e-06, + "loss": 1.5818, + "step": 517 + }, + { + "epoch": 0.5321006676938881, + "grad_norm": 2.3594319820404053, + "learning_rate": 4.600304529881995e-06, + "loss": 1.5556, + "step": 518 + }, + { + "epoch": 0.5331278890600925, + "grad_norm": 2.325979471206665, + "learning_rate": 4.598401218119528e-06, + "loss": 1.5726, + "step": 519 + }, + { + "epoch": 0.5341551104262968, + "grad_norm": 2.2765676975250244, + "learning_rate": 4.5964979063570614e-06, + "loss": 1.5151, + "step": 520 + }, + { + "epoch": 0.5341551104262968, + "eval_loss": 1.825669288635254, + "eval_runtime": 37.7491, + "eval_samples_per_second": 26.491, + "eval_steps_per_second": 1.113, + "step": 520 + }, + { + "epoch": 0.5351823317925013, + "grad_norm": 2.357647657394409, + "learning_rate": 4.594594594594596e-06, + "loss": 1.5505, + "step": 521 + }, + { + "epoch": 0.5362095531587057, + "grad_norm": 2.3652896881103516, + "learning_rate": 4.592691282832128e-06, + "loss": 1.5621, + "step": 522 + }, + { + "epoch": 0.5372367745249101, + "grad_norm": 2.234096050262451, + "learning_rate": 4.5907879710696615e-06, + "loss": 1.4851, + "step": 523 + }, + { + "epoch": 0.5382639958911145, + "grad_norm": 2.3414828777313232, + "learning_rate": 4.588884659307195e-06, + "loss": 1.5594, + "step": 524 + }, + { + "epoch": 0.539291217257319, + "grad_norm": 2.3934929370880127, + "learning_rate": 4.586981347544728e-06, + "loss": 1.4994, + "step": 525 + }, + { + "epoch": 0.5403184386235234, + "grad_norm": 2.400479555130005, + "learning_rate": 4.585078035782262e-06, + "loss": 1.5346, + "step": 526 + }, + { + "epoch": 0.5413456599897278, + "grad_norm": 2.3817873001098633, + "learning_rate": 4.583174724019795e-06, + "loss": 1.4803, + "step": 527 + }, + { + "epoch": 0.5423728813559322, + "grad_norm": 2.4503185749053955, + "learning_rate": 4.581271412257328e-06, + "loss": 1.4711, + "step": 528 + }, + { + "epoch": 0.5434001027221366, + "grad_norm": 2.452578067779541, + "learning_rate": 4.579368100494862e-06, + "loss": 1.5645, + "step": 529 + }, + { + "epoch": 0.544427324088341, + "grad_norm": 2.3278186321258545, + "learning_rate": 4.577464788732395e-06, + "loss": 1.5428, + "step": 530 + }, + { + "epoch": 0.544427324088341, + "eval_loss": 1.8242785930633545, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 530 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 2.5192365646362305, + "learning_rate": 4.575561476969928e-06, + "loss": 1.5896, + "step": 531 + }, + { + "epoch": 0.5464817668207499, + "grad_norm": 2.308701276779175, + "learning_rate": 4.573658165207461e-06, + "loss": 1.5518, + "step": 532 + }, + { + "epoch": 0.5475089881869543, + "grad_norm": 2.4221107959747314, + "learning_rate": 4.571754853444994e-06, + "loss": 1.5387, + "step": 533 + }, + { + "epoch": 0.5485362095531587, + "grad_norm": 2.457606077194214, + "learning_rate": 4.569851541682528e-06, + "loss": 1.4997, + "step": 534 + }, + { + "epoch": 0.5495634309193631, + "grad_norm": 2.3406503200531006, + "learning_rate": 4.567948229920061e-06, + "loss": 1.4761, + "step": 535 + }, + { + "epoch": 0.5505906522855676, + "grad_norm": 2.555712938308716, + "learning_rate": 4.566044918157594e-06, + "loss": 1.5114, + "step": 536 + }, + { + "epoch": 0.551617873651772, + "grad_norm": 2.373013734817505, + "learning_rate": 4.564141606395128e-06, + "loss": 1.5516, + "step": 537 + }, + { + "epoch": 0.5526450950179764, + "grad_norm": 2.3915255069732666, + "learning_rate": 4.562238294632661e-06, + "loss": 1.5145, + "step": 538 + }, + { + "epoch": 0.5536723163841808, + "grad_norm": 2.3272323608398438, + "learning_rate": 4.5603349828701945e-06, + "loss": 1.5472, + "step": 539 + }, + { + "epoch": 0.5546995377503852, + "grad_norm": 2.416072130203247, + "learning_rate": 4.558431671107728e-06, + "loss": 1.5842, + "step": 540 + }, + { + "epoch": 0.5546995377503852, + "eval_loss": 1.825427532196045, + "eval_runtime": 37.9905, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 540 + }, + { + "epoch": 0.5557267591165896, + "grad_norm": 2.3225271701812744, + "learning_rate": 4.556528359345261e-06, + "loss": 1.5188, + "step": 541 + }, + { + "epoch": 0.556753980482794, + "grad_norm": 2.2387030124664307, + "learning_rate": 4.5546250475827946e-06, + "loss": 1.4778, + "step": 542 + }, + { + "epoch": 0.5577812018489985, + "grad_norm": 2.3054494857788086, + "learning_rate": 4.552721735820328e-06, + "loss": 1.5229, + "step": 543 + }, + { + "epoch": 0.5588084232152029, + "grad_norm": 2.4956626892089844, + "learning_rate": 4.550818424057861e-06, + "loss": 1.5367, + "step": 544 + }, + { + "epoch": 0.5598356445814073, + "grad_norm": 2.454986810684204, + "learning_rate": 4.548915112295395e-06, + "loss": 1.5444, + "step": 545 + }, + { + "epoch": 0.5608628659476117, + "grad_norm": 2.466810703277588, + "learning_rate": 4.547011800532928e-06, + "loss": 1.4914, + "step": 546 + }, + { + "epoch": 0.5618900873138162, + "grad_norm": 2.51777982711792, + "learning_rate": 4.545108488770461e-06, + "loss": 1.5775, + "step": 547 + }, + { + "epoch": 0.5629173086800205, + "grad_norm": 2.340421199798584, + "learning_rate": 4.543205177007995e-06, + "loss": 1.5185, + "step": 548 + }, + { + "epoch": 0.5639445300462249, + "grad_norm": 2.376772880554199, + "learning_rate": 4.541301865245527e-06, + "loss": 1.5588, + "step": 549 + }, + { + "epoch": 0.5649717514124294, + "grad_norm": 2.4004335403442383, + "learning_rate": 4.539398553483061e-06, + "loss": 1.5283, + "step": 550 + }, + { + "epoch": 0.5649717514124294, + "eval_loss": 1.8294963836669922, + "eval_runtime": 37.9391, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 550 + }, + { + "epoch": 0.5659989727786338, + "grad_norm": 2.424459934234619, + "learning_rate": 4.537495241720594e-06, + "loss": 1.5576, + "step": 551 + }, + { + "epoch": 0.5670261941448382, + "grad_norm": 2.4413998126983643, + "learning_rate": 4.535591929958127e-06, + "loss": 1.5444, + "step": 552 + }, + { + "epoch": 0.5680534155110426, + "grad_norm": 2.408897876739502, + "learning_rate": 4.533688618195661e-06, + "loss": 1.5269, + "step": 553 + }, + { + "epoch": 0.5690806368772471, + "grad_norm": 2.471024990081787, + "learning_rate": 4.531785306433194e-06, + "loss": 1.5109, + "step": 554 + }, + { + "epoch": 0.5701078582434514, + "grad_norm": 2.374882459640503, + "learning_rate": 4.529881994670727e-06, + "loss": 1.5205, + "step": 555 + }, + { + "epoch": 0.5711350796096559, + "grad_norm": 2.4166030883789062, + "learning_rate": 4.527978682908261e-06, + "loss": 1.5276, + "step": 556 + }, + { + "epoch": 0.5721623009758603, + "grad_norm": 2.2681620121002197, + "learning_rate": 4.526075371145794e-06, + "loss": 1.5374, + "step": 557 + }, + { + "epoch": 0.5731895223420647, + "grad_norm": 2.3933286666870117, + "learning_rate": 4.5241720593833275e-06, + "loss": 1.5349, + "step": 558 + }, + { + "epoch": 0.5742167437082691, + "grad_norm": 2.4086451530456543, + "learning_rate": 4.52226874762086e-06, + "loss": 1.5512, + "step": 559 + }, + { + "epoch": 0.5752439650744735, + "grad_norm": 2.559021472930908, + "learning_rate": 4.520365435858393e-06, + "loss": 1.5789, + "step": 560 + }, + { + "epoch": 0.5752439650744735, + "eval_loss": 1.8243234157562256, + "eval_runtime": 37.934, + "eval_samples_per_second": 26.362, + "eval_steps_per_second": 1.107, + "step": 560 + }, + { + "epoch": 0.576271186440678, + "grad_norm": 2.483494520187378, + "learning_rate": 4.518462124095928e-06, + "loss": 1.5348, + "step": 561 + }, + { + "epoch": 0.5772984078068824, + "grad_norm": 2.4039993286132812, + "learning_rate": 4.516558812333461e-06, + "loss": 1.5201, + "step": 562 + }, + { + "epoch": 0.5783256291730868, + "grad_norm": 2.5080959796905518, + "learning_rate": 4.514655500570994e-06, + "loss": 1.5864, + "step": 563 + }, + { + "epoch": 0.5793528505392912, + "grad_norm": 2.620887279510498, + "learning_rate": 4.512752188808528e-06, + "loss": 1.5901, + "step": 564 + }, + { + "epoch": 0.5803800719054957, + "grad_norm": 2.5691096782684326, + "learning_rate": 4.510848877046061e-06, + "loss": 1.5506, + "step": 565 + }, + { + "epoch": 0.5814072932717, + "grad_norm": 2.3777835369110107, + "learning_rate": 4.5089455652835936e-06, + "loss": 1.5149, + "step": 566 + }, + { + "epoch": 0.5824345146379045, + "grad_norm": 2.2791666984558105, + "learning_rate": 4.507042253521127e-06, + "loss": 1.4771, + "step": 567 + }, + { + "epoch": 0.5834617360041089, + "grad_norm": 2.4953627586364746, + "learning_rate": 4.50513894175866e-06, + "loss": 1.556, + "step": 568 + }, + { + "epoch": 0.5844889573703133, + "grad_norm": 2.3563215732574463, + "learning_rate": 4.503235629996194e-06, + "loss": 1.549, + "step": 569 + }, + { + "epoch": 0.5855161787365177, + "grad_norm": 2.49676513671875, + "learning_rate": 4.501332318233727e-06, + "loss": 1.513, + "step": 570 + }, + { + "epoch": 0.5855161787365177, + "eval_loss": 1.8353043794631958, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 570 + }, + { + "epoch": 0.5865434001027221, + "grad_norm": 2.385275363922119, + "learning_rate": 4.49942900647126e-06, + "loss": 1.5622, + "step": 571 + }, + { + "epoch": 0.5875706214689266, + "grad_norm": 2.3089914321899414, + "learning_rate": 4.497525694708794e-06, + "loss": 1.5426, + "step": 572 + }, + { + "epoch": 0.588597842835131, + "grad_norm": 2.307323455810547, + "learning_rate": 4.495622382946327e-06, + "loss": 1.5088, + "step": 573 + }, + { + "epoch": 0.5896250642013354, + "grad_norm": 2.597288131713867, + "learning_rate": 4.4937190711838605e-06, + "loss": 1.5788, + "step": 574 + }, + { + "epoch": 0.5906522855675398, + "grad_norm": 2.3923256397247314, + "learning_rate": 4.491815759421394e-06, + "loss": 1.5038, + "step": 575 + }, + { + "epoch": 0.5916795069337443, + "grad_norm": 2.4199156761169434, + "learning_rate": 4.489912447658926e-06, + "loss": 1.5057, + "step": 576 + }, + { + "epoch": 0.5927067282999486, + "grad_norm": 2.5643393993377686, + "learning_rate": 4.48800913589646e-06, + "loss": 1.5274, + "step": 577 + }, + { + "epoch": 0.593733949666153, + "grad_norm": 2.5854485034942627, + "learning_rate": 4.486105824133993e-06, + "loss": 1.5341, + "step": 578 + }, + { + "epoch": 0.5947611710323575, + "grad_norm": 2.5363597869873047, + "learning_rate": 4.4842025123715264e-06, + "loss": 1.4953, + "step": 579 + }, + { + "epoch": 0.5957883923985619, + "grad_norm": 2.3880693912506104, + "learning_rate": 4.48229920060906e-06, + "loss": 1.4868, + "step": 580 + }, + { + "epoch": 0.5957883923985619, + "eval_loss": 1.8251410722732544, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 580 + }, + { + "epoch": 0.5968156137647663, + "grad_norm": 2.5272583961486816, + "learning_rate": 4.480395888846594e-06, + "loss": 1.5851, + "step": 581 + }, + { + "epoch": 0.5978428351309707, + "grad_norm": 2.4496943950653076, + "learning_rate": 4.478492577084127e-06, + "loss": 1.4896, + "step": 582 + }, + { + "epoch": 0.5988700564971752, + "grad_norm": 2.3996055126190186, + "learning_rate": 4.47658926532166e-06, + "loss": 1.5253, + "step": 583 + }, + { + "epoch": 0.5998972778633795, + "grad_norm": 2.5964202880859375, + "learning_rate": 4.474685953559193e-06, + "loss": 1.6294, + "step": 584 + }, + { + "epoch": 0.600924499229584, + "grad_norm": 2.5567739009857178, + "learning_rate": 4.472782641796727e-06, + "loss": 1.4834, + "step": 585 + }, + { + "epoch": 0.6019517205957884, + "grad_norm": 2.440291166305542, + "learning_rate": 4.47087933003426e-06, + "loss": 1.4966, + "step": 586 + }, + { + "epoch": 0.6029789419619929, + "grad_norm": 2.525763988494873, + "learning_rate": 4.468976018271793e-06, + "loss": 1.5005, + "step": 587 + }, + { + "epoch": 0.6040061633281972, + "grad_norm": 2.4822962284088135, + "learning_rate": 4.467072706509327e-06, + "loss": 1.5358, + "step": 588 + }, + { + "epoch": 0.6050333846944016, + "grad_norm": 2.591149091720581, + "learning_rate": 4.46516939474686e-06, + "loss": 1.4975, + "step": 589 + }, + { + "epoch": 0.6060606060606061, + "grad_norm": 2.4884238243103027, + "learning_rate": 4.463266082984393e-06, + "loss": 1.6107, + "step": 590 + }, + { + "epoch": 0.6060606060606061, + "eval_loss": 1.8383466005325317, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 590 + }, + { + "epoch": 0.6070878274268104, + "grad_norm": 2.4965732097625732, + "learning_rate": 4.461362771221927e-06, + "loss": 1.4986, + "step": 591 + }, + { + "epoch": 0.6081150487930149, + "grad_norm": 2.544792413711548, + "learning_rate": 4.45945945945946e-06, + "loss": 1.5067, + "step": 592 + }, + { + "epoch": 0.6091422701592193, + "grad_norm": 2.5726170539855957, + "learning_rate": 4.457556147696993e-06, + "loss": 1.5861, + "step": 593 + }, + { + "epoch": 0.6101694915254238, + "grad_norm": 2.493027448654175, + "learning_rate": 4.455652835934526e-06, + "loss": 1.5857, + "step": 594 + }, + { + "epoch": 0.6111967128916281, + "grad_norm": 2.470771074295044, + "learning_rate": 4.453749524172059e-06, + "loss": 1.5419, + "step": 595 + }, + { + "epoch": 0.6122239342578326, + "grad_norm": 2.568052053451538, + "learning_rate": 4.451846212409593e-06, + "loss": 1.5391, + "step": 596 + }, + { + "epoch": 0.613251155624037, + "grad_norm": 2.4346213340759277, + "learning_rate": 4.449942900647126e-06, + "loss": 1.4761, + "step": 597 + }, + { + "epoch": 0.6142783769902413, + "grad_norm": 2.418133497238159, + "learning_rate": 4.4480395888846595e-06, + "loss": 1.5383, + "step": 598 + }, + { + "epoch": 0.6153055983564458, + "grad_norm": 2.4795567989349365, + "learning_rate": 4.446136277122193e-06, + "loss": 1.4777, + "step": 599 + }, + { + "epoch": 0.6163328197226502, + "grad_norm": 2.5122742652893066, + "learning_rate": 4.444232965359726e-06, + "loss": 1.5353, + "step": 600 + }, + { + "epoch": 0.6163328197226502, + "eval_loss": 1.8325241804122925, + "eval_runtime": 37.9546, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 600 + }, + { + "epoch": 0.6173600410888547, + "grad_norm": 2.604067325592041, + "learning_rate": 4.4423296535972596e-06, + "loss": 1.489, + "step": 601 + }, + { + "epoch": 0.618387262455059, + "grad_norm": 2.3741753101348877, + "learning_rate": 4.440426341834793e-06, + "loss": 1.532, + "step": 602 + }, + { + "epoch": 0.6194144838212635, + "grad_norm": 2.5373036861419678, + "learning_rate": 4.438523030072326e-06, + "loss": 1.561, + "step": 603 + }, + { + "epoch": 0.6204417051874679, + "grad_norm": 2.2958099842071533, + "learning_rate": 4.43661971830986e-06, + "loss": 1.5025, + "step": 604 + }, + { + "epoch": 0.6214689265536724, + "grad_norm": 2.417006015777588, + "learning_rate": 4.434716406547393e-06, + "loss": 1.4977, + "step": 605 + }, + { + "epoch": 0.6224961479198767, + "grad_norm": 2.335296392440796, + "learning_rate": 4.432813094784926e-06, + "loss": 1.5504, + "step": 606 + }, + { + "epoch": 0.6235233692860811, + "grad_norm": 2.4086828231811523, + "learning_rate": 4.43090978302246e-06, + "loss": 1.5217, + "step": 607 + }, + { + "epoch": 0.6245505906522856, + "grad_norm": 2.3816819190979004, + "learning_rate": 4.429006471259993e-06, + "loss": 1.5126, + "step": 608 + }, + { + "epoch": 0.6255778120184899, + "grad_norm": 2.307671546936035, + "learning_rate": 4.4271031594975265e-06, + "loss": 1.4876, + "step": 609 + }, + { + "epoch": 0.6266050333846944, + "grad_norm": 2.355837821960449, + "learning_rate": 4.425199847735059e-06, + "loss": 1.4827, + "step": 610 + }, + { + "epoch": 0.6266050333846944, + "eval_loss": 1.8269455432891846, + "eval_runtime": 37.6459, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 610 + }, + { + "epoch": 0.6276322547508988, + "grad_norm": 2.558631658554077, + "learning_rate": 4.423296535972592e-06, + "loss": 1.5172, + "step": 611 + }, + { + "epoch": 0.6286594761171033, + "grad_norm": 2.3170175552368164, + "learning_rate": 4.421393224210126e-06, + "loss": 1.4784, + "step": 612 + }, + { + "epoch": 0.6296866974833076, + "grad_norm": 2.4064524173736572, + "learning_rate": 4.419489912447659e-06, + "loss": 1.5616, + "step": 613 + }, + { + "epoch": 0.6307139188495121, + "grad_norm": 2.3719160556793213, + "learning_rate": 4.4175866006851924e-06, + "loss": 1.5292, + "step": 614 + }, + { + "epoch": 0.6317411402157165, + "grad_norm": 2.3808202743530273, + "learning_rate": 4.415683288922726e-06, + "loss": 1.5122, + "step": 615 + }, + { + "epoch": 0.632768361581921, + "grad_norm": 2.4208037853240967, + "learning_rate": 4.413779977160259e-06, + "loss": 1.5088, + "step": 616 + }, + { + "epoch": 0.6337955829481253, + "grad_norm": 2.5558979511260986, + "learning_rate": 4.4118766653977925e-06, + "loss": 1.5678, + "step": 617 + }, + { + "epoch": 0.6348228043143297, + "grad_norm": 2.443908452987671, + "learning_rate": 4.409973353635326e-06, + "loss": 1.5234, + "step": 618 + }, + { + "epoch": 0.6358500256805342, + "grad_norm": 2.420408010482788, + "learning_rate": 4.408070041872859e-06, + "loss": 1.4708, + "step": 619 + }, + { + "epoch": 0.6368772470467385, + "grad_norm": 2.3884904384613037, + "learning_rate": 4.406166730110393e-06, + "loss": 1.4821, + "step": 620 + }, + { + "epoch": 0.6368772470467385, + "eval_loss": 1.8217289447784424, + "eval_runtime": 37.7936, + "eval_samples_per_second": 26.459, + "eval_steps_per_second": 1.111, + "step": 620 + }, + { + "epoch": 0.637904468412943, + "grad_norm": 2.313720703125, + "learning_rate": 4.404263418347925e-06, + "loss": 1.4962, + "step": 621 + }, + { + "epoch": 0.6389316897791474, + "grad_norm": 2.3173866271972656, + "learning_rate": 4.402360106585459e-06, + "loss": 1.4857, + "step": 622 + }, + { + "epoch": 0.6399589111453519, + "grad_norm": 2.541757822036743, + "learning_rate": 4.400456794822993e-06, + "loss": 1.5177, + "step": 623 + }, + { + "epoch": 0.6409861325115562, + "grad_norm": 2.4382357597351074, + "learning_rate": 4.398553483060526e-06, + "loss": 1.4718, + "step": 624 + }, + { + "epoch": 0.6420133538777607, + "grad_norm": 2.2988059520721436, + "learning_rate": 4.396650171298059e-06, + "loss": 1.4883, + "step": 625 + }, + { + "epoch": 0.6430405752439651, + "grad_norm": 2.4817495346069336, + "learning_rate": 4.394746859535593e-06, + "loss": 1.4823, + "step": 626 + }, + { + "epoch": 0.6440677966101694, + "grad_norm": 2.4578585624694824, + "learning_rate": 4.392843547773125e-06, + "loss": 1.4882, + "step": 627 + }, + { + "epoch": 0.6450950179763739, + "grad_norm": 2.184972047805786, + "learning_rate": 4.390940236010659e-06, + "loss": 1.514, + "step": 628 + }, + { + "epoch": 0.6461222393425783, + "grad_norm": 2.4040279388427734, + "learning_rate": 4.389036924248192e-06, + "loss": 1.5343, + "step": 629 + }, + { + "epoch": 0.6471494607087828, + "grad_norm": 2.3676459789276123, + "learning_rate": 4.387133612485725e-06, + "loss": 1.5007, + "step": 630 + }, + { + "epoch": 0.6471494607087828, + "eval_loss": 1.8106164932250977, + "eval_runtime": 37.9383, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 630 + }, + { + "epoch": 0.6481766820749871, + "grad_norm": 2.234689235687256, + "learning_rate": 4.385230300723259e-06, + "loss": 1.5064, + "step": 631 + }, + { + "epoch": 0.6492039034411916, + "grad_norm": 2.281742572784424, + "learning_rate": 4.383326988960792e-06, + "loss": 1.4777, + "step": 632 + }, + { + "epoch": 0.650231124807396, + "grad_norm": 2.4314322471618652, + "learning_rate": 4.3814236771983255e-06, + "loss": 1.5331, + "step": 633 + }, + { + "epoch": 0.6512583461736005, + "grad_norm": 2.4092588424682617, + "learning_rate": 4.379520365435859e-06, + "loss": 1.5122, + "step": 634 + }, + { + "epoch": 0.6522855675398048, + "grad_norm": 2.517362117767334, + "learning_rate": 4.377617053673392e-06, + "loss": 1.52, + "step": 635 + }, + { + "epoch": 0.6533127889060092, + "grad_norm": 2.4784109592437744, + "learning_rate": 4.3757137419109256e-06, + "loss": 1.5771, + "step": 636 + }, + { + "epoch": 0.6543400102722137, + "grad_norm": 2.32181715965271, + "learning_rate": 4.373810430148459e-06, + "loss": 1.4718, + "step": 637 + }, + { + "epoch": 0.655367231638418, + "grad_norm": 2.5146260261535645, + "learning_rate": 4.3719071183859914e-06, + "loss": 1.5307, + "step": 638 + }, + { + "epoch": 0.6563944530046225, + "grad_norm": 2.2364494800567627, + "learning_rate": 4.370003806623525e-06, + "loss": 1.4803, + "step": 639 + }, + { + "epoch": 0.6574216743708269, + "grad_norm": 2.3735806941986084, + "learning_rate": 4.368100494861058e-06, + "loss": 1.5249, + "step": 640 + }, + { + "epoch": 0.6574216743708269, + "eval_loss": 1.8134880065917969, + "eval_runtime": 38.0376, + "eval_samples_per_second": 26.29, + "eval_steps_per_second": 1.104, + "step": 640 + }, + { + "epoch": 0.6584488957370314, + "grad_norm": 2.4600815773010254, + "learning_rate": 4.3661971830985915e-06, + "loss": 1.4893, + "step": 641 + }, + { + "epoch": 0.6594761171032357, + "grad_norm": 2.4106783866882324, + "learning_rate": 4.364293871336126e-06, + "loss": 1.5163, + "step": 642 + }, + { + "epoch": 0.6605033384694402, + "grad_norm": 2.276624917984009, + "learning_rate": 4.362390559573659e-06, + "loss": 1.482, + "step": 643 + }, + { + "epoch": 0.6615305598356446, + "grad_norm": 2.336341381072998, + "learning_rate": 4.3604872478111925e-06, + "loss": 1.4152, + "step": 644 + }, + { + "epoch": 0.662557781201849, + "grad_norm": 2.3329105377197266, + "learning_rate": 4.358583936048725e-06, + "loss": 1.5411, + "step": 645 + }, + { + "epoch": 0.6635850025680534, + "grad_norm": 2.2877161502838135, + "learning_rate": 4.356680624286258e-06, + "loss": 1.4997, + "step": 646 + }, + { + "epoch": 0.6646122239342578, + "grad_norm": 2.410820484161377, + "learning_rate": 4.354777312523792e-06, + "loss": 1.5446, + "step": 647 + }, + { + "epoch": 0.6656394453004623, + "grad_norm": 2.403038501739502, + "learning_rate": 4.352874000761325e-06, + "loss": 1.4916, + "step": 648 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 2.3261778354644775, + "learning_rate": 4.3509706889988584e-06, + "loss": 1.5561, + "step": 649 + }, + { + "epoch": 0.6676938880328711, + "grad_norm": 2.4512174129486084, + "learning_rate": 4.349067377236392e-06, + "loss": 1.556, + "step": 650 + }, + { + "epoch": 0.6676938880328711, + "eval_loss": 1.8238279819488525, + "eval_runtime": 37.9988, + "eval_samples_per_second": 26.317, + "eval_steps_per_second": 1.105, + "step": 650 + }, + { + "epoch": 0.6687211093990755, + "grad_norm": 2.4523823261260986, + "learning_rate": 4.347164065473925e-06, + "loss": 1.5676, + "step": 651 + }, + { + "epoch": 0.66974833076528, + "grad_norm": 2.3770365715026855, + "learning_rate": 4.3452607537114585e-06, + "loss": 1.4815, + "step": 652 + }, + { + "epoch": 0.6707755521314843, + "grad_norm": 2.2440760135650635, + "learning_rate": 4.343357441948992e-06, + "loss": 1.5141, + "step": 653 + }, + { + "epoch": 0.6718027734976888, + "grad_norm": 2.3341948986053467, + "learning_rate": 4.341454130186525e-06, + "loss": 1.4859, + "step": 654 + }, + { + "epoch": 0.6728299948638932, + "grad_norm": 2.507354497909546, + "learning_rate": 4.339550818424058e-06, + "loss": 1.4734, + "step": 655 + }, + { + "epoch": 0.6738572162300975, + "grad_norm": 2.395954132080078, + "learning_rate": 4.337647506661591e-06, + "loss": 1.5069, + "step": 656 + }, + { + "epoch": 0.674884437596302, + "grad_norm": 2.362211227416992, + "learning_rate": 4.3357441948991245e-06, + "loss": 1.5299, + "step": 657 + }, + { + "epoch": 0.6759116589625064, + "grad_norm": 2.460008144378662, + "learning_rate": 4.333840883136658e-06, + "loss": 1.4795, + "step": 658 + }, + { + "epoch": 0.6769388803287109, + "grad_norm": 2.3578217029571533, + "learning_rate": 4.331937571374191e-06, + "loss": 1.4906, + "step": 659 + }, + { + "epoch": 0.6779661016949152, + "grad_norm": 2.398843288421631, + "learning_rate": 4.3300342596117246e-06, + "loss": 1.4986, + "step": 660 + }, + { + "epoch": 0.6779661016949152, + "eval_loss": 1.8206881284713745, + "eval_runtime": 38.044, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 660 + }, + { + "epoch": 0.6789933230611197, + "grad_norm": 2.3587260246276855, + "learning_rate": 4.328130947849258e-06, + "loss": 1.461, + "step": 661 + }, + { + "epoch": 0.6800205444273241, + "grad_norm": 2.311952590942383, + "learning_rate": 4.326227636086791e-06, + "loss": 1.5041, + "step": 662 + }, + { + "epoch": 0.6810477657935285, + "grad_norm": 2.376188278198242, + "learning_rate": 4.324324324324325e-06, + "loss": 1.5042, + "step": 663 + }, + { + "epoch": 0.6820749871597329, + "grad_norm": 2.5807230472564697, + "learning_rate": 4.322421012561858e-06, + "loss": 1.4935, + "step": 664 + }, + { + "epoch": 0.6831022085259374, + "grad_norm": 2.368128538131714, + "learning_rate": 4.320517700799391e-06, + "loss": 1.5851, + "step": 665 + }, + { + "epoch": 0.6841294298921418, + "grad_norm": 2.4517524242401123, + "learning_rate": 4.318614389036925e-06, + "loss": 1.5528, + "step": 666 + }, + { + "epoch": 0.6851566512583461, + "grad_norm": 2.5436811447143555, + "learning_rate": 4.316711077274458e-06, + "loss": 1.4633, + "step": 667 + }, + { + "epoch": 0.6861838726245506, + "grad_norm": 2.4287915229797363, + "learning_rate": 4.3148077655119915e-06, + "loss": 1.5454, + "step": 668 + }, + { + "epoch": 0.687211093990755, + "grad_norm": 2.386528730392456, + "learning_rate": 4.312904453749525e-06, + "loss": 1.5116, + "step": 669 + }, + { + "epoch": 0.6882383153569595, + "grad_norm": 2.6602251529693604, + "learning_rate": 4.311001141987058e-06, + "loss": 1.5199, + "step": 670 + }, + { + "epoch": 0.6882383153569595, + "eval_loss": 1.8252663612365723, + "eval_runtime": 38.0791, + "eval_samples_per_second": 26.261, + "eval_steps_per_second": 1.103, + "step": 670 + }, + { + "epoch": 0.6892655367231638, + "grad_norm": 2.351128339767456, + "learning_rate": 4.3090978302245916e-06, + "loss": 1.507, + "step": 671 + }, + { + "epoch": 0.6902927580893683, + "grad_norm": 2.393441677093506, + "learning_rate": 4.307194518462124e-06, + "loss": 1.579, + "step": 672 + }, + { + "epoch": 0.6913199794555727, + "grad_norm": 2.3730578422546387, + "learning_rate": 4.3052912066996574e-06, + "loss": 1.5446, + "step": 673 + }, + { + "epoch": 0.6923472008217771, + "grad_norm": 2.2522945404052734, + "learning_rate": 4.303387894937191e-06, + "loss": 1.4705, + "step": 674 + }, + { + "epoch": 0.6933744221879815, + "grad_norm": 2.384563446044922, + "learning_rate": 4.301484583174724e-06, + "loss": 1.4636, + "step": 675 + }, + { + "epoch": 0.6944016435541859, + "grad_norm": 2.2879116535186768, + "learning_rate": 4.2995812714122575e-06, + "loss": 1.4914, + "step": 676 + }, + { + "epoch": 0.6954288649203904, + "grad_norm": 2.494788646697998, + "learning_rate": 4.297677959649791e-06, + "loss": 1.5126, + "step": 677 + }, + { + "epoch": 0.6964560862865947, + "grad_norm": 2.700012683868408, + "learning_rate": 4.295774647887324e-06, + "loss": 1.5362, + "step": 678 + }, + { + "epoch": 0.6974833076527992, + "grad_norm": 2.5130302906036377, + "learning_rate": 4.293871336124858e-06, + "loss": 1.5075, + "step": 679 + }, + { + "epoch": 0.6985105290190036, + "grad_norm": 2.486236572265625, + "learning_rate": 4.291968024362391e-06, + "loss": 1.5338, + "step": 680 + }, + { + "epoch": 0.6985105290190036, + "eval_loss": 1.8229657411575317, + "eval_runtime": 38.0642, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 680 + }, + { + "epoch": 0.699537750385208, + "grad_norm": 2.441683530807495, + "learning_rate": 4.290064712599924e-06, + "loss": 1.537, + "step": 681 + }, + { + "epoch": 0.7005649717514124, + "grad_norm": 2.396798610687256, + "learning_rate": 4.288161400837458e-06, + "loss": 1.4553, + "step": 682 + }, + { + "epoch": 0.7015921931176169, + "grad_norm": 2.3636221885681152, + "learning_rate": 4.286258089074991e-06, + "loss": 1.5138, + "step": 683 + }, + { + "epoch": 0.7026194144838213, + "grad_norm": 2.4584708213806152, + "learning_rate": 4.284354777312524e-06, + "loss": 1.4725, + "step": 684 + }, + { + "epoch": 0.7036466358500256, + "grad_norm": 2.850332498550415, + "learning_rate": 4.282451465550058e-06, + "loss": 1.5075, + "step": 685 + }, + { + "epoch": 0.7046738572162301, + "grad_norm": 2.6345887184143066, + "learning_rate": 4.280548153787591e-06, + "loss": 1.5054, + "step": 686 + }, + { + "epoch": 0.7057010785824345, + "grad_norm": 2.3535311222076416, + "learning_rate": 4.2786448420251245e-06, + "loss": 1.4607, + "step": 687 + }, + { + "epoch": 0.706728299948639, + "grad_norm": 2.490431547164917, + "learning_rate": 4.276741530262658e-06, + "loss": 1.4852, + "step": 688 + }, + { + "epoch": 0.7077555213148433, + "grad_norm": 2.4353036880493164, + "learning_rate": 4.27483821850019e-06, + "loss": 1.5503, + "step": 689 + }, + { + "epoch": 0.7087827426810478, + "grad_norm": 2.3356940746307373, + "learning_rate": 4.272934906737724e-06, + "loss": 1.5416, + "step": 690 + }, + { + "epoch": 0.7087827426810478, + "eval_loss": 1.8164218664169312, + "eval_runtime": 38.0592, + "eval_samples_per_second": 26.275, + "eval_steps_per_second": 1.104, + "step": 690 + }, + { + "epoch": 0.7098099640472522, + "grad_norm": 2.4772775173187256, + "learning_rate": 4.271031594975257e-06, + "loss": 1.5392, + "step": 691 + }, + { + "epoch": 0.7108371854134566, + "grad_norm": 2.2921438217163086, + "learning_rate": 4.2691282832127905e-06, + "loss": 1.4935, + "step": 692 + }, + { + "epoch": 0.711864406779661, + "grad_norm": 2.3176305294036865, + "learning_rate": 4.267224971450324e-06, + "loss": 1.4451, + "step": 693 + }, + { + "epoch": 0.7128916281458655, + "grad_norm": 2.6825685501098633, + "learning_rate": 4.265321659687857e-06, + "loss": 1.4878, + "step": 694 + }, + { + "epoch": 0.7139188495120699, + "grad_norm": 2.6141414642333984, + "learning_rate": 4.2634183479253906e-06, + "loss": 1.4695, + "step": 695 + }, + { + "epoch": 0.7149460708782742, + "grad_norm": 2.5179429054260254, + "learning_rate": 4.261515036162924e-06, + "loss": 1.492, + "step": 696 + }, + { + "epoch": 0.7159732922444787, + "grad_norm": 2.4744884967803955, + "learning_rate": 4.259611724400457e-06, + "loss": 1.4752, + "step": 697 + }, + { + "epoch": 0.7170005136106831, + "grad_norm": 2.4911048412323, + "learning_rate": 4.257708412637991e-06, + "loss": 1.423, + "step": 698 + }, + { + "epoch": 0.7180277349768875, + "grad_norm": 2.5116214752197266, + "learning_rate": 4.255805100875523e-06, + "loss": 1.4535, + "step": 699 + }, + { + "epoch": 0.7190549563430919, + "grad_norm": 2.476651430130005, + "learning_rate": 4.2539017891130565e-06, + "loss": 1.5349, + "step": 700 + }, + { + "epoch": 0.7190549563430919, + "eval_loss": 1.8069179058074951, + "eval_runtime": 38.0292, + "eval_samples_per_second": 26.296, + "eval_steps_per_second": 1.104, + "step": 700 + }, + { + "epoch": 0.7200821777092964, + "grad_norm": 2.5234575271606445, + "learning_rate": 4.25199847735059e-06, + "loss": 1.4596, + "step": 701 + }, + { + "epoch": 0.7211093990755008, + "grad_norm": 2.33290433883667, + "learning_rate": 4.250095165588123e-06, + "loss": 1.5147, + "step": 702 + }, + { + "epoch": 0.7221366204417052, + "grad_norm": 2.479883909225464, + "learning_rate": 4.2481918538256575e-06, + "loss": 1.47, + "step": 703 + }, + { + "epoch": 0.7231638418079096, + "grad_norm": 2.432253837585449, + "learning_rate": 4.246288542063191e-06, + "loss": 1.5152, + "step": 704 + }, + { + "epoch": 0.724191063174114, + "grad_norm": 2.358304738998413, + "learning_rate": 4.244385230300724e-06, + "loss": 1.4562, + "step": 705 + }, + { + "epoch": 0.7252182845403184, + "grad_norm": 2.480877161026001, + "learning_rate": 4.242481918538257e-06, + "loss": 1.4906, + "step": 706 + }, + { + "epoch": 0.7262455059065228, + "grad_norm": 2.4287192821502686, + "learning_rate": 4.24057860677579e-06, + "loss": 1.5294, + "step": 707 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 2.4782888889312744, + "learning_rate": 4.2386752950133234e-06, + "loss": 1.4767, + "step": 708 + }, + { + "epoch": 0.7282999486389317, + "grad_norm": 2.4948253631591797, + "learning_rate": 4.236771983250857e-06, + "loss": 1.4664, + "step": 709 + }, + { + "epoch": 0.7293271700051361, + "grad_norm": 2.4625205993652344, + "learning_rate": 4.23486867148839e-06, + "loss": 1.495, + "step": 710 + }, + { + "epoch": 0.7293271700051361, + "eval_loss": 1.8140079975128174, + "eval_runtime": 37.6964, + "eval_samples_per_second": 26.528, + "eval_steps_per_second": 1.114, + "step": 710 + }, + { + "epoch": 0.7303543913713405, + "grad_norm": 2.669027805328369, + "learning_rate": 4.2329653597259235e-06, + "loss": 1.5204, + "step": 711 + }, + { + "epoch": 0.731381612737545, + "grad_norm": 2.5024707317352295, + "learning_rate": 4.231062047963457e-06, + "loss": 1.5283, + "step": 712 + }, + { + "epoch": 0.7324088341037494, + "grad_norm": 2.498258590698242, + "learning_rate": 4.22915873620099e-06, + "loss": 1.5139, + "step": 713 + }, + { + "epoch": 0.7334360554699538, + "grad_norm": 2.4305667877197266, + "learning_rate": 4.227255424438524e-06, + "loss": 1.5337, + "step": 714 + }, + { + "epoch": 0.7344632768361582, + "grad_norm": 2.3153419494628906, + "learning_rate": 4.225352112676057e-06, + "loss": 1.4646, + "step": 715 + }, + { + "epoch": 0.7354904982023626, + "grad_norm": 2.2835583686828613, + "learning_rate": 4.2234488009135895e-06, + "loss": 1.4328, + "step": 716 + }, + { + "epoch": 0.736517719568567, + "grad_norm": 2.401418924331665, + "learning_rate": 4.221545489151123e-06, + "loss": 1.4685, + "step": 717 + }, + { + "epoch": 0.7375449409347714, + "grad_norm": 2.454864978790283, + "learning_rate": 4.219642177388656e-06, + "loss": 1.4925, + "step": 718 + }, + { + "epoch": 0.7385721623009759, + "grad_norm": 2.2332873344421387, + "learning_rate": 4.2177388656261896e-06, + "loss": 1.4481, + "step": 719 + }, + { + "epoch": 0.7395993836671803, + "grad_norm": 2.46628475189209, + "learning_rate": 4.215835553863723e-06, + "loss": 1.4173, + "step": 720 + }, + { + "epoch": 0.7395993836671803, + "eval_loss": 1.817592740058899, + "eval_runtime": 37.8086, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 720 + }, + { + "epoch": 0.7406266050333847, + "grad_norm": 2.346245765686035, + "learning_rate": 4.213932242101256e-06, + "loss": 1.4688, + "step": 721 + }, + { + "epoch": 0.7416538263995891, + "grad_norm": 2.2708640098571777, + "learning_rate": 4.21202893033879e-06, + "loss": 1.4937, + "step": 722 + }, + { + "epoch": 0.7426810477657936, + "grad_norm": 2.3716516494750977, + "learning_rate": 4.210125618576323e-06, + "loss": 1.4869, + "step": 723 + }, + { + "epoch": 0.743708269131998, + "grad_norm": 2.443058490753174, + "learning_rate": 4.208222306813856e-06, + "loss": 1.542, + "step": 724 + }, + { + "epoch": 0.7447354904982023, + "grad_norm": 2.327270984649658, + "learning_rate": 4.20631899505139e-06, + "loss": 1.3933, + "step": 725 + }, + { + "epoch": 0.7457627118644068, + "grad_norm": 2.3734512329101562, + "learning_rate": 4.204415683288923e-06, + "loss": 1.5109, + "step": 726 + }, + { + "epoch": 0.7467899332306112, + "grad_norm": 2.3631174564361572, + "learning_rate": 4.2025123715264565e-06, + "loss": 1.474, + "step": 727 + }, + { + "epoch": 0.7478171545968156, + "grad_norm": 2.306068181991577, + "learning_rate": 4.20060905976399e-06, + "loss": 1.42, + "step": 728 + }, + { + "epoch": 0.74884437596302, + "grad_norm": 2.391889810562134, + "learning_rate": 4.198705748001523e-06, + "loss": 1.5178, + "step": 729 + }, + { + "epoch": 0.7498715973292245, + "grad_norm": 2.3151705265045166, + "learning_rate": 4.1968024362390566e-06, + "loss": 1.4808, + "step": 730 + }, + { + "epoch": 0.7498715973292245, + "eval_loss": 1.8157070875167847, + "eval_runtime": 37.8799, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 730 + }, + { + "epoch": 0.7508988186954288, + "grad_norm": 2.336479902267456, + "learning_rate": 4.19489912447659e-06, + "loss": 1.5438, + "step": 731 + }, + { + "epoch": 0.7519260400616333, + "grad_norm": 2.338242292404175, + "learning_rate": 4.192995812714123e-06, + "loss": 1.4869, + "step": 732 + }, + { + "epoch": 0.7529532614278377, + "grad_norm": 2.268650770187378, + "learning_rate": 4.191092500951656e-06, + "loss": 1.4787, + "step": 733 + }, + { + "epoch": 0.7539804827940421, + "grad_norm": 2.3010005950927734, + "learning_rate": 4.189189189189189e-06, + "loss": 1.4711, + "step": 734 + }, + { + "epoch": 0.7550077041602465, + "grad_norm": 2.54770827293396, + "learning_rate": 4.1872858774267225e-06, + "loss": 1.5351, + "step": 735 + }, + { + "epoch": 0.7560349255264509, + "grad_norm": 2.3412787914276123, + "learning_rate": 4.185382565664256e-06, + "loss": 1.4655, + "step": 736 + }, + { + "epoch": 0.7570621468926554, + "grad_norm": 2.266413927078247, + "learning_rate": 4.183479253901789e-06, + "loss": 1.4402, + "step": 737 + }, + { + "epoch": 0.7580893682588598, + "grad_norm": 2.2785308361053467, + "learning_rate": 4.181575942139323e-06, + "loss": 1.4874, + "step": 738 + }, + { + "epoch": 0.7591165896250642, + "grad_norm": 2.2710752487182617, + "learning_rate": 4.179672630376856e-06, + "loss": 1.4644, + "step": 739 + }, + { + "epoch": 0.7601438109912686, + "grad_norm": 2.2460498809814453, + "learning_rate": 4.177769318614389e-06, + "loss": 1.5267, + "step": 740 + }, + { + "epoch": 0.7601438109912686, + "eval_loss": 1.8187947273254395, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 740 + }, + { + "epoch": 0.7611710323574731, + "grad_norm": 2.3589370250701904, + "learning_rate": 4.175866006851923e-06, + "loss": 1.5115, + "step": 741 + }, + { + "epoch": 0.7621982537236774, + "grad_norm": 2.2818758487701416, + "learning_rate": 4.173962695089456e-06, + "loss": 1.4945, + "step": 742 + }, + { + "epoch": 0.7632254750898819, + "grad_norm": 2.338085889816284, + "learning_rate": 4.1720593833269894e-06, + "loss": 1.5076, + "step": 743 + }, + { + "epoch": 0.7642526964560863, + "grad_norm": 2.3874809741973877, + "learning_rate": 4.170156071564523e-06, + "loss": 1.5665, + "step": 744 + }, + { + "epoch": 0.7652799178222907, + "grad_norm": 2.2418739795684814, + "learning_rate": 4.168252759802056e-06, + "loss": 1.4827, + "step": 745 + }, + { + "epoch": 0.7663071391884951, + "grad_norm": 2.441315174102783, + "learning_rate": 4.1663494480395895e-06, + "loss": 1.5171, + "step": 746 + }, + { + "epoch": 0.7673343605546995, + "grad_norm": 2.3201963901519775, + "learning_rate": 4.164446136277123e-06, + "loss": 1.5153, + "step": 747 + }, + { + "epoch": 0.768361581920904, + "grad_norm": 2.3395135402679443, + "learning_rate": 4.162542824514656e-06, + "loss": 1.4911, + "step": 748 + }, + { + "epoch": 0.7693888032871083, + "grad_norm": 2.3455941677093506, + "learning_rate": 4.16063951275219e-06, + "loss": 1.4493, + "step": 749 + }, + { + "epoch": 0.7704160246533128, + "grad_norm": 2.3737382888793945, + "learning_rate": 4.158736200989722e-06, + "loss": 1.4893, + "step": 750 + }, + { + "epoch": 0.7704160246533128, + "eval_loss": 1.8153141736984253, + "eval_runtime": 37.9216, + "eval_samples_per_second": 26.37, + "eval_steps_per_second": 1.108, + "step": 750 + }, + { + "epoch": 0.7714432460195172, + "grad_norm": 2.503803014755249, + "learning_rate": 4.1568328892272555e-06, + "loss": 1.499, + "step": 751 + }, + { + "epoch": 0.7724704673857217, + "grad_norm": 2.469083547592163, + "learning_rate": 4.154929577464789e-06, + "loss": 1.4984, + "step": 752 + }, + { + "epoch": 0.773497688751926, + "grad_norm": 2.387991428375244, + "learning_rate": 4.153026265702322e-06, + "loss": 1.4283, + "step": 753 + }, + { + "epoch": 0.7745249101181304, + "grad_norm": 2.3355393409729004, + "learning_rate": 4.1511229539398556e-06, + "loss": 1.4869, + "step": 754 + }, + { + "epoch": 0.7755521314843349, + "grad_norm": 2.629345417022705, + "learning_rate": 4.149219642177389e-06, + "loss": 1.4885, + "step": 755 + }, + { + "epoch": 0.7765793528505393, + "grad_norm": 2.458944797515869, + "learning_rate": 4.147316330414922e-06, + "loss": 1.415, + "step": 756 + }, + { + "epoch": 0.7776065742167437, + "grad_norm": 2.40051531791687, + "learning_rate": 4.145413018652456e-06, + "loss": 1.5129, + "step": 757 + }, + { + "epoch": 0.7786337955829481, + "grad_norm": 2.5249011516571045, + "learning_rate": 4.143509706889989e-06, + "loss": 1.5002, + "step": 758 + }, + { + "epoch": 0.7796610169491526, + "grad_norm": 2.413422107696533, + "learning_rate": 4.141606395127522e-06, + "loss": 1.4837, + "step": 759 + }, + { + "epoch": 0.7806882383153569, + "grad_norm": 2.3438029289245605, + "learning_rate": 4.139703083365056e-06, + "loss": 1.5213, + "step": 760 + }, + { + "epoch": 0.7806882383153569, + "eval_loss": 1.8037596940994263, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 760 + }, + { + "epoch": 0.7817154596815614, + "grad_norm": 2.326828718185425, + "learning_rate": 4.137799771602588e-06, + "loss": 1.5031, + "step": 761 + }, + { + "epoch": 0.7827426810477658, + "grad_norm": 2.4902493953704834, + "learning_rate": 4.135896459840122e-06, + "loss": 1.4602, + "step": 762 + }, + { + "epoch": 0.7837699024139703, + "grad_norm": 2.488004684448242, + "learning_rate": 4.133993148077656e-06, + "loss": 1.4764, + "step": 763 + }, + { + "epoch": 0.7847971237801746, + "grad_norm": 2.5515761375427246, + "learning_rate": 4.132089836315189e-06, + "loss": 1.4946, + "step": 764 + }, + { + "epoch": 0.785824345146379, + "grad_norm": 2.556278705596924, + "learning_rate": 4.1301865245527226e-06, + "loss": 1.4338, + "step": 765 + }, + { + "epoch": 0.7868515665125835, + "grad_norm": 2.3017969131469727, + "learning_rate": 4.128283212790256e-06, + "loss": 1.4053, + "step": 766 + }, + { + "epoch": 0.7878787878787878, + "grad_norm": 2.4371304512023926, + "learning_rate": 4.1263799010277884e-06, + "loss": 1.5238, + "step": 767 + }, + { + "epoch": 0.7889060092449923, + "grad_norm": 2.6256046295166016, + "learning_rate": 4.124476589265322e-06, + "loss": 1.4672, + "step": 768 + }, + { + "epoch": 0.7899332306111967, + "grad_norm": 2.3725669384002686, + "learning_rate": 4.122573277502855e-06, + "loss": 1.4677, + "step": 769 + }, + { + "epoch": 0.7909604519774012, + "grad_norm": 2.2928454875946045, + "learning_rate": 4.1206699657403885e-06, + "loss": 1.4927, + "step": 770 + }, + { + "epoch": 0.7909604519774012, + "eval_loss": 1.792726755142212, + "eval_runtime": 37.9273, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 770 + }, + { + "epoch": 0.7919876733436055, + "grad_norm": 2.315674066543579, + "learning_rate": 4.118766653977922e-06, + "loss": 1.4632, + "step": 771 + }, + { + "epoch": 0.79301489470981, + "grad_norm": 2.418849468231201, + "learning_rate": 4.116863342215455e-06, + "loss": 1.446, + "step": 772 + }, + { + "epoch": 0.7940421160760144, + "grad_norm": 2.3209879398345947, + "learning_rate": 4.114960030452989e-06, + "loss": 1.4755, + "step": 773 + }, + { + "epoch": 0.7950693374422187, + "grad_norm": 2.2322115898132324, + "learning_rate": 4.113056718690522e-06, + "loss": 1.4149, + "step": 774 + }, + { + "epoch": 0.7960965588084232, + "grad_norm": 2.6517558097839355, + "learning_rate": 4.111153406928055e-06, + "loss": 1.5323, + "step": 775 + }, + { + "epoch": 0.7971237801746276, + "grad_norm": 2.4602737426757812, + "learning_rate": 4.109250095165589e-06, + "loss": 1.5179, + "step": 776 + }, + { + "epoch": 0.7981510015408321, + "grad_norm": 2.4015536308288574, + "learning_rate": 4.107346783403122e-06, + "loss": 1.4711, + "step": 777 + }, + { + "epoch": 0.7991782229070364, + "grad_norm": 2.4367337226867676, + "learning_rate": 4.105443471640655e-06, + "loss": 1.4425, + "step": 778 + }, + { + "epoch": 0.8002054442732409, + "grad_norm": 2.5161020755767822, + "learning_rate": 4.103540159878188e-06, + "loss": 1.4832, + "step": 779 + }, + { + "epoch": 0.8012326656394453, + "grad_norm": 2.483154773712158, + "learning_rate": 4.101636848115721e-06, + "loss": 1.4508, + "step": 780 + }, + { + "epoch": 0.8012326656394453, + "eval_loss": 1.792397379875183, + "eval_runtime": 37.8937, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 780 + }, + { + "epoch": 0.8022598870056498, + "grad_norm": 2.4649806022644043, + "learning_rate": 4.099733536353255e-06, + "loss": 1.4919, + "step": 781 + }, + { + "epoch": 0.8032871083718541, + "grad_norm": 2.578634262084961, + "learning_rate": 4.097830224590788e-06, + "loss": 1.565, + "step": 782 + }, + { + "epoch": 0.8043143297380585, + "grad_norm": 2.444427967071533, + "learning_rate": 4.095926912828322e-06, + "loss": 1.5309, + "step": 783 + }, + { + "epoch": 0.805341551104263, + "grad_norm": 2.32051420211792, + "learning_rate": 4.094023601065855e-06, + "loss": 1.4777, + "step": 784 + }, + { + "epoch": 0.8063687724704673, + "grad_norm": 2.3207180500030518, + "learning_rate": 4.092120289303388e-06, + "loss": 1.4863, + "step": 785 + }, + { + "epoch": 0.8073959938366718, + "grad_norm": 2.1404120922088623, + "learning_rate": 4.0902169775409215e-06, + "loss": 1.437, + "step": 786 + }, + { + "epoch": 0.8084232152028762, + "grad_norm": 2.2980856895446777, + "learning_rate": 4.088313665778455e-06, + "loss": 1.4571, + "step": 787 + }, + { + "epoch": 0.8094504365690807, + "grad_norm": 2.2555243968963623, + "learning_rate": 4.086410354015988e-06, + "loss": 1.4451, + "step": 788 + }, + { + "epoch": 0.810477657935285, + "grad_norm": 2.3837077617645264, + "learning_rate": 4.0845070422535216e-06, + "loss": 1.4641, + "step": 789 + }, + { + "epoch": 0.8115048793014895, + "grad_norm": 2.5199060440063477, + "learning_rate": 4.082603730491055e-06, + "loss": 1.5048, + "step": 790 + }, + { + "epoch": 0.8115048793014895, + "eval_loss": 1.8019492626190186, + "eval_runtime": 37.8805, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 790 + }, + { + "epoch": 0.8125321006676939, + "grad_norm": 2.4010138511657715, + "learning_rate": 4.080700418728588e-06, + "loss": 1.4881, + "step": 791 + }, + { + "epoch": 0.8135593220338984, + "grad_norm": 2.342094898223877, + "learning_rate": 4.078797106966122e-06, + "loss": 1.3792, + "step": 792 + }, + { + "epoch": 0.8145865434001027, + "grad_norm": 2.276772975921631, + "learning_rate": 4.076893795203655e-06, + "loss": 1.4539, + "step": 793 + }, + { + "epoch": 0.8156137647663071, + "grad_norm": 2.569275140762329, + "learning_rate": 4.074990483441188e-06, + "loss": 1.5059, + "step": 794 + }, + { + "epoch": 0.8166409861325116, + "grad_norm": 2.345761299133301, + "learning_rate": 4.073087171678721e-06, + "loss": 1.494, + "step": 795 + }, + { + "epoch": 0.8176682074987159, + "grad_norm": 2.4425504207611084, + "learning_rate": 4.071183859916254e-06, + "loss": 1.471, + "step": 796 + }, + { + "epoch": 0.8186954288649204, + "grad_norm": 2.5758635997772217, + "learning_rate": 4.069280548153788e-06, + "loss": 1.4378, + "step": 797 + }, + { + "epoch": 0.8197226502311248, + "grad_norm": 2.4320778846740723, + "learning_rate": 4.067377236391321e-06, + "loss": 1.474, + "step": 798 + }, + { + "epoch": 0.8207498715973293, + "grad_norm": 2.466978073120117, + "learning_rate": 4.065473924628854e-06, + "loss": 1.4997, + "step": 799 + }, + { + "epoch": 0.8217770929635336, + "grad_norm": 2.462489128112793, + "learning_rate": 4.063570612866388e-06, + "loss": 1.5, + "step": 800 + }, + { + "epoch": 0.8217770929635336, + "eval_loss": 1.783312201499939, + "eval_runtime": 37.8832, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 800 + }, + { + "epoch": 0.8228043143297381, + "grad_norm": 2.367455005645752, + "learning_rate": 4.061667301103921e-06, + "loss": 1.4658, + "step": 801 + }, + { + "epoch": 0.8238315356959425, + "grad_norm": 2.292250394821167, + "learning_rate": 4.0597639893414544e-06, + "loss": 1.476, + "step": 802 + }, + { + "epoch": 0.8248587570621468, + "grad_norm": 2.3460025787353516, + "learning_rate": 4.057860677578988e-06, + "loss": 1.4932, + "step": 803 + }, + { + "epoch": 0.8258859784283513, + "grad_norm": 2.3708906173706055, + "learning_rate": 4.055957365816521e-06, + "loss": 1.4396, + "step": 804 + }, + { + "epoch": 0.8269131997945557, + "grad_norm": 2.353513479232788, + "learning_rate": 4.0540540540540545e-06, + "loss": 1.4285, + "step": 805 + }, + { + "epoch": 0.8279404211607602, + "grad_norm": 2.319577932357788, + "learning_rate": 4.052150742291588e-06, + "loss": 1.4326, + "step": 806 + }, + { + "epoch": 0.8289676425269645, + "grad_norm": 2.450364112854004, + "learning_rate": 4.050247430529121e-06, + "loss": 1.5109, + "step": 807 + }, + { + "epoch": 0.829994863893169, + "grad_norm": 2.393907070159912, + "learning_rate": 4.048344118766655e-06, + "loss": 1.4948, + "step": 808 + }, + { + "epoch": 0.8310220852593734, + "grad_norm": 2.336002826690674, + "learning_rate": 4.046440807004188e-06, + "loss": 1.4851, + "step": 809 + }, + { + "epoch": 0.8320493066255779, + "grad_norm": 2.281139850616455, + "learning_rate": 4.044537495241721e-06, + "loss": 1.4445, + "step": 810 + }, + { + "epoch": 0.8320493066255779, + "eval_loss": 1.784231185913086, + "eval_runtime": 37.6614, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 810 + }, + { + "epoch": 0.8330765279917822, + "grad_norm": 2.4848175048828125, + "learning_rate": 4.042634183479255e-06, + "loss": 1.4982, + "step": 811 + }, + { + "epoch": 0.8341037493579867, + "grad_norm": 2.263347864151001, + "learning_rate": 4.040730871716787e-06, + "loss": 1.4795, + "step": 812 + }, + { + "epoch": 0.8351309707241911, + "grad_norm": 2.394005060195923, + "learning_rate": 4.038827559954321e-06, + "loss": 1.5231, + "step": 813 + }, + { + "epoch": 0.8361581920903954, + "grad_norm": 2.5051357746124268, + "learning_rate": 4.036924248191854e-06, + "loss": 1.4886, + "step": 814 + }, + { + "epoch": 0.8371854134565999, + "grad_norm": 2.3580374717712402, + "learning_rate": 4.035020936429387e-06, + "loss": 1.4713, + "step": 815 + }, + { + "epoch": 0.8382126348228043, + "grad_norm": 2.3179264068603516, + "learning_rate": 4.033117624666921e-06, + "loss": 1.4369, + "step": 816 + }, + { + "epoch": 0.8392398561890088, + "grad_norm": 2.269911289215088, + "learning_rate": 4.031214312904454e-06, + "loss": 1.4326, + "step": 817 + }, + { + "epoch": 0.8402670775552131, + "grad_norm": 2.3177807331085205, + "learning_rate": 4.029311001141987e-06, + "loss": 1.4096, + "step": 818 + }, + { + "epoch": 0.8412942989214176, + "grad_norm": 2.394324541091919, + "learning_rate": 4.027407689379521e-06, + "loss": 1.4647, + "step": 819 + }, + { + "epoch": 0.842321520287622, + "grad_norm": 2.288536548614502, + "learning_rate": 4.025504377617054e-06, + "loss": 1.4472, + "step": 820 + }, + { + "epoch": 0.842321520287622, + "eval_loss": 1.7811495065689087, + "eval_runtime": 37.8132, + "eval_samples_per_second": 26.446, + "eval_steps_per_second": 1.111, + "step": 820 + }, + { + "epoch": 0.8433487416538265, + "grad_norm": 2.376574754714966, + "learning_rate": 4.0236010658545875e-06, + "loss": 1.4918, + "step": 821 + }, + { + "epoch": 0.8443759630200308, + "grad_norm": 2.5850327014923096, + "learning_rate": 4.02169775409212e-06, + "loss": 1.5342, + "step": 822 + }, + { + "epoch": 0.8454031843862352, + "grad_norm": 2.247206211090088, + "learning_rate": 4.019794442329653e-06, + "loss": 1.4177, + "step": 823 + }, + { + "epoch": 0.8464304057524397, + "grad_norm": 2.4411795139312744, + "learning_rate": 4.0178911305671876e-06, + "loss": 1.4992, + "step": 824 + }, + { + "epoch": 0.847457627118644, + "grad_norm": 2.3377397060394287, + "learning_rate": 4.015987818804721e-06, + "loss": 1.4359, + "step": 825 + }, + { + "epoch": 0.8484848484848485, + "grad_norm": 2.329857349395752, + "learning_rate": 4.014084507042254e-06, + "loss": 1.4096, + "step": 826 + }, + { + "epoch": 0.8495120698510529, + "grad_norm": 2.563685178756714, + "learning_rate": 4.012181195279788e-06, + "loss": 1.5174, + "step": 827 + }, + { + "epoch": 0.8505392912172574, + "grad_norm": 2.4521079063415527, + "learning_rate": 4.010277883517321e-06, + "loss": 1.5028, + "step": 828 + }, + { + "epoch": 0.8515665125834617, + "grad_norm": 2.4091477394104004, + "learning_rate": 4.0083745717548535e-06, + "loss": 1.4615, + "step": 829 + }, + { + "epoch": 0.8525937339496662, + "grad_norm": 2.4010677337646484, + "learning_rate": 4.006471259992387e-06, + "loss": 1.4834, + "step": 830 + }, + { + "epoch": 0.8525937339496662, + "eval_loss": 1.7797678709030151, + "eval_runtime": 37.9378, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 830 + }, + { + "epoch": 0.8536209553158706, + "grad_norm": 2.3712635040283203, + "learning_rate": 4.00456794822992e-06, + "loss": 1.5196, + "step": 831 + }, + { + "epoch": 0.8546481766820749, + "grad_norm": 2.4096601009368896, + "learning_rate": 4.002664636467454e-06, + "loss": 1.4769, + "step": 832 + }, + { + "epoch": 0.8556753980482794, + "grad_norm": 2.408466100692749, + "learning_rate": 4.000761324704987e-06, + "loss": 1.4268, + "step": 833 + }, + { + "epoch": 0.8567026194144838, + "grad_norm": 2.6857047080993652, + "learning_rate": 3.99885801294252e-06, + "loss": 1.492, + "step": 834 + }, + { + "epoch": 0.8577298407806883, + "grad_norm": 2.297550916671753, + "learning_rate": 3.996954701180054e-06, + "loss": 1.4184, + "step": 835 + }, + { + "epoch": 0.8587570621468926, + "grad_norm": 2.423224925994873, + "learning_rate": 3.995051389417587e-06, + "loss": 1.4337, + "step": 836 + }, + { + "epoch": 0.8597842835130971, + "grad_norm": 2.260493040084839, + "learning_rate": 3.9931480776551204e-06, + "loss": 1.5267, + "step": 837 + }, + { + "epoch": 0.8608115048793015, + "grad_norm": 2.3803012371063232, + "learning_rate": 3.991244765892654e-06, + "loss": 1.4426, + "step": 838 + }, + { + "epoch": 0.861838726245506, + "grad_norm": 2.1425721645355225, + "learning_rate": 3.989341454130186e-06, + "loss": 1.4148, + "step": 839 + }, + { + "epoch": 0.8628659476117103, + "grad_norm": 2.316624641418457, + "learning_rate": 3.98743814236772e-06, + "loss": 1.4544, + "step": 840 + }, + { + "epoch": 0.8628659476117103, + "eval_loss": 1.7824597358703613, + "eval_runtime": 37.9676, + "eval_samples_per_second": 26.338, + "eval_steps_per_second": 1.106, + "step": 840 + }, + { + "epoch": 0.8638931689779148, + "grad_norm": 2.3723156452178955, + "learning_rate": 3.985534830605253e-06, + "loss": 1.473, + "step": 841 + }, + { + "epoch": 0.8649203903441192, + "grad_norm": 2.5167760848999023, + "learning_rate": 3.983631518842786e-06, + "loss": 1.4919, + "step": 842 + }, + { + "epoch": 0.8659476117103235, + "grad_norm": 2.3018670082092285, + "learning_rate": 3.98172820708032e-06, + "loss": 1.447, + "step": 843 + }, + { + "epoch": 0.866974833076528, + "grad_norm": 2.420483112335205, + "learning_rate": 3.979824895317854e-06, + "loss": 1.4159, + "step": 844 + }, + { + "epoch": 0.8680020544427324, + "grad_norm": 2.39443039894104, + "learning_rate": 3.977921583555387e-06, + "loss": 1.4984, + "step": 845 + }, + { + "epoch": 0.8690292758089369, + "grad_norm": 2.2953920364379883, + "learning_rate": 3.97601827179292e-06, + "loss": 1.4297, + "step": 846 + }, + { + "epoch": 0.8700564971751412, + "grad_norm": 2.2919528484344482, + "learning_rate": 3.974114960030453e-06, + "loss": 1.4728, + "step": 847 + }, + { + "epoch": 0.8710837185413457, + "grad_norm": 2.327914237976074, + "learning_rate": 3.9722116482679866e-06, + "loss": 1.4584, + "step": 848 + }, + { + "epoch": 0.8721109399075501, + "grad_norm": 2.305720567703247, + "learning_rate": 3.97030833650552e-06, + "loss": 1.4688, + "step": 849 + }, + { + "epoch": 0.8731381612737545, + "grad_norm": 2.2969236373901367, + "learning_rate": 3.968405024743053e-06, + "loss": 1.3972, + "step": 850 + }, + { + "epoch": 0.8731381612737545, + "eval_loss": 1.775634765625, + "eval_runtime": 37.9916, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 850 + }, + { + "epoch": 0.8741653826399589, + "grad_norm": 2.314387083053589, + "learning_rate": 3.966501712980587e-06, + "loss": 1.4476, + "step": 851 + }, + { + "epoch": 0.8751926040061633, + "grad_norm": 2.3271584510803223, + "learning_rate": 3.96459840121812e-06, + "loss": 1.4615, + "step": 852 + }, + { + "epoch": 0.8762198253723678, + "grad_norm": 2.494795322418213, + "learning_rate": 3.962695089455653e-06, + "loss": 1.5062, + "step": 853 + }, + { + "epoch": 0.8772470467385721, + "grad_norm": 2.4132444858551025, + "learning_rate": 3.960791777693187e-06, + "loss": 1.4861, + "step": 854 + }, + { + "epoch": 0.8782742681047766, + "grad_norm": 2.3869900703430176, + "learning_rate": 3.95888846593072e-06, + "loss": 1.4923, + "step": 855 + }, + { + "epoch": 0.879301489470981, + "grad_norm": 2.5167367458343506, + "learning_rate": 3.956985154168253e-06, + "loss": 1.5137, + "step": 856 + }, + { + "epoch": 0.8803287108371854, + "grad_norm": 2.510361671447754, + "learning_rate": 3.955081842405786e-06, + "loss": 1.5414, + "step": 857 + }, + { + "epoch": 0.8813559322033898, + "grad_norm": 2.554427146911621, + "learning_rate": 3.953178530643319e-06, + "loss": 1.5569, + "step": 858 + }, + { + "epoch": 0.8823831535695943, + "grad_norm": 2.306509256362915, + "learning_rate": 3.951275218880853e-06, + "loss": 1.3982, + "step": 859 + }, + { + "epoch": 0.8834103749357987, + "grad_norm": 2.3805363178253174, + "learning_rate": 3.949371907118386e-06, + "loss": 1.4977, + "step": 860 + }, + { + "epoch": 0.8834103749357987, + "eval_loss": 1.7882694005966187, + "eval_runtime": 38.0072, + "eval_samples_per_second": 26.311, + "eval_steps_per_second": 1.105, + "step": 860 + }, + { + "epoch": 0.884437596302003, + "grad_norm": 2.5413448810577393, + "learning_rate": 3.9474685953559194e-06, + "loss": 1.472, + "step": 861 + }, + { + "epoch": 0.8854648176682075, + "grad_norm": 2.484834909439087, + "learning_rate": 3.945565283593453e-06, + "loss": 1.4597, + "step": 862 + }, + { + "epoch": 0.8864920390344119, + "grad_norm": 2.6060585975646973, + "learning_rate": 3.943661971830986e-06, + "loss": 1.509, + "step": 863 + }, + { + "epoch": 0.8875192604006163, + "grad_norm": 2.6536927223205566, + "learning_rate": 3.9417586600685195e-06, + "loss": 1.4836, + "step": 864 + }, + { + "epoch": 0.8885464817668207, + "grad_norm": 2.5198960304260254, + "learning_rate": 3.939855348306053e-06, + "loss": 1.486, + "step": 865 + }, + { + "epoch": 0.8895737031330252, + "grad_norm": 2.335204839706421, + "learning_rate": 3.937952036543586e-06, + "loss": 1.4701, + "step": 866 + }, + { + "epoch": 0.8906009244992296, + "grad_norm": 2.7058098316192627, + "learning_rate": 3.93604872478112e-06, + "loss": 1.4798, + "step": 867 + }, + { + "epoch": 0.891628145865434, + "grad_norm": 2.4798922538757324, + "learning_rate": 3.934145413018653e-06, + "loss": 1.4364, + "step": 868 + }, + { + "epoch": 0.8926553672316384, + "grad_norm": 2.3090760707855225, + "learning_rate": 3.932242101256186e-06, + "loss": 1.4405, + "step": 869 + }, + { + "epoch": 0.8936825885978429, + "grad_norm": 2.4251503944396973, + "learning_rate": 3.93033878949372e-06, + "loss": 1.4113, + "step": 870 + }, + { + "epoch": 0.8936825885978429, + "eval_loss": 1.794334888458252, + "eval_runtime": 38.0197, + "eval_samples_per_second": 26.302, + "eval_steps_per_second": 1.105, + "step": 870 + }, + { + "epoch": 0.8947098099640473, + "grad_norm": 2.594945192337036, + "learning_rate": 3.928435477731253e-06, + "loss": 1.502, + "step": 871 + }, + { + "epoch": 0.8957370313302516, + "grad_norm": 2.4224843978881836, + "learning_rate": 3.9265321659687864e-06, + "loss": 1.5005, + "step": 872 + }, + { + "epoch": 0.8967642526964561, + "grad_norm": 2.3968186378479004, + "learning_rate": 3.924628854206319e-06, + "loss": 1.481, + "step": 873 + }, + { + "epoch": 0.8977914740626605, + "grad_norm": 2.4923524856567383, + "learning_rate": 3.922725542443852e-06, + "loss": 1.4259, + "step": 874 + }, + { + "epoch": 0.8988186954288649, + "grad_norm": 2.5386104583740234, + "learning_rate": 3.920822230681386e-06, + "loss": 1.4497, + "step": 875 + }, + { + "epoch": 0.8998459167950693, + "grad_norm": 2.4014909267425537, + "learning_rate": 3.918918918918919e-06, + "loss": 1.4816, + "step": 876 + }, + { + "epoch": 0.9008731381612738, + "grad_norm": 2.4019317626953125, + "learning_rate": 3.917015607156452e-06, + "loss": 1.4669, + "step": 877 + }, + { + "epoch": 0.9019003595274782, + "grad_norm": 2.5408380031585693, + "learning_rate": 3.915112295393986e-06, + "loss": 1.5343, + "step": 878 + }, + { + "epoch": 0.9029275808936826, + "grad_norm": 2.364053964614868, + "learning_rate": 3.913208983631519e-06, + "loss": 1.5229, + "step": 879 + }, + { + "epoch": 0.903954802259887, + "grad_norm": 2.307412624359131, + "learning_rate": 3.9113056718690525e-06, + "loss": 1.4444, + "step": 880 + }, + { + "epoch": 0.903954802259887, + "eval_loss": 1.7910975217819214, + "eval_runtime": 38.0659, + "eval_samples_per_second": 26.27, + "eval_steps_per_second": 1.103, + "step": 880 + }, + { + "epoch": 0.9049820236260914, + "grad_norm": 2.496135950088501, + "learning_rate": 3.909402360106586e-06, + "loss": 1.4853, + "step": 881 + }, + { + "epoch": 0.9060092449922958, + "grad_norm": 2.4825589656829834, + "learning_rate": 3.907499048344119e-06, + "loss": 1.4886, + "step": 882 + }, + { + "epoch": 0.9070364663585002, + "grad_norm": 2.3458824157714844, + "learning_rate": 3.905595736581652e-06, + "loss": 1.4882, + "step": 883 + }, + { + "epoch": 0.9080636877247047, + "grad_norm": 2.2699403762817383, + "learning_rate": 3.903692424819186e-06, + "loss": 1.4417, + "step": 884 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 2.348593235015869, + "learning_rate": 3.901789113056719e-06, + "loss": 1.4991, + "step": 885 + }, + { + "epoch": 0.9101181304571135, + "grad_norm": 2.3292129039764404, + "learning_rate": 3.899885801294253e-06, + "loss": 1.4752, + "step": 886 + }, + { + "epoch": 0.9111453518233179, + "grad_norm": 2.487183094024658, + "learning_rate": 3.897982489531786e-06, + "loss": 1.5317, + "step": 887 + }, + { + "epoch": 0.9121725731895224, + "grad_norm": 2.3300845623016357, + "learning_rate": 3.896079177769319e-06, + "loss": 1.4964, + "step": 888 + }, + { + "epoch": 0.9131997945557268, + "grad_norm": 2.3292312622070312, + "learning_rate": 3.894175866006853e-06, + "loss": 1.5221, + "step": 889 + }, + { + "epoch": 0.9142270159219312, + "grad_norm": 2.3556439876556396, + "learning_rate": 3.892272554244385e-06, + "loss": 1.4471, + "step": 890 + }, + { + "epoch": 0.9142270159219312, + "eval_loss": 1.7807039022445679, + "eval_runtime": 38.0614, + "eval_samples_per_second": 26.273, + "eval_steps_per_second": 1.103, + "step": 890 + }, + { + "epoch": 0.9152542372881356, + "grad_norm": 2.3302760124206543, + "learning_rate": 3.890369242481919e-06, + "loss": 1.5179, + "step": 891 + }, + { + "epoch": 0.91628145865434, + "grad_norm": 2.327072858810425, + "learning_rate": 3.888465930719452e-06, + "loss": 1.4474, + "step": 892 + }, + { + "epoch": 0.9173086800205444, + "grad_norm": 2.2773101329803467, + "learning_rate": 3.886562618956985e-06, + "loss": 1.4783, + "step": 893 + }, + { + "epoch": 0.9183359013867488, + "grad_norm": 2.2717974185943604, + "learning_rate": 3.884659307194519e-06, + "loss": 1.4062, + "step": 894 + }, + { + "epoch": 0.9193631227529533, + "grad_norm": 2.3350367546081543, + "learning_rate": 3.882755995432052e-06, + "loss": 1.483, + "step": 895 + }, + { + "epoch": 0.9203903441191577, + "grad_norm": 2.302811622619629, + "learning_rate": 3.8808526836695854e-06, + "loss": 1.4407, + "step": 896 + }, + { + "epoch": 0.9214175654853621, + "grad_norm": 2.4031271934509277, + "learning_rate": 3.878949371907119e-06, + "loss": 1.4331, + "step": 897 + }, + { + "epoch": 0.9224447868515665, + "grad_norm": 2.4806716442108154, + "learning_rate": 3.877046060144652e-06, + "loss": 1.4341, + "step": 898 + }, + { + "epoch": 0.923472008217771, + "grad_norm": 2.4243357181549072, + "learning_rate": 3.8751427483821855e-06, + "loss": 1.4701, + "step": 899 + }, + { + "epoch": 0.9244992295839753, + "grad_norm": 2.4953269958496094, + "learning_rate": 3.873239436619718e-06, + "loss": 1.4287, + "step": 900 + }, + { + "epoch": 0.9244992295839753, + "eval_loss": 1.7872728109359741, + "eval_runtime": 38.0906, + "eval_samples_per_second": 26.253, + "eval_steps_per_second": 1.103, + "step": 900 + }, + { + "epoch": 0.9255264509501797, + "grad_norm": 2.307490587234497, + "learning_rate": 3.871336124857251e-06, + "loss": 1.4693, + "step": 901 + }, + { + "epoch": 0.9265536723163842, + "grad_norm": 2.4745755195617676, + "learning_rate": 3.869432813094785e-06, + "loss": 1.4779, + "step": 902 + }, + { + "epoch": 0.9275808936825886, + "grad_norm": 2.367715358734131, + "learning_rate": 3.867529501332318e-06, + "loss": 1.4771, + "step": 903 + }, + { + "epoch": 0.928608115048793, + "grad_norm": 2.388549566268921, + "learning_rate": 3.8656261895698515e-06, + "loss": 1.4648, + "step": 904 + }, + { + "epoch": 0.9296353364149974, + "grad_norm": 2.39432954788208, + "learning_rate": 3.863722877807386e-06, + "loss": 1.4932, + "step": 905 + }, + { + "epoch": 0.9306625577812019, + "grad_norm": 2.3921315670013428, + "learning_rate": 3.861819566044919e-06, + "loss": 1.5312, + "step": 906 + }, + { + "epoch": 0.9316897791474062, + "grad_norm": 2.400240659713745, + "learning_rate": 3.859916254282452e-06, + "loss": 1.4651, + "step": 907 + }, + { + "epoch": 0.9327170005136107, + "grad_norm": 2.394559860229492, + "learning_rate": 3.858012942519985e-06, + "loss": 1.4759, + "step": 908 + }, + { + "epoch": 0.9337442218798151, + "grad_norm": 2.3365273475646973, + "learning_rate": 3.856109630757518e-06, + "loss": 1.4502, + "step": 909 + }, + { + "epoch": 0.9347714432460195, + "grad_norm": 2.465691328048706, + "learning_rate": 3.854206318995052e-06, + "loss": 1.5036, + "step": 910 + }, + { + "epoch": 0.9347714432460195, + "eval_loss": 1.7682362794876099, + "eval_runtime": 37.6624, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 910 + }, + { + "epoch": 0.9357986646122239, + "grad_norm": 2.3376758098602295, + "learning_rate": 3.852303007232585e-06, + "loss": 1.4342, + "step": 911 + }, + { + "epoch": 0.9368258859784283, + "grad_norm": 2.345855236053467, + "learning_rate": 3.850399695470118e-06, + "loss": 1.4444, + "step": 912 + }, + { + "epoch": 0.9378531073446328, + "grad_norm": 2.4444501399993896, + "learning_rate": 3.848496383707652e-06, + "loss": 1.4315, + "step": 913 + }, + { + "epoch": 0.9388803287108372, + "grad_norm": 2.4896373748779297, + "learning_rate": 3.846593071945185e-06, + "loss": 1.4228, + "step": 914 + }, + { + "epoch": 0.9399075500770416, + "grad_norm": 2.3970463275909424, + "learning_rate": 3.8446897601827185e-06, + "loss": 1.5283, + "step": 915 + }, + { + "epoch": 0.940934771443246, + "grad_norm": 2.5194883346557617, + "learning_rate": 3.842786448420252e-06, + "loss": 1.5445, + "step": 916 + }, + { + "epoch": 0.9419619928094505, + "grad_norm": 2.446465253829956, + "learning_rate": 3.840883136657785e-06, + "loss": 1.5248, + "step": 917 + }, + { + "epoch": 0.9429892141756548, + "grad_norm": 2.348477840423584, + "learning_rate": 3.838979824895318e-06, + "loss": 1.402, + "step": 918 + }, + { + "epoch": 0.9440164355418593, + "grad_norm": 2.4511539936065674, + "learning_rate": 3.837076513132851e-06, + "loss": 1.4893, + "step": 919 + }, + { + "epoch": 0.9450436569080637, + "grad_norm": 2.3801777362823486, + "learning_rate": 3.8351732013703844e-06, + "loss": 1.4457, + "step": 920 + }, + { + "epoch": 0.9450436569080637, + "eval_loss": 1.7810813188552856, + "eval_runtime": 37.7647, + "eval_samples_per_second": 26.48, + "eval_steps_per_second": 1.112, + "step": 920 + }, + { + "epoch": 0.9460708782742681, + "grad_norm": 2.5260043144226074, + "learning_rate": 3.833269889607918e-06, + "loss": 1.4267, + "step": 921 + }, + { + "epoch": 0.9470980996404725, + "grad_norm": 2.270653009414673, + "learning_rate": 3.831366577845451e-06, + "loss": 1.4313, + "step": 922 + }, + { + "epoch": 0.9481253210066769, + "grad_norm": 2.5123250484466553, + "learning_rate": 3.8294632660829845e-06, + "loss": 1.4588, + "step": 923 + }, + { + "epoch": 0.9491525423728814, + "grad_norm": 2.4242913722991943, + "learning_rate": 3.827559954320518e-06, + "loss": 1.4659, + "step": 924 + }, + { + "epoch": 0.9501797637390857, + "grad_norm": 2.4337117671966553, + "learning_rate": 3.825656642558051e-06, + "loss": 1.4838, + "step": 925 + }, + { + "epoch": 0.9512069851052902, + "grad_norm": 2.3577871322631836, + "learning_rate": 3.823753330795585e-06, + "loss": 1.4674, + "step": 926 + }, + { + "epoch": 0.9522342064714946, + "grad_norm": 2.2574679851531982, + "learning_rate": 3.821850019033118e-06, + "loss": 1.4011, + "step": 927 + }, + { + "epoch": 0.9532614278376991, + "grad_norm": 2.38043475151062, + "learning_rate": 3.819946707270651e-06, + "loss": 1.459, + "step": 928 + }, + { + "epoch": 0.9542886492039034, + "grad_norm": 2.4249954223632812, + "learning_rate": 3.818043395508185e-06, + "loss": 1.4649, + "step": 929 + }, + { + "epoch": 0.9553158705701078, + "grad_norm": 2.4420788288116455, + "learning_rate": 3.816140083745718e-06, + "loss": 1.532, + "step": 930 + }, + { + "epoch": 0.9553158705701078, + "eval_loss": 1.7805145978927612, + "eval_runtime": 37.8606, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 930 + }, + { + "epoch": 0.9563430919363123, + "grad_norm": 2.391042470932007, + "learning_rate": 3.8142367719832514e-06, + "loss": 1.4365, + "step": 931 + }, + { + "epoch": 0.9573703133025167, + "grad_norm": 2.462303876876831, + "learning_rate": 3.8123334602207844e-06, + "loss": 1.4367, + "step": 932 + }, + { + "epoch": 0.9583975346687211, + "grad_norm": 2.330714702606201, + "learning_rate": 3.8104301484583177e-06, + "loss": 1.427, + "step": 933 + }, + { + "epoch": 0.9594247560349255, + "grad_norm": 2.3475441932678223, + "learning_rate": 3.808526836695851e-06, + "loss": 1.4836, + "step": 934 + }, + { + "epoch": 0.96045197740113, + "grad_norm": 2.4546470642089844, + "learning_rate": 3.8066235249333845e-06, + "loss": 1.4601, + "step": 935 + }, + { + "epoch": 0.9614791987673343, + "grad_norm": 2.4604640007019043, + "learning_rate": 3.804720213170918e-06, + "loss": 1.4616, + "step": 936 + }, + { + "epoch": 0.9625064201335388, + "grad_norm": 2.3087422847747803, + "learning_rate": 3.8028169014084508e-06, + "loss": 1.4816, + "step": 937 + }, + { + "epoch": 0.9635336414997432, + "grad_norm": 2.375377893447876, + "learning_rate": 3.800913589645984e-06, + "loss": 1.471, + "step": 938 + }, + { + "epoch": 0.9645608628659477, + "grad_norm": 2.358741283416748, + "learning_rate": 3.7990102778835175e-06, + "loss": 1.4754, + "step": 939 + }, + { + "epoch": 0.965588084232152, + "grad_norm": 2.3990139961242676, + "learning_rate": 3.797106966121051e-06, + "loss": 1.534, + "step": 940 + }, + { + "epoch": 0.965588084232152, + "eval_loss": 1.7745739221572876, + "eval_runtime": 37.8883, + "eval_samples_per_second": 26.393, + "eval_steps_per_second": 1.109, + "step": 940 + }, + { + "epoch": 0.9666153055983564, + "grad_norm": 2.417396068572998, + "learning_rate": 3.7952036543585842e-06, + "loss": 1.4994, + "step": 941 + }, + { + "epoch": 0.9676425269645609, + "grad_norm": 2.323125123977661, + "learning_rate": 3.793300342596117e-06, + "loss": 1.5045, + "step": 942 + }, + { + "epoch": 0.9686697483307652, + "grad_norm": 2.4424829483032227, + "learning_rate": 3.7913970308336505e-06, + "loss": 1.4936, + "step": 943 + }, + { + "epoch": 0.9696969696969697, + "grad_norm": 2.2769429683685303, + "learning_rate": 3.789493719071184e-06, + "loss": 1.4453, + "step": 944 + }, + { + "epoch": 0.9707241910631741, + "grad_norm": 2.183701276779175, + "learning_rate": 3.7875904073087177e-06, + "loss": 1.4005, + "step": 945 + }, + { + "epoch": 0.9717514124293786, + "grad_norm": 3.4538655281066895, + "learning_rate": 3.785687095546251e-06, + "loss": 1.4521, + "step": 946 + }, + { + "epoch": 0.9727786337955829, + "grad_norm": 2.3735523223876953, + "learning_rate": 3.7837837837837844e-06, + "loss": 1.4328, + "step": 947 + }, + { + "epoch": 0.9738058551617874, + "grad_norm": 2.6214165687561035, + "learning_rate": 3.7818804720213178e-06, + "loss": 1.4004, + "step": 948 + }, + { + "epoch": 0.9748330765279918, + "grad_norm": 2.365091562271118, + "learning_rate": 3.7799771602588507e-06, + "loss": 1.4006, + "step": 949 + }, + { + "epoch": 0.9758602978941961, + "grad_norm": 2.1926634311676025, + "learning_rate": 3.778073848496384e-06, + "loss": 1.4587, + "step": 950 + }, + { + "epoch": 0.9758602978941961, + "eval_loss": 1.767502784729004, + "eval_runtime": 37.9287, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 950 + }, + { + "epoch": 0.9768875192604006, + "grad_norm": 2.3047192096710205, + "learning_rate": 3.7761705367339174e-06, + "loss": 1.4826, + "step": 951 + }, + { + "epoch": 0.977914740626605, + "grad_norm": 2.3306028842926025, + "learning_rate": 3.7742672249714508e-06, + "loss": 1.4914, + "step": 952 + }, + { + "epoch": 0.9789419619928095, + "grad_norm": 2.3409693241119385, + "learning_rate": 3.772363913208984e-06, + "loss": 1.4556, + "step": 953 + }, + { + "epoch": 0.9799691833590138, + "grad_norm": 2.3337934017181396, + "learning_rate": 3.770460601446517e-06, + "loss": 1.4751, + "step": 954 + }, + { + "epoch": 0.9809964047252183, + "grad_norm": 2.3354830741882324, + "learning_rate": 3.7685572896840504e-06, + "loss": 1.4176, + "step": 955 + }, + { + "epoch": 0.9820236260914227, + "grad_norm": 2.4049782752990723, + "learning_rate": 3.766653977921584e-06, + "loss": 1.3715, + "step": 956 + }, + { + "epoch": 0.9830508474576272, + "grad_norm": 2.5127065181732178, + "learning_rate": 3.764750666159117e-06, + "loss": 1.468, + "step": 957 + }, + { + "epoch": 0.9840780688238315, + "grad_norm": 2.5576820373535156, + "learning_rate": 3.7628473543966505e-06, + "loss": 1.4237, + "step": 958 + }, + { + "epoch": 0.9851052901900359, + "grad_norm": 2.394098997116089, + "learning_rate": 3.7609440426341835e-06, + "loss": 1.5126, + "step": 959 + }, + { + "epoch": 0.9861325115562404, + "grad_norm": 2.3358702659606934, + "learning_rate": 3.759040730871717e-06, + "loss": 1.4109, + "step": 960 + }, + { + "epoch": 0.9861325115562404, + "eval_loss": 1.7665778398513794, + "eval_runtime": 37.9028, + "eval_samples_per_second": 26.383, + "eval_steps_per_second": 1.108, + "step": 960 + }, + { + "epoch": 0.9871597329224447, + "grad_norm": 2.4213271141052246, + "learning_rate": 3.75713741910925e-06, + "loss": 1.3995, + "step": 961 + }, + { + "epoch": 0.9881869542886492, + "grad_norm": 2.367835760116577, + "learning_rate": 3.7552341073467836e-06, + "loss": 1.4491, + "step": 962 + }, + { + "epoch": 0.9892141756548536, + "grad_norm": 2.433840751647949, + "learning_rate": 3.753330795584317e-06, + "loss": 1.5078, + "step": 963 + }, + { + "epoch": 0.9902413970210581, + "grad_norm": 2.3976447582244873, + "learning_rate": 3.75142748382185e-06, + "loss": 1.4107, + "step": 964 + }, + { + "epoch": 0.9912686183872624, + "grad_norm": 2.4369053840637207, + "learning_rate": 3.749524172059384e-06, + "loss": 1.4569, + "step": 965 + }, + { + "epoch": 0.9922958397534669, + "grad_norm": 2.3203182220458984, + "learning_rate": 3.747620860296917e-06, + "loss": 1.4976, + "step": 966 + }, + { + "epoch": 0.9933230611196713, + "grad_norm": 2.431885242462158, + "learning_rate": 3.7457175485344504e-06, + "loss": 1.4945, + "step": 967 + }, + { + "epoch": 0.9943502824858758, + "grad_norm": 2.2997021675109863, + "learning_rate": 3.7438142367719837e-06, + "loss": 1.48, + "step": 968 + }, + { + "epoch": 0.9953775038520801, + "grad_norm": 2.2502849102020264, + "learning_rate": 3.741910925009517e-06, + "loss": 1.4322, + "step": 969 + }, + { + "epoch": 0.9964047252182845, + "grad_norm": 2.311049222946167, + "learning_rate": 3.7400076132470505e-06, + "loss": 1.4498, + "step": 970 + }, + { + "epoch": 0.9964047252182845, + "eval_loss": 1.771975040435791, + "eval_runtime": 37.852, + "eval_samples_per_second": 26.419, + "eval_steps_per_second": 1.11, + "step": 970 + }, + { + "epoch": 0.997431946584489, + "grad_norm": 2.431821584701538, + "learning_rate": 3.7381043014845834e-06, + "loss": 1.4796, + "step": 971 + }, + { + "epoch": 0.9984591679506933, + "grad_norm": 2.319145441055298, + "learning_rate": 3.7362009897221168e-06, + "loss": 1.459, + "step": 972 + }, + { + "epoch": 0.9994863893168978, + "grad_norm": 2.4445269107818604, + "learning_rate": 3.73429767795965e-06, + "loss": 1.4764, + "step": 973 + }, + { + "epoch": 1.0005136106831023, + "grad_norm": 2.272233009338379, + "learning_rate": 3.7323943661971835e-06, + "loss": 1.3799, + "step": 974 + }, + { + "epoch": 1.0015408320493067, + "grad_norm": 2.5421202182769775, + "learning_rate": 3.730491054434717e-06, + "loss": 1.3257, + "step": 975 + }, + { + "epoch": 1.002568053415511, + "grad_norm": 2.3988611698150635, + "learning_rate": 3.72858774267225e-06, + "loss": 1.4244, + "step": 976 + }, + { + "epoch": 1.0035952747817154, + "grad_norm": 2.5080084800720215, + "learning_rate": 3.726684430909783e-06, + "loss": 1.4461, + "step": 977 + }, + { + "epoch": 1.00462249614792, + "grad_norm": 2.411081314086914, + "learning_rate": 3.7247811191473165e-06, + "loss": 1.45, + "step": 978 + }, + { + "epoch": 1.0056497175141244, + "grad_norm": 2.729546070098877, + "learning_rate": 3.72287780738485e-06, + "loss": 1.4296, + "step": 979 + }, + { + "epoch": 1.0066769388803287, + "grad_norm": 2.5932579040527344, + "learning_rate": 3.7209744956223832e-06, + "loss": 1.4312, + "step": 980 + }, + { + "epoch": 1.0066769388803287, + "eval_loss": 1.770633339881897, + "eval_runtime": 37.8977, + "eval_samples_per_second": 26.387, + "eval_steps_per_second": 1.108, + "step": 980 + }, + { + "epoch": 1.007704160246533, + "grad_norm": 2.4512336254119873, + "learning_rate": 3.719071183859916e-06, + "loss": 1.3637, + "step": 981 + }, + { + "epoch": 1.0087313816127375, + "grad_norm": 2.5416886806488037, + "learning_rate": 3.7171678720974495e-06, + "loss": 1.4591, + "step": 982 + }, + { + "epoch": 1.009758602978942, + "grad_norm": 2.464679002761841, + "learning_rate": 3.715264560334983e-06, + "loss": 1.4306, + "step": 983 + }, + { + "epoch": 1.0107858243451464, + "grad_norm": 2.3745839595794678, + "learning_rate": 3.7133612485725163e-06, + "loss": 1.4426, + "step": 984 + }, + { + "epoch": 1.0118130457113508, + "grad_norm": 2.3606340885162354, + "learning_rate": 3.71145793681005e-06, + "loss": 1.3618, + "step": 985 + }, + { + "epoch": 1.0128402670775551, + "grad_norm": 2.5836591720581055, + "learning_rate": 3.7095546250475834e-06, + "loss": 1.4776, + "step": 986 + }, + { + "epoch": 1.0138674884437597, + "grad_norm": 2.4108684062957764, + "learning_rate": 3.7076513132851168e-06, + "loss": 1.4047, + "step": 987 + }, + { + "epoch": 1.014894709809964, + "grad_norm": 2.4316983222961426, + "learning_rate": 3.7057480015226497e-06, + "loss": 1.3772, + "step": 988 + }, + { + "epoch": 1.0159219311761685, + "grad_norm": 2.5510315895080566, + "learning_rate": 3.703844689760183e-06, + "loss": 1.4536, + "step": 989 + }, + { + "epoch": 1.0169491525423728, + "grad_norm": 2.4947116374969482, + "learning_rate": 3.7019413779977164e-06, + "loss": 1.434, + "step": 990 + }, + { + "epoch": 1.0169491525423728, + "eval_loss": 1.7928223609924316, + "eval_runtime": 37.8838, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 990 + }, + { + "epoch": 1.0179763739085772, + "grad_norm": 2.5938122272491455, + "learning_rate": 3.70003806623525e-06, + "loss": 1.4083, + "step": 991 + }, + { + "epoch": 1.0190035952747818, + "grad_norm": 2.472865581512451, + "learning_rate": 3.698134754472783e-06, + "loss": 1.4696, + "step": 992 + }, + { + "epoch": 1.0200308166409862, + "grad_norm": 2.374344825744629, + "learning_rate": 3.696231442710316e-06, + "loss": 1.403, + "step": 993 + }, + { + "epoch": 1.0210580380071905, + "grad_norm": 2.6928529739379883, + "learning_rate": 3.6943281309478495e-06, + "loss": 1.4216, + "step": 994 + }, + { + "epoch": 1.022085259373395, + "grad_norm": 2.640890598297119, + "learning_rate": 3.692424819185383e-06, + "loss": 1.4362, + "step": 995 + }, + { + "epoch": 1.0231124807395995, + "grad_norm": 2.45357084274292, + "learning_rate": 3.690521507422916e-06, + "loss": 1.382, + "step": 996 + }, + { + "epoch": 1.0241397021058039, + "grad_norm": 2.712517023086548, + "learning_rate": 3.6886181956604496e-06, + "loss": 1.395, + "step": 997 + }, + { + "epoch": 1.0251669234720082, + "grad_norm": 2.7127771377563477, + "learning_rate": 3.686714883897983e-06, + "loss": 1.4589, + "step": 998 + }, + { + "epoch": 1.0261941448382126, + "grad_norm": 2.5379409790039062, + "learning_rate": 3.684811572135516e-06, + "loss": 1.4551, + "step": 999 + }, + { + "epoch": 1.027221366204417, + "grad_norm": 2.5015134811401367, + "learning_rate": 3.6829082603730492e-06, + "loss": 1.4112, + "step": 1000 + }, + { + "epoch": 1.027221366204417, + "eval_loss": 1.7806837558746338, + "eval_runtime": 37.9091, + "eval_samples_per_second": 26.379, + "eval_steps_per_second": 1.108, + "step": 1000 + }, + { + "epoch": 1.0282485875706215, + "grad_norm": 2.481306314468384, + "learning_rate": 3.6810049486105826e-06, + "loss": 1.4469, + "step": 1001 + }, + { + "epoch": 1.029275808936826, + "grad_norm": 2.481520891189575, + "learning_rate": 3.679101636848116e-06, + "loss": 1.4284, + "step": 1002 + }, + { + "epoch": 1.0303030303030303, + "grad_norm": 2.4291019439697266, + "learning_rate": 3.6771983250856493e-06, + "loss": 1.4067, + "step": 1003 + }, + { + "epoch": 1.0313302516692346, + "grad_norm": 2.5361173152923584, + "learning_rate": 3.6752950133231822e-06, + "loss": 1.4347, + "step": 1004 + }, + { + "epoch": 1.0323574730354392, + "grad_norm": 2.5588085651397705, + "learning_rate": 3.6733917015607156e-06, + "loss": 1.4455, + "step": 1005 + }, + { + "epoch": 1.0333846944016436, + "grad_norm": 2.555091619491577, + "learning_rate": 3.6714883897982494e-06, + "loss": 1.399, + "step": 1006 + }, + { + "epoch": 1.034411915767848, + "grad_norm": 2.5489652156829834, + "learning_rate": 3.6695850780357828e-06, + "loss": 1.3894, + "step": 1007 + }, + { + "epoch": 1.0354391371340523, + "grad_norm": 2.5258426666259766, + "learning_rate": 3.667681766273316e-06, + "loss": 1.3962, + "step": 1008 + }, + { + "epoch": 1.0364663585002567, + "grad_norm": 2.455916166305542, + "learning_rate": 3.6657784545108495e-06, + "loss": 1.4773, + "step": 1009 + }, + { + "epoch": 1.0374935798664613, + "grad_norm": 2.5466208457946777, + "learning_rate": 3.6638751427483824e-06, + "loss": 1.4337, + "step": 1010 + }, + { + "epoch": 1.0374935798664613, + "eval_loss": 1.7840261459350586, + "eval_runtime": 37.6685, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 1010 + }, + { + "epoch": 1.0385208012326657, + "grad_norm": 2.5446081161499023, + "learning_rate": 3.6619718309859158e-06, + "loss": 1.4777, + "step": 1011 + }, + { + "epoch": 1.03954802259887, + "grad_norm": 2.509493350982666, + "learning_rate": 3.660068519223449e-06, + "loss": 1.4295, + "step": 1012 + }, + { + "epoch": 1.0405752439650744, + "grad_norm": 2.417632818222046, + "learning_rate": 3.6581652074609825e-06, + "loss": 1.4044, + "step": 1013 + }, + { + "epoch": 1.041602465331279, + "grad_norm": 2.457902193069458, + "learning_rate": 3.656261895698516e-06, + "loss": 1.3948, + "step": 1014 + }, + { + "epoch": 1.0426296866974833, + "grad_norm": 2.4869253635406494, + "learning_rate": 3.6543585839360492e-06, + "loss": 1.397, + "step": 1015 + }, + { + "epoch": 1.0436569080636877, + "grad_norm": 2.495419979095459, + "learning_rate": 3.652455272173582e-06, + "loss": 1.4037, + "step": 1016 + }, + { + "epoch": 1.044684129429892, + "grad_norm": 2.5231642723083496, + "learning_rate": 3.6505519604111155e-06, + "loss": 1.4156, + "step": 1017 + }, + { + "epoch": 1.0457113507960965, + "grad_norm": 2.8302087783813477, + "learning_rate": 3.648648648648649e-06, + "loss": 1.4529, + "step": 1018 + }, + { + "epoch": 1.046738572162301, + "grad_norm": 2.467430353164673, + "learning_rate": 3.6467453368861823e-06, + "loss": 1.4078, + "step": 1019 + }, + { + "epoch": 1.0477657935285054, + "grad_norm": 2.469128131866455, + "learning_rate": 3.6448420251237156e-06, + "loss": 1.4344, + "step": 1020 + }, + { + "epoch": 1.0477657935285054, + "eval_loss": 1.7800015211105347, + "eval_runtime": 37.854, + "eval_samples_per_second": 26.417, + "eval_steps_per_second": 1.11, + "step": 1020 + }, + { + "epoch": 1.0487930148947098, + "grad_norm": 2.59389328956604, + "learning_rate": 3.6429387133612486e-06, + "loss": 1.451, + "step": 1021 + }, + { + "epoch": 1.0498202362609141, + "grad_norm": 2.5882468223571777, + "learning_rate": 3.641035401598782e-06, + "loss": 1.4026, + "step": 1022 + }, + { + "epoch": 1.0508474576271187, + "grad_norm": 2.601926565170288, + "learning_rate": 3.6391320898363153e-06, + "loss": 1.4382, + "step": 1023 + }, + { + "epoch": 1.051874678993323, + "grad_norm": 2.5336716175079346, + "learning_rate": 3.6372287780738487e-06, + "loss": 1.4184, + "step": 1024 + }, + { + "epoch": 1.0529019003595275, + "grad_norm": 2.503507614135742, + "learning_rate": 3.635325466311382e-06, + "loss": 1.4426, + "step": 1025 + }, + { + "epoch": 1.0539291217257318, + "grad_norm": 2.483013153076172, + "learning_rate": 3.633422154548916e-06, + "loss": 1.4563, + "step": 1026 + }, + { + "epoch": 1.0549563430919364, + "grad_norm": 2.4803354740142822, + "learning_rate": 3.6315188427864487e-06, + "loss": 1.4284, + "step": 1027 + }, + { + "epoch": 1.0559835644581408, + "grad_norm": 2.3735713958740234, + "learning_rate": 3.629615531023982e-06, + "loss": 1.3772, + "step": 1028 + }, + { + "epoch": 1.0570107858243452, + "grad_norm": 2.471226215362549, + "learning_rate": 3.6277122192615155e-06, + "loss": 1.4399, + "step": 1029 + }, + { + "epoch": 1.0580380071905495, + "grad_norm": 2.3827438354492188, + "learning_rate": 3.625808907499049e-06, + "loss": 1.3951, + "step": 1030 + }, + { + "epoch": 1.0580380071905495, + "eval_loss": 1.7787058353424072, + "eval_runtime": 37.9484, + "eval_samples_per_second": 26.352, + "eval_steps_per_second": 1.107, + "step": 1030 + }, + { + "epoch": 1.059065228556754, + "grad_norm": 2.403416633605957, + "learning_rate": 3.623905595736582e-06, + "loss": 1.3946, + "step": 1031 + }, + { + "epoch": 1.0600924499229585, + "grad_norm": 2.455740213394165, + "learning_rate": 3.6220022839741156e-06, + "loss": 1.3859, + "step": 1032 + }, + { + "epoch": 1.0611196712891628, + "grad_norm": 2.6349902153015137, + "learning_rate": 3.6200989722116485e-06, + "loss": 1.4092, + "step": 1033 + }, + { + "epoch": 1.0621468926553672, + "grad_norm": 2.6292974948883057, + "learning_rate": 3.618195660449182e-06, + "loss": 1.419, + "step": 1034 + }, + { + "epoch": 1.0631741140215716, + "grad_norm": 2.4338958263397217, + "learning_rate": 3.6162923486867152e-06, + "loss": 1.3546, + "step": 1035 + }, + { + "epoch": 1.064201335387776, + "grad_norm": 2.376420259475708, + "learning_rate": 3.6143890369242486e-06, + "loss": 1.3899, + "step": 1036 + }, + { + "epoch": 1.0652285567539805, + "grad_norm": 2.428738832473755, + "learning_rate": 3.612485725161782e-06, + "loss": 1.3955, + "step": 1037 + }, + { + "epoch": 1.066255778120185, + "grad_norm": 2.4316651821136475, + "learning_rate": 3.610582413399315e-06, + "loss": 1.3875, + "step": 1038 + }, + { + "epoch": 1.0672829994863893, + "grad_norm": 2.3969266414642334, + "learning_rate": 3.6086791016368482e-06, + "loss": 1.404, + "step": 1039 + }, + { + "epoch": 1.0683102208525936, + "grad_norm": 2.662283420562744, + "learning_rate": 3.6067757898743816e-06, + "loss": 1.4639, + "step": 1040 + }, + { + "epoch": 1.0683102208525936, + "eval_loss": 1.7603929042816162, + "eval_runtime": 38.0098, + "eval_samples_per_second": 26.309, + "eval_steps_per_second": 1.105, + "step": 1040 + }, + { + "epoch": 1.0693374422187982, + "grad_norm": 2.601452112197876, + "learning_rate": 3.604872478111915e-06, + "loss": 1.47, + "step": 1041 + }, + { + "epoch": 1.0703646635850026, + "grad_norm": 2.4221863746643066, + "learning_rate": 3.6029691663494483e-06, + "loss": 1.4162, + "step": 1042 + }, + { + "epoch": 1.071391884951207, + "grad_norm": 2.4996650218963623, + "learning_rate": 3.6010658545869813e-06, + "loss": 1.443, + "step": 1043 + }, + { + "epoch": 1.0724191063174113, + "grad_norm": 2.547239303588867, + "learning_rate": 3.5991625428245146e-06, + "loss": 1.4464, + "step": 1044 + }, + { + "epoch": 1.073446327683616, + "grad_norm": 2.481187343597412, + "learning_rate": 3.597259231062048e-06, + "loss": 1.3749, + "step": 1045 + }, + { + "epoch": 1.0744735490498203, + "grad_norm": 2.5504934787750244, + "learning_rate": 3.5953559192995818e-06, + "loss": 1.4231, + "step": 1046 + }, + { + "epoch": 1.0755007704160247, + "grad_norm": 2.465390205383301, + "learning_rate": 3.593452607537115e-06, + "loss": 1.4805, + "step": 1047 + }, + { + "epoch": 1.076527991782229, + "grad_norm": 2.4399969577789307, + "learning_rate": 3.5915492957746485e-06, + "loss": 1.4091, + "step": 1048 + }, + { + "epoch": 1.0775552131484334, + "grad_norm": 2.4819061756134033, + "learning_rate": 3.589645984012182e-06, + "loss": 1.4367, + "step": 1049 + }, + { + "epoch": 1.078582434514638, + "grad_norm": 2.459517002105713, + "learning_rate": 3.587742672249715e-06, + "loss": 1.3882, + "step": 1050 + }, + { + "epoch": 1.078582434514638, + "eval_loss": 1.7557458877563477, + "eval_runtime": 38.0271, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1050 + }, + { + "epoch": 1.0796096558808423, + "grad_norm": 2.5434813499450684, + "learning_rate": 3.585839360487248e-06, + "loss": 1.3926, + "step": 1051 + }, + { + "epoch": 1.0806368772470467, + "grad_norm": 2.6173205375671387, + "learning_rate": 3.5839360487247815e-06, + "loss": 1.4044, + "step": 1052 + }, + { + "epoch": 1.081664098613251, + "grad_norm": 2.677584171295166, + "learning_rate": 3.582032736962315e-06, + "loss": 1.4695, + "step": 1053 + }, + { + "epoch": 1.0826913199794557, + "grad_norm": 2.4151623249053955, + "learning_rate": 3.5801294251998483e-06, + "loss": 1.3699, + "step": 1054 + }, + { + "epoch": 1.08371854134566, + "grad_norm": 2.4683942794799805, + "learning_rate": 3.578226113437381e-06, + "loss": 1.4132, + "step": 1055 + }, + { + "epoch": 1.0847457627118644, + "grad_norm": 2.6076581478118896, + "learning_rate": 3.5763228016749146e-06, + "loss": 1.4729, + "step": 1056 + }, + { + "epoch": 1.0857729840780688, + "grad_norm": 2.4766876697540283, + "learning_rate": 3.574419489912448e-06, + "loss": 1.4196, + "step": 1057 + }, + { + "epoch": 1.0868002054442731, + "grad_norm": 2.466458559036255, + "learning_rate": 3.5725161781499813e-06, + "loss": 1.4049, + "step": 1058 + }, + { + "epoch": 1.0878274268104777, + "grad_norm": 2.4220409393310547, + "learning_rate": 3.5706128663875147e-06, + "loss": 1.4246, + "step": 1059 + }, + { + "epoch": 1.088854648176682, + "grad_norm": 2.5571770668029785, + "learning_rate": 3.5687095546250476e-06, + "loss": 1.4397, + "step": 1060 + }, + { + "epoch": 1.088854648176682, + "eval_loss": 1.7701127529144287, + "eval_runtime": 38.0754, + "eval_samples_per_second": 26.264, + "eval_steps_per_second": 1.103, + "step": 1060 + }, + { + "epoch": 1.0898818695428865, + "grad_norm": 2.648768186569214, + "learning_rate": 3.566806242862581e-06, + "loss": 1.423, + "step": 1061 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 2.403651714324951, + "learning_rate": 3.5649029311001143e-06, + "loss": 1.3995, + "step": 1062 + }, + { + "epoch": 1.0919363122752954, + "grad_norm": 2.465538263320923, + "learning_rate": 3.5629996193376477e-06, + "loss": 1.4139, + "step": 1063 + }, + { + "epoch": 1.0929635336414998, + "grad_norm": 2.464662551879883, + "learning_rate": 3.561096307575181e-06, + "loss": 1.4499, + "step": 1064 + }, + { + "epoch": 1.0939907550077042, + "grad_norm": 2.4943597316741943, + "learning_rate": 3.559192995812714e-06, + "loss": 1.4555, + "step": 1065 + }, + { + "epoch": 1.0950179763739085, + "grad_norm": 2.6332199573516846, + "learning_rate": 3.557289684050248e-06, + "loss": 1.4484, + "step": 1066 + }, + { + "epoch": 1.0960451977401129, + "grad_norm": 2.4370667934417725, + "learning_rate": 3.555386372287781e-06, + "loss": 1.4588, + "step": 1067 + }, + { + "epoch": 1.0970724191063175, + "grad_norm": 2.568368673324585, + "learning_rate": 3.5534830605253145e-06, + "loss": 1.4269, + "step": 1068 + }, + { + "epoch": 1.0980996404725218, + "grad_norm": 2.4618077278137207, + "learning_rate": 3.551579748762848e-06, + "loss": 1.4165, + "step": 1069 + }, + { + "epoch": 1.0991268618387262, + "grad_norm": 2.551320791244507, + "learning_rate": 3.5496764370003812e-06, + "loss": 1.3991, + "step": 1070 + }, + { + "epoch": 1.0991268618387262, + "eval_loss": 1.7648736238479614, + "eval_runtime": 38.0273, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 1070 + }, + { + "epoch": 1.1001540832049306, + "grad_norm": 2.350905418395996, + "learning_rate": 3.5477731252379146e-06, + "loss": 1.3556, + "step": 1071 + }, + { + "epoch": 1.1011813045711352, + "grad_norm": 2.4294588565826416, + "learning_rate": 3.5458698134754475e-06, + "loss": 1.3801, + "step": 1072 + }, + { + "epoch": 1.1022085259373395, + "grad_norm": 2.576774835586548, + "learning_rate": 3.543966501712981e-06, + "loss": 1.3876, + "step": 1073 + }, + { + "epoch": 1.103235747303544, + "grad_norm": 2.716935873031616, + "learning_rate": 3.5420631899505142e-06, + "loss": 1.4907, + "step": 1074 + }, + { + "epoch": 1.1042629686697483, + "grad_norm": 2.425305128097534, + "learning_rate": 3.5401598781880476e-06, + "loss": 1.3161, + "step": 1075 + }, + { + "epoch": 1.1052901900359529, + "grad_norm": 2.54738712310791, + "learning_rate": 3.538256566425581e-06, + "loss": 1.4462, + "step": 1076 + }, + { + "epoch": 1.1063174114021572, + "grad_norm": 2.5109341144561768, + "learning_rate": 3.536353254663114e-06, + "loss": 1.4022, + "step": 1077 + }, + { + "epoch": 1.1073446327683616, + "grad_norm": 2.468749523162842, + "learning_rate": 3.5344499429006473e-06, + "loss": 1.3857, + "step": 1078 + }, + { + "epoch": 1.108371854134566, + "grad_norm": 2.5285134315490723, + "learning_rate": 3.5325466311381806e-06, + "loss": 1.4309, + "step": 1079 + }, + { + "epoch": 1.1093990755007703, + "grad_norm": 2.6391055583953857, + "learning_rate": 3.530643319375714e-06, + "loss": 1.3644, + "step": 1080 + }, + { + "epoch": 1.1093990755007703, + "eval_loss": 1.7846598625183105, + "eval_runtime": 38.0604, + "eval_samples_per_second": 26.274, + "eval_steps_per_second": 1.104, + "step": 1080 + }, + { + "epoch": 1.110426296866975, + "grad_norm": 2.660356283187866, + "learning_rate": 3.5287400076132474e-06, + "loss": 1.3922, + "step": 1081 + }, + { + "epoch": 1.1114535182331793, + "grad_norm": 2.5648741722106934, + "learning_rate": 3.5268366958507803e-06, + "loss": 1.4282, + "step": 1082 + }, + { + "epoch": 1.1124807395993837, + "grad_norm": 2.6862876415252686, + "learning_rate": 3.5249333840883137e-06, + "loss": 1.4208, + "step": 1083 + }, + { + "epoch": 1.113507960965588, + "grad_norm": 2.6200191974639893, + "learning_rate": 3.523030072325847e-06, + "loss": 1.3883, + "step": 1084 + }, + { + "epoch": 1.1145351823317924, + "grad_norm": 2.5802111625671387, + "learning_rate": 3.5211267605633804e-06, + "loss": 1.4033, + "step": 1085 + }, + { + "epoch": 1.115562403697997, + "grad_norm": 2.4148080348968506, + "learning_rate": 3.519223448800914e-06, + "loss": 1.4085, + "step": 1086 + }, + { + "epoch": 1.1165896250642013, + "grad_norm": 2.7452127933502197, + "learning_rate": 3.5173201370384475e-06, + "loss": 1.4021, + "step": 1087 + }, + { + "epoch": 1.1176168464304057, + "grad_norm": 2.6437106132507324, + "learning_rate": 3.515416825275981e-06, + "loss": 1.4161, + "step": 1088 + }, + { + "epoch": 1.11864406779661, + "grad_norm": 2.4709842205047607, + "learning_rate": 3.513513513513514e-06, + "loss": 1.4322, + "step": 1089 + }, + { + "epoch": 1.1196712891628147, + "grad_norm": 2.5138099193573, + "learning_rate": 3.511610201751047e-06, + "loss": 1.3435, + "step": 1090 + }, + { + "epoch": 1.1196712891628147, + "eval_loss": 1.7687023878097534, + "eval_runtime": 38.0137, + "eval_samples_per_second": 26.306, + "eval_steps_per_second": 1.105, + "step": 1090 + }, + { + "epoch": 1.120698510529019, + "grad_norm": 2.4194881916046143, + "learning_rate": 3.5097068899885806e-06, + "loss": 1.4133, + "step": 1091 + }, + { + "epoch": 1.1217257318952234, + "grad_norm": 2.6564371585845947, + "learning_rate": 3.507803578226114e-06, + "loss": 1.4759, + "step": 1092 + }, + { + "epoch": 1.1227529532614278, + "grad_norm": 2.4480180740356445, + "learning_rate": 3.5059002664636473e-06, + "loss": 1.3577, + "step": 1093 + }, + { + "epoch": 1.1237801746276324, + "grad_norm": 2.5538978576660156, + "learning_rate": 3.5039969547011802e-06, + "loss": 1.3984, + "step": 1094 + }, + { + "epoch": 1.1248073959938367, + "grad_norm": 2.6467089653015137, + "learning_rate": 3.5020936429387136e-06, + "loss": 1.4049, + "step": 1095 + }, + { + "epoch": 1.125834617360041, + "grad_norm": 2.6502673625946045, + "learning_rate": 3.500190331176247e-06, + "loss": 1.462, + "step": 1096 + }, + { + "epoch": 1.1268618387262455, + "grad_norm": 2.5696403980255127, + "learning_rate": 3.4982870194137803e-06, + "loss": 1.3926, + "step": 1097 + }, + { + "epoch": 1.1278890600924498, + "grad_norm": 2.5951733589172363, + "learning_rate": 3.4963837076513137e-06, + "loss": 1.4031, + "step": 1098 + }, + { + "epoch": 1.1289162814586544, + "grad_norm": 2.4910218715667725, + "learning_rate": 3.4944803958888466e-06, + "loss": 1.3348, + "step": 1099 + }, + { + "epoch": 1.1299435028248588, + "grad_norm": 2.6325888633728027, + "learning_rate": 3.49257708412638e-06, + "loss": 1.4641, + "step": 1100 + }, + { + "epoch": 1.1299435028248588, + "eval_loss": 1.7720683813095093, + "eval_runtime": 37.9585, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 1100 + }, + { + "epoch": 1.1309707241910631, + "grad_norm": 2.6712775230407715, + "learning_rate": 3.4906737723639133e-06, + "loss": 1.4726, + "step": 1101 + }, + { + "epoch": 1.1319979455572675, + "grad_norm": 2.580404043197632, + "learning_rate": 3.4887704606014467e-06, + "loss": 1.4158, + "step": 1102 + }, + { + "epoch": 1.133025166923472, + "grad_norm": 2.467453718185425, + "learning_rate": 3.48686714883898e-06, + "loss": 1.4636, + "step": 1103 + }, + { + "epoch": 1.1340523882896765, + "grad_norm": 2.627136468887329, + "learning_rate": 3.484963837076513e-06, + "loss": 1.4082, + "step": 1104 + }, + { + "epoch": 1.1350796096558808, + "grad_norm": 2.6198604106903076, + "learning_rate": 3.4830605253140464e-06, + "loss": 1.3628, + "step": 1105 + }, + { + "epoch": 1.1361068310220852, + "grad_norm": 2.439512014389038, + "learning_rate": 3.4811572135515797e-06, + "loss": 1.3885, + "step": 1106 + }, + { + "epoch": 1.1371340523882898, + "grad_norm": 2.5543010234832764, + "learning_rate": 3.4792539017891135e-06, + "loss": 1.3674, + "step": 1107 + }, + { + "epoch": 1.1381612737544942, + "grad_norm": 2.7263364791870117, + "learning_rate": 3.477350590026647e-06, + "loss": 1.4387, + "step": 1108 + }, + { + "epoch": 1.1391884951206985, + "grad_norm": 2.5030503273010254, + "learning_rate": 3.4754472782641802e-06, + "loss": 1.4222, + "step": 1109 + }, + { + "epoch": 1.140215716486903, + "grad_norm": 2.5090320110321045, + "learning_rate": 3.4735439665017136e-06, + "loss": 1.3981, + "step": 1110 + }, + { + "epoch": 1.140215716486903, + "eval_loss": 1.769408941268921, + "eval_runtime": 37.608, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 1110 + }, + { + "epoch": 1.1412429378531073, + "grad_norm": 2.575934648513794, + "learning_rate": 3.4716406547392465e-06, + "loss": 1.4074, + "step": 1111 + }, + { + "epoch": 1.1422701592193119, + "grad_norm": 2.6381359100341797, + "learning_rate": 3.46973734297678e-06, + "loss": 1.4162, + "step": 1112 + }, + { + "epoch": 1.1432973805855162, + "grad_norm": 2.7605674266815186, + "learning_rate": 3.4678340312143133e-06, + "loss": 1.4616, + "step": 1113 + }, + { + "epoch": 1.1443246019517206, + "grad_norm": 2.5548479557037354, + "learning_rate": 3.4659307194518466e-06, + "loss": 1.4511, + "step": 1114 + }, + { + "epoch": 1.145351823317925, + "grad_norm": 2.471168279647827, + "learning_rate": 3.46402740768938e-06, + "loss": 1.3777, + "step": 1115 + }, + { + "epoch": 1.1463790446841293, + "grad_norm": 2.4416346549987793, + "learning_rate": 3.462124095926913e-06, + "loss": 1.3805, + "step": 1116 + }, + { + "epoch": 1.147406266050334, + "grad_norm": 2.4972171783447266, + "learning_rate": 3.4602207841644463e-06, + "loss": 1.4282, + "step": 1117 + }, + { + "epoch": 1.1484334874165383, + "grad_norm": 2.4379162788391113, + "learning_rate": 3.4583174724019797e-06, + "loss": 1.4744, + "step": 1118 + }, + { + "epoch": 1.1494607087827426, + "grad_norm": 2.508047580718994, + "learning_rate": 3.456414160639513e-06, + "loss": 1.4303, + "step": 1119 + }, + { + "epoch": 1.150487930148947, + "grad_norm": 2.4299631118774414, + "learning_rate": 3.4545108488770464e-06, + "loss": 1.4179, + "step": 1120 + }, + { + "epoch": 1.150487930148947, + "eval_loss": 1.7768158912658691, + "eval_runtime": 37.736, + "eval_samples_per_second": 26.5, + "eval_steps_per_second": 1.113, + "step": 1120 + }, + { + "epoch": 1.1515151515151516, + "grad_norm": 2.41538143157959, + "learning_rate": 3.4526075371145793e-06, + "loss": 1.4235, + "step": 1121 + }, + { + "epoch": 1.152542372881356, + "grad_norm": 2.5614655017852783, + "learning_rate": 3.4507042253521127e-06, + "loss": 1.4879, + "step": 1122 + }, + { + "epoch": 1.1535695942475603, + "grad_norm": 2.4826605319976807, + "learning_rate": 3.448800913589646e-06, + "loss": 1.3968, + "step": 1123 + }, + { + "epoch": 1.1545968156137647, + "grad_norm": 2.5341391563415527, + "learning_rate": 3.4468976018271794e-06, + "loss": 1.4903, + "step": 1124 + }, + { + "epoch": 1.1556240369799693, + "grad_norm": 2.493818998336792, + "learning_rate": 3.4449942900647128e-06, + "loss": 1.4466, + "step": 1125 + }, + { + "epoch": 1.1566512583461737, + "grad_norm": 2.6057965755462646, + "learning_rate": 3.443090978302246e-06, + "loss": 1.4425, + "step": 1126 + }, + { + "epoch": 1.157678479712378, + "grad_norm": 2.5481390953063965, + "learning_rate": 3.44118766653978e-06, + "loss": 1.4296, + "step": 1127 + }, + { + "epoch": 1.1587057010785824, + "grad_norm": 2.5753650665283203, + "learning_rate": 3.439284354777313e-06, + "loss": 1.4408, + "step": 1128 + }, + { + "epoch": 1.1597329224447868, + "grad_norm": 2.5423803329467773, + "learning_rate": 3.4373810430148462e-06, + "loss": 1.373, + "step": 1129 + }, + { + "epoch": 1.1607601438109914, + "grad_norm": 2.593487024307251, + "learning_rate": 3.4354777312523796e-06, + "loss": 1.3992, + "step": 1130 + }, + { + "epoch": 1.1607601438109914, + "eval_loss": 1.7784411907196045, + "eval_runtime": 37.8687, + "eval_samples_per_second": 26.407, + "eval_steps_per_second": 1.109, + "step": 1130 + }, + { + "epoch": 1.1617873651771957, + "grad_norm": 2.4693503379821777, + "learning_rate": 3.433574419489913e-06, + "loss": 1.4339, + "step": 1131 + }, + { + "epoch": 1.1628145865434, + "grad_norm": 2.5477423667907715, + "learning_rate": 3.4316711077274463e-06, + "loss": 1.4035, + "step": 1132 + }, + { + "epoch": 1.1638418079096045, + "grad_norm": 2.5650672912597656, + "learning_rate": 3.4297677959649792e-06, + "loss": 1.4517, + "step": 1133 + }, + { + "epoch": 1.1648690292758088, + "grad_norm": 2.3217973709106445, + "learning_rate": 3.4278644842025126e-06, + "loss": 1.3983, + "step": 1134 + }, + { + "epoch": 1.1658962506420134, + "grad_norm": 2.674342155456543, + "learning_rate": 3.425961172440046e-06, + "loss": 1.4799, + "step": 1135 + }, + { + "epoch": 1.1669234720082178, + "grad_norm": 2.4861397743225098, + "learning_rate": 3.4240578606775793e-06, + "loss": 1.414, + "step": 1136 + }, + { + "epoch": 1.1679506933744221, + "grad_norm": 2.4306106567382812, + "learning_rate": 3.4221545489151127e-06, + "loss": 1.3918, + "step": 1137 + }, + { + "epoch": 1.1689779147406265, + "grad_norm": 2.6873955726623535, + "learning_rate": 3.4202512371526456e-06, + "loss": 1.4979, + "step": 1138 + }, + { + "epoch": 1.170005136106831, + "grad_norm": 2.559042453765869, + "learning_rate": 3.418347925390179e-06, + "loss": 1.4424, + "step": 1139 + }, + { + "epoch": 1.1710323574730355, + "grad_norm": 2.397489547729492, + "learning_rate": 3.4164446136277124e-06, + "loss": 1.418, + "step": 1140 + }, + { + "epoch": 1.1710323574730355, + "eval_loss": 1.7702736854553223, + "eval_runtime": 37.8963, + "eval_samples_per_second": 26.388, + "eval_steps_per_second": 1.108, + "step": 1140 + }, + { + "epoch": 1.1720595788392398, + "grad_norm": 2.4295787811279297, + "learning_rate": 3.4145413018652457e-06, + "loss": 1.4436, + "step": 1141 + }, + { + "epoch": 1.1730868002054442, + "grad_norm": 2.503202199935913, + "learning_rate": 3.412637990102779e-06, + "loss": 1.4598, + "step": 1142 + }, + { + "epoch": 1.1741140215716488, + "grad_norm": 2.540195941925049, + "learning_rate": 3.4107346783403125e-06, + "loss": 1.3704, + "step": 1143 + }, + { + "epoch": 1.1751412429378532, + "grad_norm": 2.4832522869110107, + "learning_rate": 3.4088313665778454e-06, + "loss": 1.4081, + "step": 1144 + }, + { + "epoch": 1.1761684643040575, + "grad_norm": 2.5639121532440186, + "learning_rate": 3.4069280548153788e-06, + "loss": 1.3866, + "step": 1145 + }, + { + "epoch": 1.177195685670262, + "grad_norm": 2.526610851287842, + "learning_rate": 3.405024743052912e-06, + "loss": 1.4181, + "step": 1146 + }, + { + "epoch": 1.1782229070364663, + "grad_norm": 2.6945254802703857, + "learning_rate": 3.403121431290446e-06, + "loss": 1.3871, + "step": 1147 + }, + { + "epoch": 1.1792501284026708, + "grad_norm": 2.6561484336853027, + "learning_rate": 3.4012181195279793e-06, + "loss": 1.3987, + "step": 1148 + }, + { + "epoch": 1.1802773497688752, + "grad_norm": 2.6104650497436523, + "learning_rate": 3.3993148077655126e-06, + "loss": 1.4057, + "step": 1149 + }, + { + "epoch": 1.1813045711350796, + "grad_norm": 2.571707248687744, + "learning_rate": 3.3974114960030456e-06, + "loss": 1.3606, + "step": 1150 + }, + { + "epoch": 1.1813045711350796, + "eval_loss": 1.7780840396881104, + "eval_runtime": 37.8616, + "eval_samples_per_second": 26.412, + "eval_steps_per_second": 1.109, + "step": 1150 + }, + { + "epoch": 1.182331792501284, + "grad_norm": 2.5804402828216553, + "learning_rate": 3.395508184240579e-06, + "loss": 1.3876, + "step": 1151 + }, + { + "epoch": 1.1833590138674885, + "grad_norm": 2.832446336746216, + "learning_rate": 3.3936048724781123e-06, + "loss": 1.4674, + "step": 1152 + }, + { + "epoch": 1.184386235233693, + "grad_norm": 2.6478688716888428, + "learning_rate": 3.3917015607156457e-06, + "loss": 1.3643, + "step": 1153 + }, + { + "epoch": 1.1854134565998973, + "grad_norm": 2.579416513442993, + "learning_rate": 3.389798248953179e-06, + "loss": 1.3874, + "step": 1154 + }, + { + "epoch": 1.1864406779661016, + "grad_norm": 2.487861156463623, + "learning_rate": 3.387894937190712e-06, + "loss": 1.4323, + "step": 1155 + }, + { + "epoch": 1.1874678993323062, + "grad_norm": 2.5020716190338135, + "learning_rate": 3.3859916254282453e-06, + "loss": 1.423, + "step": 1156 + }, + { + "epoch": 1.1884951206985106, + "grad_norm": 2.6588852405548096, + "learning_rate": 3.3840883136657787e-06, + "loss": 1.4808, + "step": 1157 + }, + { + "epoch": 1.189522342064715, + "grad_norm": 2.5702178478240967, + "learning_rate": 3.382185001903312e-06, + "loss": 1.4007, + "step": 1158 + }, + { + "epoch": 1.1905495634309193, + "grad_norm": 2.5082061290740967, + "learning_rate": 3.3802816901408454e-06, + "loss": 1.4278, + "step": 1159 + }, + { + "epoch": 1.1915767847971237, + "grad_norm": 2.4165053367614746, + "learning_rate": 3.3783783783783788e-06, + "loss": 1.4088, + "step": 1160 + }, + { + "epoch": 1.1915767847971237, + "eval_loss": 1.7722209692001343, + "eval_runtime": 37.8843, + "eval_samples_per_second": 26.396, + "eval_steps_per_second": 1.109, + "step": 1160 + }, + { + "epoch": 1.1926040061633283, + "grad_norm": 2.4510512351989746, + "learning_rate": 3.3764750666159117e-06, + "loss": 1.4089, + "step": 1161 + }, + { + "epoch": 1.1936312275295327, + "grad_norm": 2.5006556510925293, + "learning_rate": 3.374571754853445e-06, + "loss": 1.3988, + "step": 1162 + }, + { + "epoch": 1.194658448895737, + "grad_norm": 2.416264057159424, + "learning_rate": 3.3726684430909784e-06, + "loss": 1.3576, + "step": 1163 + }, + { + "epoch": 1.1956856702619414, + "grad_norm": 2.465613842010498, + "learning_rate": 3.370765131328512e-06, + "loss": 1.3796, + "step": 1164 + }, + { + "epoch": 1.1967128916281458, + "grad_norm": 2.510542154312134, + "learning_rate": 3.368861819566045e-06, + "loss": 1.391, + "step": 1165 + }, + { + "epoch": 1.1977401129943503, + "grad_norm": 2.729495048522949, + "learning_rate": 3.366958507803578e-06, + "loss": 1.431, + "step": 1166 + }, + { + "epoch": 1.1987673343605547, + "grad_norm": 2.4473495483398438, + "learning_rate": 3.365055196041112e-06, + "loss": 1.376, + "step": 1167 + }, + { + "epoch": 1.199794555726759, + "grad_norm": 2.3869211673736572, + "learning_rate": 3.3631518842786452e-06, + "loss": 1.4498, + "step": 1168 + }, + { + "epoch": 1.2008217770929634, + "grad_norm": 2.4202256202697754, + "learning_rate": 3.3612485725161786e-06, + "loss": 1.3754, + "step": 1169 + }, + { + "epoch": 1.201848998459168, + "grad_norm": 2.452707529067993, + "learning_rate": 3.359345260753712e-06, + "loss": 1.4647, + "step": 1170 + }, + { + "epoch": 1.201848998459168, + "eval_loss": 1.769252896308899, + "eval_runtime": 37.9385, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1170 + }, + { + "epoch": 1.2028762198253724, + "grad_norm": 2.56616473197937, + "learning_rate": 3.3574419489912453e-06, + "loss": 1.443, + "step": 1171 + }, + { + "epoch": 1.2039034411915768, + "grad_norm": 2.6483540534973145, + "learning_rate": 3.3555386372287783e-06, + "loss": 1.4755, + "step": 1172 + }, + { + "epoch": 1.2049306625577811, + "grad_norm": 2.5579090118408203, + "learning_rate": 3.3536353254663116e-06, + "loss": 1.3732, + "step": 1173 + }, + { + "epoch": 1.2059578839239857, + "grad_norm": 2.5310683250427246, + "learning_rate": 3.351732013703845e-06, + "loss": 1.4077, + "step": 1174 + }, + { + "epoch": 1.20698510529019, + "grad_norm": 2.616957187652588, + "learning_rate": 3.3498287019413784e-06, + "loss": 1.4237, + "step": 1175 + }, + { + "epoch": 1.2080123266563945, + "grad_norm": 2.6008999347686768, + "learning_rate": 3.3479253901789117e-06, + "loss": 1.4381, + "step": 1176 + }, + { + "epoch": 1.2090395480225988, + "grad_norm": 2.5235824584960938, + "learning_rate": 3.346022078416445e-06, + "loss": 1.3471, + "step": 1177 + }, + { + "epoch": 1.2100667693888032, + "grad_norm": 2.3749704360961914, + "learning_rate": 3.344118766653978e-06, + "loss": 1.417, + "step": 1178 + }, + { + "epoch": 1.2110939907550078, + "grad_norm": 2.493380069732666, + "learning_rate": 3.3422154548915114e-06, + "loss": 1.3924, + "step": 1179 + }, + { + "epoch": 1.2121212121212122, + "grad_norm": 2.4233086109161377, + "learning_rate": 3.3403121431290448e-06, + "loss": 1.4257, + "step": 1180 + }, + { + "epoch": 1.2121212121212122, + "eval_loss": 1.7660146951675415, + "eval_runtime": 37.9268, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 1180 + }, + { + "epoch": 1.2131484334874165, + "grad_norm": 2.535982608795166, + "learning_rate": 3.338408831366578e-06, + "loss": 1.3587, + "step": 1181 + }, + { + "epoch": 1.214175654853621, + "grad_norm": 2.5744845867156982, + "learning_rate": 3.3365055196041115e-06, + "loss": 1.436, + "step": 1182 + }, + { + "epoch": 1.2152028762198253, + "grad_norm": 2.5544042587280273, + "learning_rate": 3.3346022078416444e-06, + "loss": 1.464, + "step": 1183 + }, + { + "epoch": 1.2162300975860298, + "grad_norm": 2.4980483055114746, + "learning_rate": 3.3326988960791778e-06, + "loss": 1.3915, + "step": 1184 + }, + { + "epoch": 1.2172573189522342, + "grad_norm": 2.403468608856201, + "learning_rate": 3.330795584316711e-06, + "loss": 1.3239, + "step": 1185 + }, + { + "epoch": 1.2182845403184386, + "grad_norm": 2.6861846446990967, + "learning_rate": 3.3288922725542445e-06, + "loss": 1.4013, + "step": 1186 + }, + { + "epoch": 1.219311761684643, + "grad_norm": 2.525033712387085, + "learning_rate": 3.326988960791778e-06, + "loss": 1.4135, + "step": 1187 + }, + { + "epoch": 1.2203389830508475, + "grad_norm": 2.6705894470214844, + "learning_rate": 3.3250856490293117e-06, + "loss": 1.4587, + "step": 1188 + }, + { + "epoch": 1.221366204417052, + "grad_norm": 2.6108217239379883, + "learning_rate": 3.323182337266845e-06, + "loss": 1.4482, + "step": 1189 + }, + { + "epoch": 1.2223934257832563, + "grad_norm": 2.5227413177490234, + "learning_rate": 3.321279025504378e-06, + "loss": 1.432, + "step": 1190 + }, + { + "epoch": 1.2223934257832563, + "eval_loss": 1.756821870803833, + "eval_runtime": 37.9929, + "eval_samples_per_second": 26.321, + "eval_steps_per_second": 1.105, + "step": 1190 + }, + { + "epoch": 1.2234206471494606, + "grad_norm": 2.469766139984131, + "learning_rate": 3.3193757137419113e-06, + "loss": 1.4215, + "step": 1191 + }, + { + "epoch": 1.2244478685156652, + "grad_norm": 2.5257861614227295, + "learning_rate": 3.3174724019794447e-06, + "loss": 1.4624, + "step": 1192 + }, + { + "epoch": 1.2254750898818696, + "grad_norm": 2.5281760692596436, + "learning_rate": 3.315569090216978e-06, + "loss": 1.3799, + "step": 1193 + }, + { + "epoch": 1.226502311248074, + "grad_norm": 2.701521396636963, + "learning_rate": 3.3136657784545114e-06, + "loss": 1.4631, + "step": 1194 + }, + { + "epoch": 1.2275295326142783, + "grad_norm": 2.4239444732666016, + "learning_rate": 3.3117624666920443e-06, + "loss": 1.3746, + "step": 1195 + }, + { + "epoch": 1.2285567539804827, + "grad_norm": 2.430377721786499, + "learning_rate": 3.3098591549295777e-06, + "loss": 1.3584, + "step": 1196 + }, + { + "epoch": 1.2295839753466873, + "grad_norm": 2.7177658081054688, + "learning_rate": 3.307955843167111e-06, + "loss": 1.4301, + "step": 1197 + }, + { + "epoch": 1.2306111967128917, + "grad_norm": 2.512943983078003, + "learning_rate": 3.3060525314046444e-06, + "loss": 1.4111, + "step": 1198 + }, + { + "epoch": 1.231638418079096, + "grad_norm": 2.5269784927368164, + "learning_rate": 3.304149219642178e-06, + "loss": 1.4096, + "step": 1199 + }, + { + "epoch": 1.2326656394453004, + "grad_norm": 2.476409673690796, + "learning_rate": 3.3022459078797107e-06, + "loss": 1.4322, + "step": 1200 + }, + { + "epoch": 1.2326656394453004, + "eval_loss": 1.7645230293273926, + "eval_runtime": 37.9898, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1200 + }, + { + "epoch": 1.2336928608115048, + "grad_norm": 2.5339248180389404, + "learning_rate": 3.300342596117244e-06, + "loss": 1.4093, + "step": 1201 + }, + { + "epoch": 1.2347200821777093, + "grad_norm": 2.3505656719207764, + "learning_rate": 3.2984392843547775e-06, + "loss": 1.3623, + "step": 1202 + }, + { + "epoch": 1.2357473035439137, + "grad_norm": 2.400259256362915, + "learning_rate": 3.296535972592311e-06, + "loss": 1.3512, + "step": 1203 + }, + { + "epoch": 1.236774524910118, + "grad_norm": 2.561558723449707, + "learning_rate": 3.294632660829844e-06, + "loss": 1.3946, + "step": 1204 + }, + { + "epoch": 1.2378017462763227, + "grad_norm": 2.623100519180298, + "learning_rate": 3.292729349067377e-06, + "loss": 1.3972, + "step": 1205 + }, + { + "epoch": 1.238828967642527, + "grad_norm": 2.4889469146728516, + "learning_rate": 3.2908260373049105e-06, + "loss": 1.381, + "step": 1206 + }, + { + "epoch": 1.2398561890087314, + "grad_norm": 2.457334518432617, + "learning_rate": 3.288922725542444e-06, + "loss": 1.4345, + "step": 1207 + }, + { + "epoch": 1.2408834103749358, + "grad_norm": 2.440988540649414, + "learning_rate": 3.2870194137799776e-06, + "loss": 1.3901, + "step": 1208 + }, + { + "epoch": 1.2419106317411401, + "grad_norm": 2.4992308616638184, + "learning_rate": 3.285116102017511e-06, + "loss": 1.3694, + "step": 1209 + }, + { + "epoch": 1.2429378531073447, + "grad_norm": 2.611128091812134, + "learning_rate": 3.2832127902550444e-06, + "loss": 1.4344, + "step": 1210 + }, + { + "epoch": 1.2429378531073447, + "eval_loss": 1.763407826423645, + "eval_runtime": 37.7142, + "eval_samples_per_second": 26.515, + "eval_steps_per_second": 1.114, + "step": 1210 + }, + { + "epoch": 1.243965074473549, + "grad_norm": 2.6573522090911865, + "learning_rate": 3.2813094784925777e-06, + "loss": 1.4204, + "step": 1211 + }, + { + "epoch": 1.2449922958397535, + "grad_norm": 2.4785778522491455, + "learning_rate": 3.2794061667301107e-06, + "loss": 1.3617, + "step": 1212 + }, + { + "epoch": 1.2460195172059578, + "grad_norm": 2.4831974506378174, + "learning_rate": 3.277502854967644e-06, + "loss": 1.4004, + "step": 1213 + }, + { + "epoch": 1.2470467385721622, + "grad_norm": 2.6142327785491943, + "learning_rate": 3.2755995432051774e-06, + "loss": 1.3881, + "step": 1214 + }, + { + "epoch": 1.2480739599383668, + "grad_norm": 2.447962999343872, + "learning_rate": 3.2736962314427107e-06, + "loss": 1.3984, + "step": 1215 + }, + { + "epoch": 1.2491011813045712, + "grad_norm": 2.3895885944366455, + "learning_rate": 3.271792919680244e-06, + "loss": 1.4099, + "step": 1216 + }, + { + "epoch": 1.2501284026707755, + "grad_norm": 2.599364995956421, + "learning_rate": 3.269889607917777e-06, + "loss": 1.4575, + "step": 1217 + }, + { + "epoch": 1.25115562403698, + "grad_norm": 2.6294758319854736, + "learning_rate": 3.2679862961553104e-06, + "loss": 1.3752, + "step": 1218 + }, + { + "epoch": 1.2521828454031843, + "grad_norm": 2.50923228263855, + "learning_rate": 3.2660829843928438e-06, + "loss": 1.4017, + "step": 1219 + }, + { + "epoch": 1.2532100667693888, + "grad_norm": 2.626434564590454, + "learning_rate": 3.264179672630377e-06, + "loss": 1.4505, + "step": 1220 + }, + { + "epoch": 1.2532100667693888, + "eval_loss": 1.7672324180603027, + "eval_runtime": 37.8605, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 1220 + }, + { + "epoch": 1.2542372881355932, + "grad_norm": 2.541663885116577, + "learning_rate": 3.2622763608679105e-06, + "loss": 1.4129, + "step": 1221 + }, + { + "epoch": 1.2552645095017976, + "grad_norm": 2.559877395629883, + "learning_rate": 3.2603730491054434e-06, + "loss": 1.3793, + "step": 1222 + }, + { + "epoch": 1.2562917308680022, + "grad_norm": 2.4085795879364014, + "learning_rate": 3.258469737342977e-06, + "loss": 1.3895, + "step": 1223 + }, + { + "epoch": 1.2573189522342065, + "grad_norm": 2.2942678928375244, + "learning_rate": 3.25656642558051e-06, + "loss": 1.3713, + "step": 1224 + }, + { + "epoch": 1.258346173600411, + "grad_norm": 2.561086893081665, + "learning_rate": 3.2546631138180435e-06, + "loss": 1.3835, + "step": 1225 + }, + { + "epoch": 1.2593733949666153, + "grad_norm": 2.4947667121887207, + "learning_rate": 3.252759802055577e-06, + "loss": 1.3793, + "step": 1226 + }, + { + "epoch": 1.2604006163328196, + "grad_norm": 2.6293253898620605, + "learning_rate": 3.25085649029311e-06, + "loss": 1.3935, + "step": 1227 + }, + { + "epoch": 1.2614278376990242, + "grad_norm": 2.5461983680725098, + "learning_rate": 3.248953178530644e-06, + "loss": 1.4201, + "step": 1228 + }, + { + "epoch": 1.2624550590652286, + "grad_norm": 2.381537675857544, + "learning_rate": 3.247049866768177e-06, + "loss": 1.3932, + "step": 1229 + }, + { + "epoch": 1.263482280431433, + "grad_norm": 2.381558418273926, + "learning_rate": 3.2451465550057103e-06, + "loss": 1.3425, + "step": 1230 + }, + { + "epoch": 1.263482280431433, + "eval_loss": 1.7713288068771362, + "eval_runtime": 37.9629, + "eval_samples_per_second": 26.342, + "eval_steps_per_second": 1.106, + "step": 1230 + }, + { + "epoch": 1.2645095017976373, + "grad_norm": 2.5877468585968018, + "learning_rate": 3.2432432432432437e-06, + "loss": 1.3917, + "step": 1231 + }, + { + "epoch": 1.2655367231638417, + "grad_norm": 2.6057848930358887, + "learning_rate": 3.241339931480777e-06, + "loss": 1.4217, + "step": 1232 + }, + { + "epoch": 1.2665639445300463, + "grad_norm": 2.5157413482666016, + "learning_rate": 3.2394366197183104e-06, + "loss": 1.427, + "step": 1233 + }, + { + "epoch": 1.2675911658962506, + "grad_norm": 2.4153692722320557, + "learning_rate": 3.2375333079558434e-06, + "loss": 1.3822, + "step": 1234 + }, + { + "epoch": 1.268618387262455, + "grad_norm": 2.5459582805633545, + "learning_rate": 3.2356299961933767e-06, + "loss": 1.4152, + "step": 1235 + }, + { + "epoch": 1.2696456086286596, + "grad_norm": 2.606468915939331, + "learning_rate": 3.23372668443091e-06, + "loss": 1.4278, + "step": 1236 + }, + { + "epoch": 1.270672829994864, + "grad_norm": 2.5061633586883545, + "learning_rate": 3.2318233726684435e-06, + "loss": 1.4512, + "step": 1237 + }, + { + "epoch": 1.2717000513610683, + "grad_norm": 2.3693087100982666, + "learning_rate": 3.229920060905977e-06, + "loss": 1.3791, + "step": 1238 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 2.479583978652954, + "learning_rate": 3.2280167491435098e-06, + "loss": 1.3814, + "step": 1239 + }, + { + "epoch": 1.273754494093477, + "grad_norm": 2.3258883953094482, + "learning_rate": 3.226113437381043e-06, + "loss": 1.3904, + "step": 1240 + }, + { + "epoch": 1.273754494093477, + "eval_loss": 1.7602728605270386, + "eval_runtime": 37.9604, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 1240 + }, + { + "epoch": 1.2747817154596817, + "grad_norm": 2.4170308113098145, + "learning_rate": 3.2242101256185765e-06, + "loss": 1.4045, + "step": 1241 + }, + { + "epoch": 1.275808936825886, + "grad_norm": 2.578066825866699, + "learning_rate": 3.22230681385611e-06, + "loss": 1.4556, + "step": 1242 + }, + { + "epoch": 1.2768361581920904, + "grad_norm": 2.558187246322632, + "learning_rate": 3.220403502093643e-06, + "loss": 1.4036, + "step": 1243 + }, + { + "epoch": 1.2778633795582948, + "grad_norm": 2.598916530609131, + "learning_rate": 3.218500190331176e-06, + "loss": 1.4462, + "step": 1244 + }, + { + "epoch": 1.2788906009244991, + "grad_norm": 2.4528257846832275, + "learning_rate": 3.2165968785687095e-06, + "loss": 1.3855, + "step": 1245 + }, + { + "epoch": 1.2799178222907037, + "grad_norm": 2.5937459468841553, + "learning_rate": 3.214693566806243e-06, + "loss": 1.4191, + "step": 1246 + }, + { + "epoch": 1.280945043656908, + "grad_norm": 2.5324676036834717, + "learning_rate": 3.2127902550437762e-06, + "loss": 1.4162, + "step": 1247 + }, + { + "epoch": 1.2819722650231125, + "grad_norm": 2.4849791526794434, + "learning_rate": 3.21088694328131e-06, + "loss": 1.4037, + "step": 1248 + }, + { + "epoch": 1.2829994863893168, + "grad_norm": 2.389582633972168, + "learning_rate": 3.2089836315188434e-06, + "loss": 1.3966, + "step": 1249 + }, + { + "epoch": 1.2840267077555212, + "grad_norm": 2.387909173965454, + "learning_rate": 3.2070803197563767e-06, + "loss": 1.396, + "step": 1250 + }, + { + "epoch": 1.2840267077555212, + "eval_loss": 1.7642556428909302, + "eval_runtime": 37.9446, + "eval_samples_per_second": 26.354, + "eval_steps_per_second": 1.107, + "step": 1250 + }, + { + "epoch": 1.2850539291217258, + "grad_norm": 2.5014612674713135, + "learning_rate": 3.2051770079939097e-06, + "loss": 1.4128, + "step": 1251 + }, + { + "epoch": 1.2860811504879301, + "grad_norm": 2.497950553894043, + "learning_rate": 3.203273696231443e-06, + "loss": 1.4412, + "step": 1252 + }, + { + "epoch": 1.2871083718541345, + "grad_norm": 2.4371297359466553, + "learning_rate": 3.2013703844689764e-06, + "loss": 1.3754, + "step": 1253 + }, + { + "epoch": 1.288135593220339, + "grad_norm": 2.4336745738983154, + "learning_rate": 3.1994670727065098e-06, + "loss": 1.4258, + "step": 1254 + }, + { + "epoch": 1.2891628145865435, + "grad_norm": 2.5165536403656006, + "learning_rate": 3.197563760944043e-06, + "loss": 1.4414, + "step": 1255 + }, + { + "epoch": 1.2901900359527478, + "grad_norm": 2.5060458183288574, + "learning_rate": 3.195660449181576e-06, + "loss": 1.4336, + "step": 1256 + }, + { + "epoch": 1.2912172573189522, + "grad_norm": 2.584284543991089, + "learning_rate": 3.1937571374191094e-06, + "loss": 1.3768, + "step": 1257 + }, + { + "epoch": 1.2922444786851566, + "grad_norm": 2.615629196166992, + "learning_rate": 3.191853825656643e-06, + "loss": 1.401, + "step": 1258 + }, + { + "epoch": 1.2932717000513612, + "grad_norm": 2.5989596843719482, + "learning_rate": 3.189950513894176e-06, + "loss": 1.3915, + "step": 1259 + }, + { + "epoch": 1.2942989214175655, + "grad_norm": 2.5507843494415283, + "learning_rate": 3.1880472021317095e-06, + "loss": 1.3979, + "step": 1260 + }, + { + "epoch": 1.2942989214175655, + "eval_loss": 1.7715612649917603, + "eval_runtime": 37.9859, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1260 + }, + { + "epoch": 1.29532614278377, + "grad_norm": 2.512140989303589, + "learning_rate": 3.1861438903692425e-06, + "loss": 1.3523, + "step": 1261 + }, + { + "epoch": 1.2963533641499743, + "grad_norm": 2.4311885833740234, + "learning_rate": 3.184240578606776e-06, + "loss": 1.3828, + "step": 1262 + }, + { + "epoch": 1.2973805855161786, + "grad_norm": 2.439265489578247, + "learning_rate": 3.182337266844309e-06, + "loss": 1.402, + "step": 1263 + }, + { + "epoch": 1.2984078068823832, + "grad_norm": 2.5773651599884033, + "learning_rate": 3.1804339550818426e-06, + "loss": 1.4362, + "step": 1264 + }, + { + "epoch": 1.2994350282485876, + "grad_norm": 2.456021547317505, + "learning_rate": 3.178530643319376e-06, + "loss": 1.3918, + "step": 1265 + }, + { + "epoch": 1.300462249614792, + "grad_norm": 2.44404673576355, + "learning_rate": 3.176627331556909e-06, + "loss": 1.3662, + "step": 1266 + }, + { + "epoch": 1.3014894709809965, + "grad_norm": 2.778027057647705, + "learning_rate": 3.1747240197944422e-06, + "loss": 1.4112, + "step": 1267 + }, + { + "epoch": 1.3025166923472007, + "grad_norm": 2.476496934890747, + "learning_rate": 3.172820708031976e-06, + "loss": 1.3654, + "step": 1268 + }, + { + "epoch": 1.3035439137134053, + "grad_norm": 2.6415998935699463, + "learning_rate": 3.1709173962695094e-06, + "loss": 1.4441, + "step": 1269 + }, + { + "epoch": 1.3045711350796096, + "grad_norm": 2.5046234130859375, + "learning_rate": 3.1690140845070427e-06, + "loss": 1.4316, + "step": 1270 + }, + { + "epoch": 1.3045711350796096, + "eval_loss": 1.7736135721206665, + "eval_runtime": 37.9392, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 1270 + }, + { + "epoch": 1.305598356445814, + "grad_norm": 2.6234474182128906, + "learning_rate": 3.167110772744576e-06, + "loss": 1.4, + "step": 1271 + }, + { + "epoch": 1.3066255778120186, + "grad_norm": 2.488297700881958, + "learning_rate": 3.1652074609821095e-06, + "loss": 1.3908, + "step": 1272 + }, + { + "epoch": 1.307652799178223, + "grad_norm": 2.3759512901306152, + "learning_rate": 3.1633041492196424e-06, + "loss": 1.3908, + "step": 1273 + }, + { + "epoch": 1.3086800205444273, + "grad_norm": 2.487281084060669, + "learning_rate": 3.1614008374571758e-06, + "loss": 1.4217, + "step": 1274 + }, + { + "epoch": 1.3097072419106317, + "grad_norm": 2.5366780757904053, + "learning_rate": 3.159497525694709e-06, + "loss": 1.4229, + "step": 1275 + }, + { + "epoch": 1.310734463276836, + "grad_norm": 2.416141986846924, + "learning_rate": 3.1575942139322425e-06, + "loss": 1.3865, + "step": 1276 + }, + { + "epoch": 1.3117616846430407, + "grad_norm": 2.537964344024658, + "learning_rate": 3.155690902169776e-06, + "loss": 1.4141, + "step": 1277 + }, + { + "epoch": 1.312788906009245, + "grad_norm": 2.481692314147949, + "learning_rate": 3.1537875904073088e-06, + "loss": 1.383, + "step": 1278 + }, + { + "epoch": 1.3138161273754494, + "grad_norm": 2.4898593425750732, + "learning_rate": 3.151884278644842e-06, + "loss": 1.4078, + "step": 1279 + }, + { + "epoch": 1.3148433487416538, + "grad_norm": 2.4692323207855225, + "learning_rate": 3.1499809668823755e-06, + "loss": 1.3818, + "step": 1280 + }, + { + "epoch": 1.3148433487416538, + "eval_loss": 1.7660051584243774, + "eval_runtime": 37.8922, + "eval_samples_per_second": 26.391, + "eval_steps_per_second": 1.108, + "step": 1280 + }, + { + "epoch": 1.3158705701078581, + "grad_norm": 2.475862503051758, + "learning_rate": 3.148077655119909e-06, + "loss": 1.4384, + "step": 1281 + }, + { + "epoch": 1.3168977914740627, + "grad_norm": 2.2839648723602295, + "learning_rate": 3.1461743433574422e-06, + "loss": 1.3722, + "step": 1282 + }, + { + "epoch": 1.317925012840267, + "grad_norm": 2.3965585231781006, + "learning_rate": 3.144271031594975e-06, + "loss": 1.4262, + "step": 1283 + }, + { + "epoch": 1.3189522342064715, + "grad_norm": 2.4107236862182617, + "learning_rate": 3.1423677198325085e-06, + "loss": 1.4216, + "step": 1284 + }, + { + "epoch": 1.319979455572676, + "grad_norm": 2.438899040222168, + "learning_rate": 3.140464408070042e-06, + "loss": 1.383, + "step": 1285 + }, + { + "epoch": 1.3210066769388804, + "grad_norm": 2.4146060943603516, + "learning_rate": 3.1385610963075753e-06, + "loss": 1.3827, + "step": 1286 + }, + { + "epoch": 1.3220338983050848, + "grad_norm": 2.4193496704101562, + "learning_rate": 3.1366577845451086e-06, + "loss": 1.3707, + "step": 1287 + }, + { + "epoch": 1.3230611196712891, + "grad_norm": 2.5284323692321777, + "learning_rate": 3.134754472782642e-06, + "loss": 1.3623, + "step": 1288 + }, + { + "epoch": 1.3240883410374935, + "grad_norm": 2.421509265899658, + "learning_rate": 3.1328511610201758e-06, + "loss": 1.35, + "step": 1289 + }, + { + "epoch": 1.325115562403698, + "grad_norm": 2.3964431285858154, + "learning_rate": 3.1309478492577087e-06, + "loss": 1.3486, + "step": 1290 + }, + { + "epoch": 1.325115562403698, + "eval_loss": 1.7699168920516968, + "eval_runtime": 37.9003, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 1290 + }, + { + "epoch": 1.3261427837699025, + "grad_norm": 2.4371817111968994, + "learning_rate": 3.129044537495242e-06, + "loss": 1.3958, + "step": 1291 + }, + { + "epoch": 1.3271700051361068, + "grad_norm": 2.45233154296875, + "learning_rate": 3.1271412257327754e-06, + "loss": 1.4175, + "step": 1292 + }, + { + "epoch": 1.3281972265023112, + "grad_norm": 2.314384937286377, + "learning_rate": 3.125237913970309e-06, + "loss": 1.3701, + "step": 1293 + }, + { + "epoch": 1.3292244478685156, + "grad_norm": 2.5603694915771484, + "learning_rate": 3.123334602207842e-06, + "loss": 1.3489, + "step": 1294 + }, + { + "epoch": 1.3302516692347202, + "grad_norm": 2.5919623374938965, + "learning_rate": 3.121431290445375e-06, + "loss": 1.4472, + "step": 1295 + }, + { + "epoch": 1.3312788906009245, + "grad_norm": 2.6912758350372314, + "learning_rate": 3.1195279786829085e-06, + "loss": 1.4917, + "step": 1296 + }, + { + "epoch": 1.332306111967129, + "grad_norm": 2.439089298248291, + "learning_rate": 3.117624666920442e-06, + "loss": 1.3776, + "step": 1297 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 2.5207953453063965, + "learning_rate": 3.115721355157975e-06, + "loss": 1.3784, + "step": 1298 + }, + { + "epoch": 1.3343605546995376, + "grad_norm": 2.507791519165039, + "learning_rate": 3.1138180433955085e-06, + "loss": 1.3346, + "step": 1299 + }, + { + "epoch": 1.3353877760657422, + "grad_norm": 2.450531482696533, + "learning_rate": 3.1119147316330415e-06, + "loss": 1.4151, + "step": 1300 + }, + { + "epoch": 1.3353877760657422, + "eval_loss": 1.7761608362197876, + "eval_runtime": 37.9041, + "eval_samples_per_second": 26.382, + "eval_steps_per_second": 1.108, + "step": 1300 + }, + { + "epoch": 1.3364149974319466, + "grad_norm": 2.488255262374878, + "learning_rate": 3.110011419870575e-06, + "loss": 1.4475, + "step": 1301 + }, + { + "epoch": 1.337442218798151, + "grad_norm": 2.5059807300567627, + "learning_rate": 3.1081081081081082e-06, + "loss": 1.3812, + "step": 1302 + }, + { + "epoch": 1.3384694401643555, + "grad_norm": 2.5579488277435303, + "learning_rate": 3.1062047963456416e-06, + "loss": 1.4134, + "step": 1303 + }, + { + "epoch": 1.33949666153056, + "grad_norm": 2.6170148849487305, + "learning_rate": 3.104301484583175e-06, + "loss": 1.459, + "step": 1304 + }, + { + "epoch": 1.3405238828967643, + "grad_norm": 2.5572125911712646, + "learning_rate": 3.1023981728207083e-06, + "loss": 1.4354, + "step": 1305 + }, + { + "epoch": 1.3415511042629686, + "grad_norm": 2.396883487701416, + "learning_rate": 3.1004948610582412e-06, + "loss": 1.4016, + "step": 1306 + }, + { + "epoch": 1.342578325629173, + "grad_norm": 2.5869390964508057, + "learning_rate": 3.0985915492957746e-06, + "loss": 1.3945, + "step": 1307 + }, + { + "epoch": 1.3436055469953776, + "grad_norm": 2.4271085262298584, + "learning_rate": 3.096688237533308e-06, + "loss": 1.3506, + "step": 1308 + }, + { + "epoch": 1.344632768361582, + "grad_norm": 2.398407220840454, + "learning_rate": 3.0947849257708418e-06, + "loss": 1.4475, + "step": 1309 + }, + { + "epoch": 1.3456599897277863, + "grad_norm": 2.441934823989868, + "learning_rate": 3.092881614008375e-06, + "loss": 1.3972, + "step": 1310 + }, + { + "epoch": 1.3456599897277863, + "eval_loss": 1.774910569190979, + "eval_runtime": 37.6105, + "eval_samples_per_second": 26.588, + "eval_steps_per_second": 1.117, + "step": 1310 + }, + { + "epoch": 1.3466872110939907, + "grad_norm": 2.5015370845794678, + "learning_rate": 3.0909783022459085e-06, + "loss": 1.4418, + "step": 1311 + }, + { + "epoch": 1.347714432460195, + "grad_norm": 2.4616997241973877, + "learning_rate": 3.0890749904834414e-06, + "loss": 1.3833, + "step": 1312 + }, + { + "epoch": 1.3487416538263997, + "grad_norm": 2.5993902683258057, + "learning_rate": 3.0871716787209748e-06, + "loss": 1.4385, + "step": 1313 + }, + { + "epoch": 1.349768875192604, + "grad_norm": 2.518195867538452, + "learning_rate": 3.085268366958508e-06, + "loss": 1.4127, + "step": 1314 + }, + { + "epoch": 1.3507960965588084, + "grad_norm": 2.545189380645752, + "learning_rate": 3.0833650551960415e-06, + "loss": 1.42, + "step": 1315 + }, + { + "epoch": 1.3518233179250128, + "grad_norm": 2.461818218231201, + "learning_rate": 3.081461743433575e-06, + "loss": 1.3796, + "step": 1316 + }, + { + "epoch": 1.3528505392912171, + "grad_norm": 2.4081408977508545, + "learning_rate": 3.079558431671108e-06, + "loss": 1.3939, + "step": 1317 + }, + { + "epoch": 1.3538777606574217, + "grad_norm": 2.2885539531707764, + "learning_rate": 3.077655119908641e-06, + "loss": 1.3645, + "step": 1318 + }, + { + "epoch": 1.354904982023626, + "grad_norm": 2.453791618347168, + "learning_rate": 3.0757518081461745e-06, + "loss": 1.3906, + "step": 1319 + }, + { + "epoch": 1.3559322033898304, + "grad_norm": 2.4538352489471436, + "learning_rate": 3.073848496383708e-06, + "loss": 1.4127, + "step": 1320 + }, + { + "epoch": 1.3559322033898304, + "eval_loss": 1.7821755409240723, + "eval_runtime": 37.7585, + "eval_samples_per_second": 26.484, + "eval_steps_per_second": 1.112, + "step": 1320 + }, + { + "epoch": 1.356959424756035, + "grad_norm": 2.438425064086914, + "learning_rate": 3.0719451846212413e-06, + "loss": 1.3763, + "step": 1321 + }, + { + "epoch": 1.3579866461222394, + "grad_norm": 2.4398279190063477, + "learning_rate": 3.0700418728587746e-06, + "loss": 1.3963, + "step": 1322 + }, + { + "epoch": 1.3590138674884438, + "grad_norm": 2.6763861179351807, + "learning_rate": 3.0681385610963076e-06, + "loss": 1.3837, + "step": 1323 + }, + { + "epoch": 1.3600410888546481, + "grad_norm": 2.685056209564209, + "learning_rate": 3.066235249333841e-06, + "loss": 1.4158, + "step": 1324 + }, + { + "epoch": 1.3610683102208525, + "grad_norm": 2.6663355827331543, + "learning_rate": 3.0643319375713743e-06, + "loss": 1.4533, + "step": 1325 + }, + { + "epoch": 1.362095531587057, + "grad_norm": 2.4613428115844727, + "learning_rate": 3.0624286258089076e-06, + "loss": 1.3927, + "step": 1326 + }, + { + "epoch": 1.3631227529532615, + "grad_norm": 2.348334789276123, + "learning_rate": 3.060525314046441e-06, + "loss": 1.3782, + "step": 1327 + }, + { + "epoch": 1.3641499743194658, + "grad_norm": 2.5011332035064697, + "learning_rate": 3.058622002283974e-06, + "loss": 1.4175, + "step": 1328 + }, + { + "epoch": 1.3651771956856702, + "grad_norm": 2.4692256450653076, + "learning_rate": 3.0567186905215077e-06, + "loss": 1.4319, + "step": 1329 + }, + { + "epoch": 1.3662044170518746, + "grad_norm": 2.616990089416504, + "learning_rate": 3.054815378759041e-06, + "loss": 1.4264, + "step": 1330 + }, + { + "epoch": 1.3662044170518746, + "eval_loss": 1.781583309173584, + "eval_runtime": 37.9291, + "eval_samples_per_second": 26.365, + "eval_steps_per_second": 1.107, + "step": 1330 + }, + { + "epoch": 1.3672316384180792, + "grad_norm": 2.5416364669799805, + "learning_rate": 3.0529120669965745e-06, + "loss": 1.4249, + "step": 1331 + }, + { + "epoch": 1.3682588597842835, + "grad_norm": 2.5346689224243164, + "learning_rate": 3.051008755234108e-06, + "loss": 1.3647, + "step": 1332 + }, + { + "epoch": 1.3692860811504879, + "grad_norm": 2.585714340209961, + "learning_rate": 3.049105443471641e-06, + "loss": 1.3714, + "step": 1333 + }, + { + "epoch": 1.3703133025166925, + "grad_norm": 2.5958940982818604, + "learning_rate": 3.0472021317091745e-06, + "loss": 1.5041, + "step": 1334 + }, + { + "epoch": 1.3713405238828968, + "grad_norm": 2.4795925617218018, + "learning_rate": 3.0452988199467075e-06, + "loss": 1.4317, + "step": 1335 + }, + { + "epoch": 1.3723677452491012, + "grad_norm": 2.4608285427093506, + "learning_rate": 3.043395508184241e-06, + "loss": 1.4259, + "step": 1336 + }, + { + "epoch": 1.3733949666153056, + "grad_norm": 2.5970380306243896, + "learning_rate": 3.041492196421774e-06, + "loss": 1.3859, + "step": 1337 + }, + { + "epoch": 1.37442218798151, + "grad_norm": 2.517047166824341, + "learning_rate": 3.0395888846593076e-06, + "loss": 1.4086, + "step": 1338 + }, + { + "epoch": 1.3754494093477145, + "grad_norm": 2.4541358947753906, + "learning_rate": 3.037685572896841e-06, + "loss": 1.3959, + "step": 1339 + }, + { + "epoch": 1.376476630713919, + "grad_norm": 2.6143250465393066, + "learning_rate": 3.035782261134374e-06, + "loss": 1.4088, + "step": 1340 + }, + { + "epoch": 1.376476630713919, + "eval_loss": 1.7629026174545288, + "eval_runtime": 38.0156, + "eval_samples_per_second": 26.305, + "eval_steps_per_second": 1.105, + "step": 1340 + }, + { + "epoch": 1.3775038520801233, + "grad_norm": 2.763033151626587, + "learning_rate": 3.0338789493719072e-06, + "loss": 1.4104, + "step": 1341 + }, + { + "epoch": 1.3785310734463276, + "grad_norm": 2.523888349533081, + "learning_rate": 3.0319756376094406e-06, + "loss": 1.4073, + "step": 1342 + }, + { + "epoch": 1.379558294812532, + "grad_norm": 2.5393309593200684, + "learning_rate": 3.030072325846974e-06, + "loss": 1.4145, + "step": 1343 + }, + { + "epoch": 1.3805855161787366, + "grad_norm": 2.455733060836792, + "learning_rate": 3.0281690140845073e-06, + "loss": 1.3782, + "step": 1344 + }, + { + "epoch": 1.381612737544941, + "grad_norm": 2.600311517715454, + "learning_rate": 3.0262657023220403e-06, + "loss": 1.4158, + "step": 1345 + }, + { + "epoch": 1.3826399589111453, + "grad_norm": 2.453249931335449, + "learning_rate": 3.0243623905595736e-06, + "loss": 1.414, + "step": 1346 + }, + { + "epoch": 1.3836671802773497, + "grad_norm": 2.603515625, + "learning_rate": 3.022459078797107e-06, + "loss": 1.4102, + "step": 1347 + }, + { + "epoch": 1.384694401643554, + "grad_norm": 2.6515872478485107, + "learning_rate": 3.0205557670346404e-06, + "loss": 1.4063, + "step": 1348 + }, + { + "epoch": 1.3857216230097587, + "grad_norm": 2.412846565246582, + "learning_rate": 3.018652455272174e-06, + "loss": 1.3739, + "step": 1349 + }, + { + "epoch": 1.386748844375963, + "grad_norm": 2.748563051223755, + "learning_rate": 3.0167491435097075e-06, + "loss": 1.4909, + "step": 1350 + }, + { + "epoch": 1.386748844375963, + "eval_loss": 1.7653924226760864, + "eval_runtime": 37.9855, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 1350 + }, + { + "epoch": 1.3877760657421674, + "grad_norm": 2.7861363887786865, + "learning_rate": 3.014845831747241e-06, + "loss": 1.4028, + "step": 1351 + }, + { + "epoch": 1.388803287108372, + "grad_norm": 2.5898189544677734, + "learning_rate": 3.012942519984774e-06, + "loss": 1.3793, + "step": 1352 + }, + { + "epoch": 1.3898305084745763, + "grad_norm": 2.4240355491638184, + "learning_rate": 3.011039208222307e-06, + "loss": 1.3889, + "step": 1353 + }, + { + "epoch": 1.3908577298407807, + "grad_norm": 2.4492504596710205, + "learning_rate": 3.0091358964598405e-06, + "loss": 1.3464, + "step": 1354 + }, + { + "epoch": 1.391884951206985, + "grad_norm": 2.4425718784332275, + "learning_rate": 3.007232584697374e-06, + "loss": 1.4073, + "step": 1355 + }, + { + "epoch": 1.3929121725731894, + "grad_norm": 2.5057594776153564, + "learning_rate": 3.0053292729349073e-06, + "loss": 1.4053, + "step": 1356 + }, + { + "epoch": 1.393939393939394, + "grad_norm": 2.5386273860931396, + "learning_rate": 3.00342596117244e-06, + "loss": 1.3991, + "step": 1357 + }, + { + "epoch": 1.3949666153055984, + "grad_norm": 2.6289730072021484, + "learning_rate": 3.0015226494099736e-06, + "loss": 1.3971, + "step": 1358 + }, + { + "epoch": 1.3959938366718028, + "grad_norm": 2.5091774463653564, + "learning_rate": 2.999619337647507e-06, + "loss": 1.3815, + "step": 1359 + }, + { + "epoch": 1.3970210580380071, + "grad_norm": 2.567659616470337, + "learning_rate": 2.9977160258850403e-06, + "loss": 1.3786, + "step": 1360 + }, + { + "epoch": 1.3970210580380071, + "eval_loss": 1.7706127166748047, + "eval_runtime": 37.9897, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1360 + }, + { + "epoch": 1.3980482794042115, + "grad_norm": 2.5508553981781006, + "learning_rate": 2.9958127141225736e-06, + "loss": 1.3629, + "step": 1361 + }, + { + "epoch": 1.399075500770416, + "grad_norm": 2.624800205230713, + "learning_rate": 2.9939094023601066e-06, + "loss": 1.4355, + "step": 1362 + }, + { + "epoch": 1.4001027221366205, + "grad_norm": 2.432114839553833, + "learning_rate": 2.99200609059764e-06, + "loss": 1.3995, + "step": 1363 + }, + { + "epoch": 1.4011299435028248, + "grad_norm": 2.4633607864379883, + "learning_rate": 2.9901027788351733e-06, + "loss": 1.3499, + "step": 1364 + }, + { + "epoch": 1.4021571648690292, + "grad_norm": 2.557415246963501, + "learning_rate": 2.9881994670727067e-06, + "loss": 1.3614, + "step": 1365 + }, + { + "epoch": 1.4031843862352336, + "grad_norm": 2.411295175552368, + "learning_rate": 2.98629615531024e-06, + "loss": 1.3857, + "step": 1366 + }, + { + "epoch": 1.4042116076014381, + "grad_norm": 2.429797410964966, + "learning_rate": 2.984392843547773e-06, + "loss": 1.3701, + "step": 1367 + }, + { + "epoch": 1.4052388289676425, + "grad_norm": 2.6080522537231445, + "learning_rate": 2.9824895317853063e-06, + "loss": 1.3996, + "step": 1368 + }, + { + "epoch": 1.4062660503338469, + "grad_norm": 2.5128798484802246, + "learning_rate": 2.98058622002284e-06, + "loss": 1.3973, + "step": 1369 + }, + { + "epoch": 1.4072932717000515, + "grad_norm": 2.5906126499176025, + "learning_rate": 2.9786829082603735e-06, + "loss": 1.4412, + "step": 1370 + }, + { + "epoch": 1.4072932717000515, + "eval_loss": 1.7700779438018799, + "eval_runtime": 38.017, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 1370 + }, + { + "epoch": 1.4083204930662558, + "grad_norm": 2.4890248775482178, + "learning_rate": 2.976779596497907e-06, + "loss": 1.4251, + "step": 1371 + }, + { + "epoch": 1.4093477144324602, + "grad_norm": 2.3803439140319824, + "learning_rate": 2.97487628473544e-06, + "loss": 1.4363, + "step": 1372 + }, + { + "epoch": 1.4103749357986646, + "grad_norm": 2.461958646774292, + "learning_rate": 2.9729729729729736e-06, + "loss": 1.3948, + "step": 1373 + }, + { + "epoch": 1.411402157164869, + "grad_norm": 2.419793128967285, + "learning_rate": 2.9710696612105065e-06, + "loss": 1.4367, + "step": 1374 + }, + { + "epoch": 1.4124293785310735, + "grad_norm": 2.5748846530914307, + "learning_rate": 2.96916634944804e-06, + "loss": 1.3873, + "step": 1375 + }, + { + "epoch": 1.413456599897278, + "grad_norm": 2.564558506011963, + "learning_rate": 2.9672630376855732e-06, + "loss": 1.4395, + "step": 1376 + }, + { + "epoch": 1.4144838212634823, + "grad_norm": 2.596085548400879, + "learning_rate": 2.9653597259231066e-06, + "loss": 1.4091, + "step": 1377 + }, + { + "epoch": 1.4155110426296866, + "grad_norm": 2.49029278755188, + "learning_rate": 2.96345641416064e-06, + "loss": 1.3639, + "step": 1378 + }, + { + "epoch": 1.416538263995891, + "grad_norm": 2.506887912750244, + "learning_rate": 2.961553102398173e-06, + "loss": 1.3065, + "step": 1379 + }, + { + "epoch": 1.4175654853620956, + "grad_norm": 2.517609119415283, + "learning_rate": 2.9596497906357063e-06, + "loss": 1.4365, + "step": 1380 + }, + { + "epoch": 1.4175654853620956, + "eval_loss": 1.7673696279525757, + "eval_runtime": 37.9941, + "eval_samples_per_second": 26.32, + "eval_steps_per_second": 1.105, + "step": 1380 + }, + { + "epoch": 1.4185927067283, + "grad_norm": 2.6870028972625732, + "learning_rate": 2.9577464788732396e-06, + "loss": 1.4386, + "step": 1381 + }, + { + "epoch": 1.4196199280945043, + "grad_norm": 2.3195550441741943, + "learning_rate": 2.955843167110773e-06, + "loss": 1.3771, + "step": 1382 + }, + { + "epoch": 1.420647149460709, + "grad_norm": 2.4306302070617676, + "learning_rate": 2.9539398553483064e-06, + "loss": 1.3963, + "step": 1383 + }, + { + "epoch": 1.4216743708269133, + "grad_norm": 2.4227781295776367, + "learning_rate": 2.9520365435858393e-06, + "loss": 1.394, + "step": 1384 + }, + { + "epoch": 1.4227015921931176, + "grad_norm": 2.521425485610962, + "learning_rate": 2.9501332318233727e-06, + "loss": 1.4664, + "step": 1385 + }, + { + "epoch": 1.423728813559322, + "grad_norm": 2.533363103866577, + "learning_rate": 2.948229920060906e-06, + "loss": 1.4072, + "step": 1386 + }, + { + "epoch": 1.4247560349255264, + "grad_norm": 2.378007173538208, + "learning_rate": 2.9463266082984394e-06, + "loss": 1.38, + "step": 1387 + }, + { + "epoch": 1.425783256291731, + "grad_norm": 2.476451873779297, + "learning_rate": 2.9444232965359727e-06, + "loss": 1.3864, + "step": 1388 + }, + { + "epoch": 1.4268104776579353, + "grad_norm": 2.428365707397461, + "learning_rate": 2.9425199847735057e-06, + "loss": 1.4081, + "step": 1389 + }, + { + "epoch": 1.4278376990241397, + "grad_norm": 2.4382991790771484, + "learning_rate": 2.94061667301104e-06, + "loss": 1.3806, + "step": 1390 + }, + { + "epoch": 1.4278376990241397, + "eval_loss": 1.7905043363571167, + "eval_runtime": 38.0264, + "eval_samples_per_second": 26.298, + "eval_steps_per_second": 1.104, + "step": 1390 + }, + { + "epoch": 1.428864920390344, + "grad_norm": 2.414433240890503, + "learning_rate": 2.938713361248573e-06, + "loss": 1.3113, + "step": 1391 + }, + { + "epoch": 1.4298921417565484, + "grad_norm": 2.450570821762085, + "learning_rate": 2.936810049486106e-06, + "loss": 1.4023, + "step": 1392 + }, + { + "epoch": 1.430919363122753, + "grad_norm": 2.4372148513793945, + "learning_rate": 2.9349067377236396e-06, + "loss": 1.4057, + "step": 1393 + }, + { + "epoch": 1.4319465844889574, + "grad_norm": 2.4278311729431152, + "learning_rate": 2.933003425961173e-06, + "loss": 1.3574, + "step": 1394 + }, + { + "epoch": 1.4329738058551618, + "grad_norm": 2.448517322540283, + "learning_rate": 2.9311001141987063e-06, + "loss": 1.3722, + "step": 1395 + }, + { + "epoch": 1.4340010272213661, + "grad_norm": 2.354310989379883, + "learning_rate": 2.9291968024362392e-06, + "loss": 1.3682, + "step": 1396 + }, + { + "epoch": 1.4350282485875705, + "grad_norm": 2.3318262100219727, + "learning_rate": 2.9272934906737726e-06, + "loss": 1.3604, + "step": 1397 + }, + { + "epoch": 1.436055469953775, + "grad_norm": 2.377561092376709, + "learning_rate": 2.925390178911306e-06, + "loss": 1.372, + "step": 1398 + }, + { + "epoch": 1.4370826913199795, + "grad_norm": 2.5543324947357178, + "learning_rate": 2.9234868671488393e-06, + "loss": 1.3846, + "step": 1399 + }, + { + "epoch": 1.4381099126861838, + "grad_norm": 2.6063973903656006, + "learning_rate": 2.9215835553863727e-06, + "loss": 1.3989, + "step": 1400 + }, + { + "epoch": 1.4381099126861838, + "eval_loss": 1.7730437517166138, + "eval_runtime": 38.0407, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 1400 + }, + { + "epoch": 1.4391371340523884, + "grad_norm": 2.4241244792938232, + "learning_rate": 2.9196802436239056e-06, + "loss": 1.3945, + "step": 1401 + }, + { + "epoch": 1.4401643554185928, + "grad_norm": 2.306528329849243, + "learning_rate": 2.917776931861439e-06, + "loss": 1.3686, + "step": 1402 + }, + { + "epoch": 1.4411915767847971, + "grad_norm": 2.4980850219726562, + "learning_rate": 2.9158736200989723e-06, + "loss": 1.3714, + "step": 1403 + }, + { + "epoch": 1.4422187981510015, + "grad_norm": 2.385195732116699, + "learning_rate": 2.9139703083365057e-06, + "loss": 1.3706, + "step": 1404 + }, + { + "epoch": 1.4432460195172059, + "grad_norm": 2.4087095260620117, + "learning_rate": 2.912066996574039e-06, + "loss": 1.3967, + "step": 1405 + }, + { + "epoch": 1.4442732408834105, + "grad_norm": 2.470900535583496, + "learning_rate": 2.910163684811572e-06, + "loss": 1.3351, + "step": 1406 + }, + { + "epoch": 1.4453004622496148, + "grad_norm": 2.8330743312835693, + "learning_rate": 2.9082603730491054e-06, + "loss": 1.4413, + "step": 1407 + }, + { + "epoch": 1.4463276836158192, + "grad_norm": 2.5553359985351562, + "learning_rate": 2.9063570612866387e-06, + "loss": 1.4207, + "step": 1408 + }, + { + "epoch": 1.4473549049820236, + "grad_norm": 2.509373903274536, + "learning_rate": 2.904453749524172e-06, + "loss": 1.3387, + "step": 1409 + }, + { + "epoch": 1.448382126348228, + "grad_norm": 2.5485401153564453, + "learning_rate": 2.902550437761706e-06, + "loss": 1.3765, + "step": 1410 + }, + { + "epoch": 1.448382126348228, + "eval_loss": 1.7703789472579956, + "eval_runtime": 37.7078, + "eval_samples_per_second": 26.52, + "eval_steps_per_second": 1.114, + "step": 1410 + }, + { + "epoch": 1.4494093477144325, + "grad_norm": 2.502002716064453, + "learning_rate": 2.9006471259992392e-06, + "loss": 1.3842, + "step": 1411 + }, + { + "epoch": 1.450436569080637, + "grad_norm": 2.4621520042419434, + "learning_rate": 2.8987438142367726e-06, + "loss": 1.3809, + "step": 1412 + }, + { + "epoch": 1.4514637904468413, + "grad_norm": 2.4255263805389404, + "learning_rate": 2.8968405024743055e-06, + "loss": 1.4411, + "step": 1413 + }, + { + "epoch": 1.4524910118130456, + "grad_norm": 2.2450802326202393, + "learning_rate": 2.894937190711839e-06, + "loss": 1.3389, + "step": 1414 + }, + { + "epoch": 1.45351823317925, + "grad_norm": 2.502903938293457, + "learning_rate": 2.8930338789493723e-06, + "loss": 1.4066, + "step": 1415 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 2.3685200214385986, + "learning_rate": 2.8911305671869056e-06, + "loss": 1.3995, + "step": 1416 + }, + { + "epoch": 1.455572675911659, + "grad_norm": 2.355095624923706, + "learning_rate": 2.889227255424439e-06, + "loss": 1.3352, + "step": 1417 + }, + { + "epoch": 1.4565998972778633, + "grad_norm": 2.363035202026367, + "learning_rate": 2.887323943661972e-06, + "loss": 1.3792, + "step": 1418 + }, + { + "epoch": 1.457627118644068, + "grad_norm": 2.451078176498413, + "learning_rate": 2.8854206318995053e-06, + "loss": 1.358, + "step": 1419 + }, + { + "epoch": 1.4586543400102723, + "grad_norm": 2.554501533508301, + "learning_rate": 2.8835173201370386e-06, + "loss": 1.3811, + "step": 1420 + }, + { + "epoch": 1.4586543400102723, + "eval_loss": 1.773111343383789, + "eval_runtime": 37.7693, + "eval_samples_per_second": 26.477, + "eval_steps_per_second": 1.112, + "step": 1420 + }, + { + "epoch": 1.4596815613764766, + "grad_norm": 2.556537389755249, + "learning_rate": 2.881614008374572e-06, + "loss": 1.4397, + "step": 1421 + }, + { + "epoch": 1.460708782742681, + "grad_norm": 2.5945091247558594, + "learning_rate": 2.8797106966121054e-06, + "loss": 1.3976, + "step": 1422 + }, + { + "epoch": 1.4617360041088854, + "grad_norm": 2.5158495903015137, + "learning_rate": 2.8778073848496383e-06, + "loss": 1.381, + "step": 1423 + }, + { + "epoch": 1.46276322547509, + "grad_norm": 2.359591007232666, + "learning_rate": 2.8759040730871717e-06, + "loss": 1.4156, + "step": 1424 + }, + { + "epoch": 1.4637904468412943, + "grad_norm": 2.3388748168945312, + "learning_rate": 2.874000761324705e-06, + "loss": 1.3808, + "step": 1425 + }, + { + "epoch": 1.4648176682074987, + "grad_norm": 2.438465118408203, + "learning_rate": 2.8720974495622384e-06, + "loss": 1.4158, + "step": 1426 + }, + { + "epoch": 1.465844889573703, + "grad_norm": 2.5624589920043945, + "learning_rate": 2.8701941377997718e-06, + "loss": 1.3917, + "step": 1427 + }, + { + "epoch": 1.4668721109399074, + "grad_norm": 2.467921495437622, + "learning_rate": 2.8682908260373047e-06, + "loss": 1.4316, + "step": 1428 + }, + { + "epoch": 1.467899332306112, + "grad_norm": 2.6034083366394043, + "learning_rate": 2.866387514274838e-06, + "loss": 1.4058, + "step": 1429 + }, + { + "epoch": 1.4689265536723164, + "grad_norm": 2.560844898223877, + "learning_rate": 2.864484202512372e-06, + "loss": 1.4114, + "step": 1430 + }, + { + "epoch": 1.4689265536723164, + "eval_loss": 1.7799324989318848, + "eval_runtime": 37.8814, + "eval_samples_per_second": 26.398, + "eval_steps_per_second": 1.109, + "step": 1430 + }, + { + "epoch": 1.4699537750385208, + "grad_norm": 2.469799280166626, + "learning_rate": 2.8625808907499052e-06, + "loss": 1.3767, + "step": 1431 + }, + { + "epoch": 1.4709809964047253, + "grad_norm": 2.408665895462036, + "learning_rate": 2.8606775789874386e-06, + "loss": 1.4064, + "step": 1432 + }, + { + "epoch": 1.4720082177709297, + "grad_norm": 2.497950553894043, + "learning_rate": 2.858774267224972e-06, + "loss": 1.3747, + "step": 1433 + }, + { + "epoch": 1.473035439137134, + "grad_norm": 2.6274573802948, + "learning_rate": 2.8568709554625053e-06, + "loss": 1.3994, + "step": 1434 + }, + { + "epoch": 1.4740626605033385, + "grad_norm": 2.545614719390869, + "learning_rate": 2.8549676437000382e-06, + "loss": 1.4354, + "step": 1435 + }, + { + "epoch": 1.4750898818695428, + "grad_norm": 2.796013355255127, + "learning_rate": 2.8530643319375716e-06, + "loss": 1.4081, + "step": 1436 + }, + { + "epoch": 1.4761171032357474, + "grad_norm": 2.68571400642395, + "learning_rate": 2.851161020175105e-06, + "loss": 1.3833, + "step": 1437 + }, + { + "epoch": 1.4771443246019518, + "grad_norm": 2.455244541168213, + "learning_rate": 2.8492577084126383e-06, + "loss": 1.3587, + "step": 1438 + }, + { + "epoch": 1.4781715459681561, + "grad_norm": 2.552747964859009, + "learning_rate": 2.8473543966501717e-06, + "loss": 1.4004, + "step": 1439 + }, + { + "epoch": 1.4791987673343605, + "grad_norm": 2.9428722858428955, + "learning_rate": 2.8454510848877046e-06, + "loss": 1.437, + "step": 1440 + }, + { + "epoch": 1.4791987673343605, + "eval_loss": 1.7853318452835083, + "eval_runtime": 37.8848, + "eval_samples_per_second": 26.396, + "eval_steps_per_second": 1.109, + "step": 1440 + }, + { + "epoch": 1.4802259887005649, + "grad_norm": 2.686199188232422, + "learning_rate": 2.843547773125238e-06, + "loss": 1.3901, + "step": 1441 + }, + { + "epoch": 1.4812532100667695, + "grad_norm": 2.47476863861084, + "learning_rate": 2.8416444613627714e-06, + "loss": 1.3887, + "step": 1442 + }, + { + "epoch": 1.4822804314329738, + "grad_norm": 2.425445795059204, + "learning_rate": 2.8397411496003047e-06, + "loss": 1.3502, + "step": 1443 + }, + { + "epoch": 1.4833076527991782, + "grad_norm": 2.529341459274292, + "learning_rate": 2.837837837837838e-06, + "loss": 1.3973, + "step": 1444 + }, + { + "epoch": 1.4843348741653826, + "grad_norm": 2.3861210346221924, + "learning_rate": 2.835934526075371e-06, + "loss": 1.4072, + "step": 1445 + }, + { + "epoch": 1.485362095531587, + "grad_norm": 2.359403371810913, + "learning_rate": 2.8340312143129044e-06, + "loss": 1.3553, + "step": 1446 + }, + { + "epoch": 1.4863893168977915, + "grad_norm": 2.417792558670044, + "learning_rate": 2.8321279025504377e-06, + "loss": 1.3539, + "step": 1447 + }, + { + "epoch": 1.487416538263996, + "grad_norm": 2.7177648544311523, + "learning_rate": 2.830224590787971e-06, + "loss": 1.4575, + "step": 1448 + }, + { + "epoch": 1.4884437596302003, + "grad_norm": 2.514181613922119, + "learning_rate": 2.8283212790255045e-06, + "loss": 1.388, + "step": 1449 + }, + { + "epoch": 1.4894709809964048, + "grad_norm": 2.4293761253356934, + "learning_rate": 2.8264179672630383e-06, + "loss": 1.4143, + "step": 1450 + }, + { + "epoch": 1.4894709809964048, + "eval_loss": 1.781496286392212, + "eval_runtime": 37.8942, + "eval_samples_per_second": 26.389, + "eval_steps_per_second": 1.108, + "step": 1450 + }, + { + "epoch": 1.4904982023626092, + "grad_norm": 2.495591878890991, + "learning_rate": 2.8245146555005716e-06, + "loss": 1.3809, + "step": 1451 + }, + { + "epoch": 1.4915254237288136, + "grad_norm": 2.520158290863037, + "learning_rate": 2.8226113437381046e-06, + "loss": 1.3297, + "step": 1452 + }, + { + "epoch": 1.492552645095018, + "grad_norm": 2.403061866760254, + "learning_rate": 2.820708031975638e-06, + "loss": 1.3985, + "step": 1453 + }, + { + "epoch": 1.4935798664612223, + "grad_norm": 2.606884479522705, + "learning_rate": 2.8188047202131713e-06, + "loss": 1.4179, + "step": 1454 + }, + { + "epoch": 1.494607087827427, + "grad_norm": 2.4543404579162598, + "learning_rate": 2.8169014084507046e-06, + "loss": 1.4325, + "step": 1455 + }, + { + "epoch": 1.4956343091936313, + "grad_norm": 2.3947877883911133, + "learning_rate": 2.814998096688238e-06, + "loss": 1.3876, + "step": 1456 + }, + { + "epoch": 1.4966615305598356, + "grad_norm": 2.4632408618927, + "learning_rate": 2.813094784925771e-06, + "loss": 1.3392, + "step": 1457 + }, + { + "epoch": 1.49768875192604, + "grad_norm": 2.507312059402466, + "learning_rate": 2.8111914731633043e-06, + "loss": 1.4351, + "step": 1458 + }, + { + "epoch": 1.4987159732922444, + "grad_norm": 2.6338255405426025, + "learning_rate": 2.8092881614008377e-06, + "loss": 1.389, + "step": 1459 + }, + { + "epoch": 1.499743194658449, + "grad_norm": 2.3906662464141846, + "learning_rate": 2.807384849638371e-06, + "loss": 1.3655, + "step": 1460 + }, + { + "epoch": 1.499743194658449, + "eval_loss": 1.7716007232666016, + "eval_runtime": 37.8931, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 1460 + }, + { + "epoch": 1.5007704160246533, + "grad_norm": 2.4841926097869873, + "learning_rate": 2.8054815378759044e-06, + "loss": 1.4046, + "step": 1461 + }, + { + "epoch": 1.5017976373908577, + "grad_norm": 2.4934608936309814, + "learning_rate": 2.8035782261134378e-06, + "loss": 1.3781, + "step": 1462 + }, + { + "epoch": 1.5028248587570623, + "grad_norm": 2.538504123687744, + "learning_rate": 2.8016749143509707e-06, + "loss": 1.383, + "step": 1463 + }, + { + "epoch": 1.5038520801232664, + "grad_norm": 2.461595058441162, + "learning_rate": 2.799771602588504e-06, + "loss": 1.3892, + "step": 1464 + }, + { + "epoch": 1.504879301489471, + "grad_norm": 2.3686397075653076, + "learning_rate": 2.7978682908260374e-06, + "loss": 1.4037, + "step": 1465 + }, + { + "epoch": 1.5059065228556754, + "grad_norm": 2.47755765914917, + "learning_rate": 2.795964979063571e-06, + "loss": 1.4173, + "step": 1466 + }, + { + "epoch": 1.5069337442218798, + "grad_norm": 2.504168748855591, + "learning_rate": 2.794061667301104e-06, + "loss": 1.3955, + "step": 1467 + }, + { + "epoch": 1.5079609655880843, + "grad_norm": 2.4311411380767822, + "learning_rate": 2.792158355538637e-06, + "loss": 1.4079, + "step": 1468 + }, + { + "epoch": 1.5089881869542885, + "grad_norm": 2.523200273513794, + "learning_rate": 2.7902550437761705e-06, + "loss": 1.4019, + "step": 1469 + }, + { + "epoch": 1.510015408320493, + "grad_norm": 2.3743972778320312, + "learning_rate": 2.7883517320137042e-06, + "loss": 1.3677, + "step": 1470 + }, + { + "epoch": 1.510015408320493, + "eval_loss": 1.7622252702713013, + "eval_runtime": 37.8933, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 1470 + }, + { + "epoch": 1.5110426296866974, + "grad_norm": 2.4073667526245117, + "learning_rate": 2.7864484202512376e-06, + "loss": 1.3623, + "step": 1471 + }, + { + "epoch": 1.5120698510529018, + "grad_norm": 2.5012941360473633, + "learning_rate": 2.784545108488771e-06, + "loss": 1.382, + "step": 1472 + }, + { + "epoch": 1.5130970724191064, + "grad_norm": 2.425480365753174, + "learning_rate": 2.7826417967263043e-06, + "loss": 1.3896, + "step": 1473 + }, + { + "epoch": 1.5141242937853108, + "grad_norm": 2.4600207805633545, + "learning_rate": 2.7807384849638373e-06, + "loss": 1.4364, + "step": 1474 + }, + { + "epoch": 1.5151515151515151, + "grad_norm": 2.5489892959594727, + "learning_rate": 2.7788351732013706e-06, + "loss": 1.3832, + "step": 1475 + }, + { + "epoch": 1.5161787365177197, + "grad_norm": 2.4228315353393555, + "learning_rate": 2.776931861438904e-06, + "loss": 1.3755, + "step": 1476 + }, + { + "epoch": 1.5172059578839239, + "grad_norm": 2.519340991973877, + "learning_rate": 2.7750285496764374e-06, + "loss": 1.3915, + "step": 1477 + }, + { + "epoch": 1.5182331792501285, + "grad_norm": 2.3544020652770996, + "learning_rate": 2.7731252379139707e-06, + "loss": 1.394, + "step": 1478 + }, + { + "epoch": 1.5192604006163328, + "grad_norm": 2.4037892818450928, + "learning_rate": 2.771221926151504e-06, + "loss": 1.3396, + "step": 1479 + }, + { + "epoch": 1.5202876219825372, + "grad_norm": 2.409222364425659, + "learning_rate": 2.769318614389037e-06, + "loss": 1.4052, + "step": 1480 + }, + { + "epoch": 1.5202876219825372, + "eval_loss": 1.7622238397598267, + "eval_runtime": 37.9064, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 1480 + }, + { + "epoch": 1.5213148433487418, + "grad_norm": 2.3649749755859375, + "learning_rate": 2.7674153026265704e-06, + "loss": 1.401, + "step": 1481 + }, + { + "epoch": 1.522342064714946, + "grad_norm": 2.4484262466430664, + "learning_rate": 2.7655119908641037e-06, + "loss": 1.4026, + "step": 1482 + }, + { + "epoch": 1.5233692860811505, + "grad_norm": 2.492107629776001, + "learning_rate": 2.763608679101637e-06, + "loss": 1.3985, + "step": 1483 + }, + { + "epoch": 1.5243965074473549, + "grad_norm": 2.5192503929138184, + "learning_rate": 2.7617053673391705e-06, + "loss": 1.4282, + "step": 1484 + }, + { + "epoch": 1.5254237288135593, + "grad_norm": 2.471250534057617, + "learning_rate": 2.7598020555767034e-06, + "loss": 1.4217, + "step": 1485 + }, + { + "epoch": 1.5264509501797638, + "grad_norm": 2.2985496520996094, + "learning_rate": 2.7578987438142368e-06, + "loss": 1.3599, + "step": 1486 + }, + { + "epoch": 1.5274781715459682, + "grad_norm": 2.492047071456909, + "learning_rate": 2.75599543205177e-06, + "loss": 1.4036, + "step": 1487 + }, + { + "epoch": 1.5285053929121726, + "grad_norm": 2.5131945610046387, + "learning_rate": 2.7540921202893035e-06, + "loss": 1.4206, + "step": 1488 + }, + { + "epoch": 1.529532614278377, + "grad_norm": 2.619874954223633, + "learning_rate": 2.752188808526837e-06, + "loss": 1.4055, + "step": 1489 + }, + { + "epoch": 1.5305598356445813, + "grad_norm": 2.4526877403259277, + "learning_rate": 2.75028549676437e-06, + "loss": 1.3955, + "step": 1490 + }, + { + "epoch": 1.5305598356445813, + "eval_loss": 1.758982539176941, + "eval_runtime": 37.8756, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1490 + }, + { + "epoch": 1.531587057010786, + "grad_norm": 2.4925529956817627, + "learning_rate": 2.7483821850019036e-06, + "loss": 1.3899, + "step": 1491 + }, + { + "epoch": 1.5326142783769903, + "grad_norm": 2.3789961338043213, + "learning_rate": 2.746478873239437e-06, + "loss": 1.377, + "step": 1492 + }, + { + "epoch": 1.5336414997431946, + "grad_norm": 2.4063172340393066, + "learning_rate": 2.7445755614769703e-06, + "loss": 1.384, + "step": 1493 + }, + { + "epoch": 1.5346687211093992, + "grad_norm": 2.3189985752105713, + "learning_rate": 2.7426722497145037e-06, + "loss": 1.3635, + "step": 1494 + }, + { + "epoch": 1.5356959424756034, + "grad_norm": 2.4792535305023193, + "learning_rate": 2.740768937952037e-06, + "loss": 1.3676, + "step": 1495 + }, + { + "epoch": 1.536723163841808, + "grad_norm": 2.342355489730835, + "learning_rate": 2.7388656261895704e-06, + "loss": 1.3812, + "step": 1496 + }, + { + "epoch": 1.5377503852080123, + "grad_norm": 2.556312084197998, + "learning_rate": 2.7369623144271033e-06, + "loss": 1.4529, + "step": 1497 + }, + { + "epoch": 1.5387776065742167, + "grad_norm": 2.555600643157959, + "learning_rate": 2.7350590026646367e-06, + "loss": 1.3819, + "step": 1498 + }, + { + "epoch": 1.5398048279404213, + "grad_norm": 2.4179797172546387, + "learning_rate": 2.73315569090217e-06, + "loss": 1.3383, + "step": 1499 + }, + { + "epoch": 1.5408320493066254, + "grad_norm": 2.5075631141662598, + "learning_rate": 2.7312523791397034e-06, + "loss": 1.4217, + "step": 1500 + }, + { + "epoch": 1.5408320493066254, + "eval_loss": 1.769559621810913, + "eval_runtime": 37.8701, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 1500 + }, + { + "epoch": 1.54185927067283, + "grad_norm": 2.5675160884857178, + "learning_rate": 2.7293490673772368e-06, + "loss": 1.4437, + "step": 1501 + }, + { + "epoch": 1.5428864920390344, + "grad_norm": 2.4768013954162598, + "learning_rate": 2.7274457556147697e-06, + "loss": 1.4022, + "step": 1502 + }, + { + "epoch": 1.5439137134052388, + "grad_norm": 2.4523043632507324, + "learning_rate": 2.725542443852303e-06, + "loss": 1.385, + "step": 1503 + }, + { + "epoch": 1.5449409347714433, + "grad_norm": 2.504263401031494, + "learning_rate": 2.7236391320898365e-06, + "loss": 1.398, + "step": 1504 + }, + { + "epoch": 1.5459681561376477, + "grad_norm": 2.6203269958496094, + "learning_rate": 2.72173582032737e-06, + "loss": 1.3651, + "step": 1505 + }, + { + "epoch": 1.546995377503852, + "grad_norm": 2.3805699348449707, + "learning_rate": 2.719832508564903e-06, + "loss": 1.3861, + "step": 1506 + }, + { + "epoch": 1.5480225988700564, + "grad_norm": 2.5000486373901367, + "learning_rate": 2.717929196802436e-06, + "loss": 1.3931, + "step": 1507 + }, + { + "epoch": 1.5490498202362608, + "grad_norm": 2.5567827224731445, + "learning_rate": 2.7160258850399695e-06, + "loss": 1.3608, + "step": 1508 + }, + { + "epoch": 1.5500770416024654, + "grad_norm": 2.531614303588867, + "learning_rate": 2.714122573277503e-06, + "loss": 1.3863, + "step": 1509 + }, + { + "epoch": 1.5511042629686698, + "grad_norm": 2.419722318649292, + "learning_rate": 2.712219261515036e-06, + "loss": 1.3963, + "step": 1510 + }, + { + "epoch": 1.5511042629686698, + "eval_loss": 1.7669111490249634, + "eval_runtime": 37.6622, + "eval_samples_per_second": 26.552, + "eval_steps_per_second": 1.115, + "step": 1510 + }, + { + "epoch": 1.5521314843348741, + "grad_norm": 2.4597935676574707, + "learning_rate": 2.71031594975257e-06, + "loss": 1.4044, + "step": 1511 + }, + { + "epoch": 1.5531587057010787, + "grad_norm": 2.386899471282959, + "learning_rate": 2.7084126379901034e-06, + "loss": 1.4092, + "step": 1512 + }, + { + "epoch": 1.5541859270672829, + "grad_norm": 2.4398958683013916, + "learning_rate": 2.7065093262276367e-06, + "loss": 1.4185, + "step": 1513 + }, + { + "epoch": 1.5552131484334875, + "grad_norm": 2.319274663925171, + "learning_rate": 2.7046060144651697e-06, + "loss": 1.4316, + "step": 1514 + }, + { + "epoch": 1.5562403697996918, + "grad_norm": 2.4519565105438232, + "learning_rate": 2.702702702702703e-06, + "loss": 1.4387, + "step": 1515 + }, + { + "epoch": 1.5572675911658962, + "grad_norm": 2.2890849113464355, + "learning_rate": 2.7007993909402364e-06, + "loss": 1.3167, + "step": 1516 + }, + { + "epoch": 1.5582948125321008, + "grad_norm": 2.6062512397766113, + "learning_rate": 2.6988960791777697e-06, + "loss": 1.4332, + "step": 1517 + }, + { + "epoch": 1.559322033898305, + "grad_norm": 2.508957862854004, + "learning_rate": 2.696992767415303e-06, + "loss": 1.3995, + "step": 1518 + }, + { + "epoch": 1.5603492552645095, + "grad_norm": 2.4134879112243652, + "learning_rate": 2.695089455652836e-06, + "loss": 1.3793, + "step": 1519 + }, + { + "epoch": 1.5613764766307139, + "grad_norm": 2.498729705810547, + "learning_rate": 2.6931861438903694e-06, + "loss": 1.3564, + "step": 1520 + }, + { + "epoch": 1.5613764766307139, + "eval_loss": 1.76124107837677, + "eval_runtime": 37.831, + "eval_samples_per_second": 26.433, + "eval_steps_per_second": 1.11, + "step": 1520 + }, + { + "epoch": 1.5624036979969183, + "grad_norm": 2.455510139465332, + "learning_rate": 2.6912828321279028e-06, + "loss": 1.4109, + "step": 1521 + }, + { + "epoch": 1.5634309193631228, + "grad_norm": 2.422098159790039, + "learning_rate": 2.689379520365436e-06, + "loss": 1.3811, + "step": 1522 + }, + { + "epoch": 1.5644581407293272, + "grad_norm": 2.380890369415283, + "learning_rate": 2.6874762086029695e-06, + "loss": 1.4215, + "step": 1523 + }, + { + "epoch": 1.5654853620955316, + "grad_norm": 2.39363169670105, + "learning_rate": 2.6855728968405024e-06, + "loss": 1.3708, + "step": 1524 + }, + { + "epoch": 1.5665125834617362, + "grad_norm": 2.3624465465545654, + "learning_rate": 2.683669585078036e-06, + "loss": 1.4084, + "step": 1525 + }, + { + "epoch": 1.5675398048279403, + "grad_norm": 2.3589534759521484, + "learning_rate": 2.681766273315569e-06, + "loss": 1.3465, + "step": 1526 + }, + { + "epoch": 1.568567026194145, + "grad_norm": 2.4358372688293457, + "learning_rate": 2.6798629615531025e-06, + "loss": 1.3915, + "step": 1527 + }, + { + "epoch": 1.5695942475603493, + "grad_norm": 2.4162042140960693, + "learning_rate": 2.677959649790636e-06, + "loss": 1.3842, + "step": 1528 + }, + { + "epoch": 1.5706214689265536, + "grad_norm": 2.568127155303955, + "learning_rate": 2.676056338028169e-06, + "loss": 1.4197, + "step": 1529 + }, + { + "epoch": 1.5716486902927582, + "grad_norm": 2.5098628997802734, + "learning_rate": 2.674153026265702e-06, + "loss": 1.4132, + "step": 1530 + }, + { + "epoch": 1.5716486902927582, + "eval_loss": 1.7691000699996948, + "eval_runtime": 37.9709, + "eval_samples_per_second": 26.336, + "eval_steps_per_second": 1.106, + "step": 1530 + }, + { + "epoch": 1.5726759116589624, + "grad_norm": 2.5882396697998047, + "learning_rate": 2.672249714503236e-06, + "loss": 1.4223, + "step": 1531 + }, + { + "epoch": 1.573703133025167, + "grad_norm": 2.5202438831329346, + "learning_rate": 2.6703464027407693e-06, + "loss": 1.3804, + "step": 1532 + }, + { + "epoch": 1.5747303543913713, + "grad_norm": 2.4582319259643555, + "learning_rate": 2.6684430909783027e-06, + "loss": 1.3987, + "step": 1533 + }, + { + "epoch": 1.5757575757575757, + "grad_norm": 2.3516597747802734, + "learning_rate": 2.666539779215836e-06, + "loss": 1.3878, + "step": 1534 + }, + { + "epoch": 1.5767847971237803, + "grad_norm": 2.410268783569336, + "learning_rate": 2.6646364674533694e-06, + "loss": 1.3737, + "step": 1535 + }, + { + "epoch": 1.5778120184899846, + "grad_norm": 2.2420742511749268, + "learning_rate": 2.6627331556909024e-06, + "loss": 1.2908, + "step": 1536 + }, + { + "epoch": 1.578839239856189, + "grad_norm": 2.498166084289551, + "learning_rate": 2.6608298439284357e-06, + "loss": 1.383, + "step": 1537 + }, + { + "epoch": 1.5798664612223934, + "grad_norm": 2.389152765274048, + "learning_rate": 2.658926532165969e-06, + "loss": 1.3829, + "step": 1538 + }, + { + "epoch": 1.5808936825885977, + "grad_norm": 2.4197394847869873, + "learning_rate": 2.6570232204035024e-06, + "loss": 1.3438, + "step": 1539 + }, + { + "epoch": 1.5819209039548023, + "grad_norm": 2.321922540664673, + "learning_rate": 2.655119908641036e-06, + "loss": 1.3741, + "step": 1540 + }, + { + "epoch": 1.5819209039548023, + "eval_loss": 1.7643449306488037, + "eval_runtime": 38.0047, + "eval_samples_per_second": 26.313, + "eval_steps_per_second": 1.105, + "step": 1540 + }, + { + "epoch": 1.5829481253210067, + "grad_norm": 2.4056994915008545, + "learning_rate": 2.6532165968785687e-06, + "loss": 1.3857, + "step": 1541 + }, + { + "epoch": 1.583975346687211, + "grad_norm": 2.2752463817596436, + "learning_rate": 2.651313285116102e-06, + "loss": 1.3517, + "step": 1542 + }, + { + "epoch": 1.5850025680534157, + "grad_norm": 2.411092519760132, + "learning_rate": 2.6494099733536355e-06, + "loss": 1.3548, + "step": 1543 + }, + { + "epoch": 1.5860297894196198, + "grad_norm": 2.7468392848968506, + "learning_rate": 2.647506661591169e-06, + "loss": 1.5273, + "step": 1544 + }, + { + "epoch": 1.5870570107858244, + "grad_norm": 2.5081777572631836, + "learning_rate": 2.645603349828702e-06, + "loss": 1.4064, + "step": 1545 + }, + { + "epoch": 1.5880842321520288, + "grad_norm": 2.3444366455078125, + "learning_rate": 2.643700038066235e-06, + "loss": 1.3444, + "step": 1546 + }, + { + "epoch": 1.5891114535182331, + "grad_norm": 2.437304973602295, + "learning_rate": 2.6417967263037685e-06, + "loss": 1.4078, + "step": 1547 + }, + { + "epoch": 1.5901386748844377, + "grad_norm": 2.456725597381592, + "learning_rate": 2.639893414541302e-06, + "loss": 1.4212, + "step": 1548 + }, + { + "epoch": 1.5911658962506419, + "grad_norm": 2.4959378242492676, + "learning_rate": 2.6379901027788352e-06, + "loss": 1.4439, + "step": 1549 + }, + { + "epoch": 1.5921931176168465, + "grad_norm": 2.3856401443481445, + "learning_rate": 2.6360867910163686e-06, + "loss": 1.3748, + "step": 1550 + }, + { + "epoch": 1.5921931176168465, + "eval_loss": 1.7585556507110596, + "eval_runtime": 38.038, + "eval_samples_per_second": 26.289, + "eval_steps_per_second": 1.104, + "step": 1550 + }, + { + "epoch": 1.5932203389830508, + "grad_norm": 2.3285882472991943, + "learning_rate": 2.6341834792539024e-06, + "loss": 1.3912, + "step": 1551 + }, + { + "epoch": 1.5942475603492552, + "grad_norm": 2.365633249282837, + "learning_rate": 2.6322801674914357e-06, + "loss": 1.3922, + "step": 1552 + }, + { + "epoch": 1.5952747817154598, + "grad_norm": 2.460118293762207, + "learning_rate": 2.6303768557289687e-06, + "loss": 1.3899, + "step": 1553 + }, + { + "epoch": 1.5963020030816641, + "grad_norm": 2.244420051574707, + "learning_rate": 2.628473543966502e-06, + "loss": 1.347, + "step": 1554 + }, + { + "epoch": 1.5973292244478685, + "grad_norm": 2.404733180999756, + "learning_rate": 2.6265702322040354e-06, + "loss": 1.4163, + "step": 1555 + }, + { + "epoch": 1.5983564458140729, + "grad_norm": 2.3252789974212646, + "learning_rate": 2.6246669204415688e-06, + "loss": 1.3578, + "step": 1556 + }, + { + "epoch": 1.5993836671802772, + "grad_norm": 2.3664159774780273, + "learning_rate": 2.622763608679102e-06, + "loss": 1.3585, + "step": 1557 + }, + { + "epoch": 1.6004108885464818, + "grad_norm": 2.3188118934631348, + "learning_rate": 2.620860296916635e-06, + "loss": 1.3794, + "step": 1558 + }, + { + "epoch": 1.6014381099126862, + "grad_norm": 2.3163065910339355, + "learning_rate": 2.6189569851541684e-06, + "loss": 1.3157, + "step": 1559 + }, + { + "epoch": 1.6024653312788906, + "grad_norm": 2.4271090030670166, + "learning_rate": 2.617053673391702e-06, + "loss": 1.413, + "step": 1560 + }, + { + "epoch": 1.6024653312788906, + "eval_loss": 1.7588003873825073, + "eval_runtime": 38.0406, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 1560 + }, + { + "epoch": 1.6034925526450952, + "grad_norm": 2.5792505741119385, + "learning_rate": 2.615150361629235e-06, + "loss": 1.4431, + "step": 1561 + }, + { + "epoch": 1.6045197740112993, + "grad_norm": 2.4612159729003906, + "learning_rate": 2.6132470498667685e-06, + "loss": 1.3967, + "step": 1562 + }, + { + "epoch": 1.605546995377504, + "grad_norm": 2.4364020824432373, + "learning_rate": 2.6113437381043015e-06, + "loss": 1.3779, + "step": 1563 + }, + { + "epoch": 1.6065742167437083, + "grad_norm": 2.464442253112793, + "learning_rate": 2.609440426341835e-06, + "loss": 1.3685, + "step": 1564 + }, + { + "epoch": 1.6076014381099126, + "grad_norm": 2.4871442317962646, + "learning_rate": 2.607537114579368e-06, + "loss": 1.3979, + "step": 1565 + }, + { + "epoch": 1.6086286594761172, + "grad_norm": 2.375364303588867, + "learning_rate": 2.6056338028169015e-06, + "loss": 1.3836, + "step": 1566 + }, + { + "epoch": 1.6096558808423214, + "grad_norm": 2.364389657974243, + "learning_rate": 2.603730491054435e-06, + "loss": 1.3655, + "step": 1567 + }, + { + "epoch": 1.610683102208526, + "grad_norm": 2.421839714050293, + "learning_rate": 2.601827179291968e-06, + "loss": 1.398, + "step": 1568 + }, + { + "epoch": 1.6117103235747303, + "grad_norm": 2.3270041942596436, + "learning_rate": 2.599923867529501e-06, + "loss": 1.4061, + "step": 1569 + }, + { + "epoch": 1.6127375449409347, + "grad_norm": 2.4831383228302, + "learning_rate": 2.5980205557670346e-06, + "loss": 1.3808, + "step": 1570 + }, + { + "epoch": 1.6127375449409347, + "eval_loss": 1.773654818534851, + "eval_runtime": 38.046, + "eval_samples_per_second": 26.284, + "eval_steps_per_second": 1.104, + "step": 1570 + }, + { + "epoch": 1.6137647663071393, + "grad_norm": 2.499079942703247, + "learning_rate": 2.5961172440045684e-06, + "loss": 1.4167, + "step": 1571 + }, + { + "epoch": 1.6147919876733436, + "grad_norm": 2.3598642349243164, + "learning_rate": 2.5942139322421017e-06, + "loss": 1.4319, + "step": 1572 + }, + { + "epoch": 1.615819209039548, + "grad_norm": 2.3554258346557617, + "learning_rate": 2.592310620479635e-06, + "loss": 1.3723, + "step": 1573 + }, + { + "epoch": 1.6168464304057526, + "grad_norm": 2.2909231185913086, + "learning_rate": 2.5904073087171684e-06, + "loss": 1.339, + "step": 1574 + }, + { + "epoch": 1.6178736517719567, + "grad_norm": 2.3029019832611084, + "learning_rate": 2.5885039969547014e-06, + "loss": 1.4026, + "step": 1575 + }, + { + "epoch": 1.6189008731381613, + "grad_norm": 2.2370176315307617, + "learning_rate": 2.5866006851922347e-06, + "loss": 1.3531, + "step": 1576 + }, + { + "epoch": 1.6199280945043657, + "grad_norm": 2.2932159900665283, + "learning_rate": 2.584697373429768e-06, + "loss": 1.3585, + "step": 1577 + }, + { + "epoch": 1.62095531587057, + "grad_norm": 2.391493797302246, + "learning_rate": 2.5827940616673015e-06, + "loss": 1.3817, + "step": 1578 + }, + { + "epoch": 1.6219825372367747, + "grad_norm": 2.476572036743164, + "learning_rate": 2.580890749904835e-06, + "loss": 1.4241, + "step": 1579 + }, + { + "epoch": 1.6230097586029788, + "grad_norm": 2.3396525382995605, + "learning_rate": 2.5789874381423678e-06, + "loss": 1.3734, + "step": 1580 + }, + { + "epoch": 1.6230097586029788, + "eval_loss": 1.7641068696975708, + "eval_runtime": 38.0413, + "eval_samples_per_second": 26.287, + "eval_steps_per_second": 1.104, + "step": 1580 + }, + { + "epoch": 1.6240369799691834, + "grad_norm": 2.381436824798584, + "learning_rate": 2.577084126379901e-06, + "loss": 1.3764, + "step": 1581 + }, + { + "epoch": 1.6250642013353878, + "grad_norm": 2.391737937927246, + "learning_rate": 2.5751808146174345e-06, + "loss": 1.3419, + "step": 1582 + }, + { + "epoch": 1.6260914227015921, + "grad_norm": 2.453653573989868, + "learning_rate": 2.573277502854968e-06, + "loss": 1.398, + "step": 1583 + }, + { + "epoch": 1.6271186440677967, + "grad_norm": 2.3906450271606445, + "learning_rate": 2.5713741910925012e-06, + "loss": 1.3605, + "step": 1584 + }, + { + "epoch": 1.6281458654340009, + "grad_norm": 2.347041606903076, + "learning_rate": 2.569470879330034e-06, + "loss": 1.369, + "step": 1585 + }, + { + "epoch": 1.6291730868002054, + "grad_norm": 2.571047306060791, + "learning_rate": 2.5675675675675675e-06, + "loss": 1.421, + "step": 1586 + }, + { + "epoch": 1.6302003081664098, + "grad_norm": 2.3640646934509277, + "learning_rate": 2.565664255805101e-06, + "loss": 1.3736, + "step": 1587 + }, + { + "epoch": 1.6312275295326142, + "grad_norm": 2.419844388961792, + "learning_rate": 2.5637609440426343e-06, + "loss": 1.4008, + "step": 1588 + }, + { + "epoch": 1.6322547508988188, + "grad_norm": 2.3414361476898193, + "learning_rate": 2.5618576322801676e-06, + "loss": 1.3473, + "step": 1589 + }, + { + "epoch": 1.6332819722650231, + "grad_norm": 2.462259292602539, + "learning_rate": 2.5599543205177006e-06, + "loss": 1.4418, + "step": 1590 + }, + { + "epoch": 1.6332819722650231, + "eval_loss": 1.760501503944397, + "eval_runtime": 38.0428, + "eval_samples_per_second": 26.286, + "eval_steps_per_second": 1.104, + "step": 1590 + }, + { + "epoch": 1.6343091936312275, + "grad_norm": 2.49070405960083, + "learning_rate": 2.558051008755234e-06, + "loss": 1.3859, + "step": 1591 + }, + { + "epoch": 1.635336414997432, + "grad_norm": 2.500180244445801, + "learning_rate": 2.5561476969927677e-06, + "loss": 1.4379, + "step": 1592 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 2.4660637378692627, + "learning_rate": 2.554244385230301e-06, + "loss": 1.3656, + "step": 1593 + }, + { + "epoch": 1.6373908577298408, + "grad_norm": 2.5678634643554688, + "learning_rate": 2.5523410734678344e-06, + "loss": 1.3939, + "step": 1594 + }, + { + "epoch": 1.6384180790960452, + "grad_norm": 2.358930826187134, + "learning_rate": 2.550437761705368e-06, + "loss": 1.3281, + "step": 1595 + }, + { + "epoch": 1.6394453004622496, + "grad_norm": 2.462233543395996, + "learning_rate": 2.548534449942901e-06, + "loss": 1.3643, + "step": 1596 + }, + { + "epoch": 1.6404725218284542, + "grad_norm": 2.5294501781463623, + "learning_rate": 2.546631138180434e-06, + "loss": 1.4009, + "step": 1597 + }, + { + "epoch": 1.6414997431946583, + "grad_norm": 2.5325710773468018, + "learning_rate": 2.5447278264179675e-06, + "loss": 1.3797, + "step": 1598 + }, + { + "epoch": 1.642526964560863, + "grad_norm": 2.3293585777282715, + "learning_rate": 2.542824514655501e-06, + "loss": 1.3492, + "step": 1599 + }, + { + "epoch": 1.6435541859270673, + "grad_norm": 2.5357868671417236, + "learning_rate": 2.540921202893034e-06, + "loss": 1.4295, + "step": 1600 + }, + { + "epoch": 1.6435541859270673, + "eval_loss": 1.7499933242797852, + "eval_runtime": 38.0219, + "eval_samples_per_second": 26.301, + "eval_steps_per_second": 1.105, + "step": 1600 + }, + { + "epoch": 1.6445814072932716, + "grad_norm": 2.509333848953247, + "learning_rate": 2.5390178911305675e-06, + "loss": 1.4174, + "step": 1601 + }, + { + "epoch": 1.6456086286594762, + "grad_norm": 2.4453277587890625, + "learning_rate": 2.5371145793681005e-06, + "loss": 1.3898, + "step": 1602 + }, + { + "epoch": 1.6466358500256806, + "grad_norm": 2.506847381591797, + "learning_rate": 2.535211267605634e-06, + "loss": 1.3881, + "step": 1603 + }, + { + "epoch": 1.647663071391885, + "grad_norm": 2.2735178470611572, + "learning_rate": 2.533307955843167e-06, + "loss": 1.3268, + "step": 1604 + }, + { + "epoch": 1.6486902927580893, + "grad_norm": 2.479081630706787, + "learning_rate": 2.5314046440807006e-06, + "loss": 1.3897, + "step": 1605 + }, + { + "epoch": 1.6497175141242937, + "grad_norm": 2.3882839679718018, + "learning_rate": 2.529501332318234e-06, + "loss": 1.3734, + "step": 1606 + }, + { + "epoch": 1.6507447354904983, + "grad_norm": 2.521341323852539, + "learning_rate": 2.5275980205557673e-06, + "loss": 1.4298, + "step": 1607 + }, + { + "epoch": 1.6517719568567026, + "grad_norm": 2.55405855178833, + "learning_rate": 2.5256947087933002e-06, + "loss": 1.4497, + "step": 1608 + }, + { + "epoch": 1.652799178222907, + "grad_norm": 2.534478187561035, + "learning_rate": 2.5237913970308336e-06, + "loss": 1.438, + "step": 1609 + }, + { + "epoch": 1.6538263995891116, + "grad_norm": 2.453374147415161, + "learning_rate": 2.521888085268367e-06, + "loss": 1.3834, + "step": 1610 + }, + { + "epoch": 1.6538263995891116, + "eval_loss": 1.7550541162490845, + "eval_runtime": 37.6671, + "eval_samples_per_second": 26.548, + "eval_steps_per_second": 1.115, + "step": 1610 + }, + { + "epoch": 1.6548536209553157, + "grad_norm": 2.4754152297973633, + "learning_rate": 2.5199847735059003e-06, + "loss": 1.434, + "step": 1611 + }, + { + "epoch": 1.6558808423215203, + "grad_norm": 2.3785831928253174, + "learning_rate": 2.518081461743434e-06, + "loss": 1.3706, + "step": 1612 + }, + { + "epoch": 1.6569080636877247, + "grad_norm": 2.3936359882354736, + "learning_rate": 2.5161781499809675e-06, + "loss": 1.3955, + "step": 1613 + }, + { + "epoch": 1.657935285053929, + "grad_norm": 2.338012456893921, + "learning_rate": 2.5142748382185004e-06, + "loss": 1.3532, + "step": 1614 + }, + { + "epoch": 1.6589625064201337, + "grad_norm": 2.5027408599853516, + "learning_rate": 2.5123715264560338e-06, + "loss": 1.3522, + "step": 1615 + }, + { + "epoch": 1.6599897277863378, + "grad_norm": 2.4059648513793945, + "learning_rate": 2.510468214693567e-06, + "loss": 1.3767, + "step": 1616 + }, + { + "epoch": 1.6610169491525424, + "grad_norm": 2.324235200881958, + "learning_rate": 2.5085649029311005e-06, + "loss": 1.3248, + "step": 1617 + }, + { + "epoch": 1.6620441705187468, + "grad_norm": 2.336317300796509, + "learning_rate": 2.506661591168634e-06, + "loss": 1.288, + "step": 1618 + }, + { + "epoch": 1.6630713918849511, + "grad_norm": 2.3386120796203613, + "learning_rate": 2.504758279406167e-06, + "loss": 1.3715, + "step": 1619 + }, + { + "epoch": 1.6640986132511557, + "grad_norm": 2.5584142208099365, + "learning_rate": 2.5028549676437e-06, + "loss": 1.4157, + "step": 1620 + }, + { + "epoch": 1.6640986132511557, + "eval_loss": 1.7575956583023071, + "eval_runtime": 37.7897, + "eval_samples_per_second": 26.462, + "eval_steps_per_second": 1.111, + "step": 1620 + }, + { + "epoch": 1.66512583461736, + "grad_norm": 2.475881338119507, + "learning_rate": 2.5009516558812335e-06, + "loss": 1.357, + "step": 1621 + }, + { + "epoch": 1.6661530559835644, + "grad_norm": 2.4611690044403076, + "learning_rate": 2.499048344118767e-06, + "loss": 1.4077, + "step": 1622 + }, + { + "epoch": 1.667180277349769, + "grad_norm": 2.4230902194976807, + "learning_rate": 2.4971450323563002e-06, + "loss": 1.443, + "step": 1623 + }, + { + "epoch": 1.6682074987159732, + "grad_norm": 2.359095335006714, + "learning_rate": 2.4952417205938336e-06, + "loss": 1.3595, + "step": 1624 + }, + { + "epoch": 1.6692347200821778, + "grad_norm": 2.4128329753875732, + "learning_rate": 2.4933384088313666e-06, + "loss": 1.3636, + "step": 1625 + }, + { + "epoch": 1.6702619414483821, + "grad_norm": 2.3391597270965576, + "learning_rate": 2.4914350970689e-06, + "loss": 1.3536, + "step": 1626 + }, + { + "epoch": 1.6712891628145865, + "grad_norm": 2.478581190109253, + "learning_rate": 2.4895317853064337e-06, + "loss": 1.3642, + "step": 1627 + }, + { + "epoch": 1.672316384180791, + "grad_norm": 2.7438251972198486, + "learning_rate": 2.4876284735439666e-06, + "loss": 1.4192, + "step": 1628 + }, + { + "epoch": 1.6733436055469952, + "grad_norm": 2.508089542388916, + "learning_rate": 2.4857251617815e-06, + "loss": 1.3879, + "step": 1629 + }, + { + "epoch": 1.6743708269131998, + "grad_norm": 2.5251338481903076, + "learning_rate": 2.4838218500190334e-06, + "loss": 1.3936, + "step": 1630 + }, + { + "epoch": 1.6743708269131998, + "eval_loss": 1.7578516006469727, + "eval_runtime": 37.8753, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1630 + }, + { + "epoch": 1.6753980482794042, + "grad_norm": 2.3655474185943604, + "learning_rate": 2.4819185382565667e-06, + "loss": 1.3839, + "step": 1631 + }, + { + "epoch": 1.6764252696456086, + "grad_norm": 2.4703869819641113, + "learning_rate": 2.4800152264941e-06, + "loss": 1.3767, + "step": 1632 + }, + { + "epoch": 1.6774524910118132, + "grad_norm": 2.4739503860473633, + "learning_rate": 2.478111914731633e-06, + "loss": 1.3465, + "step": 1633 + }, + { + "epoch": 1.6784797123780173, + "grad_norm": 2.356006383895874, + "learning_rate": 2.4762086029691664e-06, + "loss": 1.3828, + "step": 1634 + }, + { + "epoch": 1.6795069337442219, + "grad_norm": 2.7019715309143066, + "learning_rate": 2.4743052912066998e-06, + "loss": 1.4875, + "step": 1635 + }, + { + "epoch": 1.6805341551104263, + "grad_norm": 2.3504486083984375, + "learning_rate": 2.472401979444233e-06, + "loss": 1.3494, + "step": 1636 + }, + { + "epoch": 1.6815613764766306, + "grad_norm": 2.6056807041168213, + "learning_rate": 2.4704986676817665e-06, + "loss": 1.4399, + "step": 1637 + }, + { + "epoch": 1.6825885978428352, + "grad_norm": 2.481422185897827, + "learning_rate": 2.4685953559193e-06, + "loss": 1.3803, + "step": 1638 + }, + { + "epoch": 1.6836158192090396, + "grad_norm": 2.483227491378784, + "learning_rate": 2.466692044156833e-06, + "loss": 1.4255, + "step": 1639 + }, + { + "epoch": 1.684643040575244, + "grad_norm": 2.4171221256256104, + "learning_rate": 2.4647887323943666e-06, + "loss": 1.3541, + "step": 1640 + }, + { + "epoch": 1.684643040575244, + "eval_loss": 1.7658556699752808, + "eval_runtime": 37.8673, + "eval_samples_per_second": 26.408, + "eval_steps_per_second": 1.109, + "step": 1640 + }, + { + "epoch": 1.6856702619414485, + "grad_norm": 2.2551865577697754, + "learning_rate": 2.4628854206319e-06, + "loss": 1.334, + "step": 1641 + }, + { + "epoch": 1.6866974833076527, + "grad_norm": 2.45120906829834, + "learning_rate": 2.460982108869433e-06, + "loss": 1.3416, + "step": 1642 + }, + { + "epoch": 1.6877247046738573, + "grad_norm": 2.4356141090393066, + "learning_rate": 2.4590787971069662e-06, + "loss": 1.4222, + "step": 1643 + }, + { + "epoch": 1.6887519260400616, + "grad_norm": 2.3313231468200684, + "learning_rate": 2.4571754853444996e-06, + "loss": 1.3654, + "step": 1644 + }, + { + "epoch": 1.689779147406266, + "grad_norm": 2.3518543243408203, + "learning_rate": 2.455272173582033e-06, + "loss": 1.3521, + "step": 1645 + }, + { + "epoch": 1.6908063687724706, + "grad_norm": 2.477158546447754, + "learning_rate": 2.4533688618195663e-06, + "loss": 1.3665, + "step": 1646 + }, + { + "epoch": 1.6918335901386747, + "grad_norm": 2.2717649936676025, + "learning_rate": 2.4514655500570997e-06, + "loss": 1.2992, + "step": 1647 + }, + { + "epoch": 1.6928608115048793, + "grad_norm": 2.2843265533447266, + "learning_rate": 2.449562238294633e-06, + "loss": 1.3458, + "step": 1648 + }, + { + "epoch": 1.6938880328710837, + "grad_norm": 2.4243931770324707, + "learning_rate": 2.4476589265321664e-06, + "loss": 1.3143, + "step": 1649 + }, + { + "epoch": 1.694915254237288, + "grad_norm": 2.4439947605133057, + "learning_rate": 2.4457556147696993e-06, + "loss": 1.41, + "step": 1650 + }, + { + "epoch": 1.694915254237288, + "eval_loss": 1.762351155281067, + "eval_runtime": 37.8973, + "eval_samples_per_second": 26.387, + "eval_steps_per_second": 1.108, + "step": 1650 + }, + { + "epoch": 1.6959424756034926, + "grad_norm": 2.3726727962493896, + "learning_rate": 2.4438523030072327e-06, + "loss": 1.388, + "step": 1651 + }, + { + "epoch": 1.696969696969697, + "grad_norm": 2.4449899196624756, + "learning_rate": 2.441948991244766e-06, + "loss": 1.4023, + "step": 1652 + }, + { + "epoch": 1.6979969183359014, + "grad_norm": 2.4462502002716064, + "learning_rate": 2.4400456794822994e-06, + "loss": 1.3761, + "step": 1653 + }, + { + "epoch": 1.6990241397021058, + "grad_norm": 2.413308620452881, + "learning_rate": 2.438142367719833e-06, + "loss": 1.4017, + "step": 1654 + }, + { + "epoch": 1.7000513610683101, + "grad_norm": 2.398103952407837, + "learning_rate": 2.4362390559573657e-06, + "loss": 1.3709, + "step": 1655 + }, + { + "epoch": 1.7010785824345147, + "grad_norm": 2.4637506008148193, + "learning_rate": 2.434335744194899e-06, + "loss": 1.3657, + "step": 1656 + }, + { + "epoch": 1.702105803800719, + "grad_norm": 2.419947385787964, + "learning_rate": 2.432432432432433e-06, + "loss": 1.4071, + "step": 1657 + }, + { + "epoch": 1.7031330251669234, + "grad_norm": 2.4477648735046387, + "learning_rate": 2.4305291206699662e-06, + "loss": 1.3288, + "step": 1658 + }, + { + "epoch": 1.704160246533128, + "grad_norm": 2.3359293937683105, + "learning_rate": 2.428625808907499e-06, + "loss": 1.396, + "step": 1659 + }, + { + "epoch": 1.7051874678993322, + "grad_norm": 2.396193742752075, + "learning_rate": 2.4267224971450325e-06, + "loss": 1.4246, + "step": 1660 + }, + { + "epoch": 1.7051874678993322, + "eval_loss": 1.7526435852050781, + "eval_runtime": 37.9178, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 1660 + }, + { + "epoch": 1.7062146892655368, + "grad_norm": 2.511453628540039, + "learning_rate": 2.424819185382566e-06, + "loss": 1.3877, + "step": 1661 + }, + { + "epoch": 1.7072419106317411, + "grad_norm": 2.4472320079803467, + "learning_rate": 2.4229158736200993e-06, + "loss": 1.3346, + "step": 1662 + }, + { + "epoch": 1.7082691319979455, + "grad_norm": 2.2219882011413574, + "learning_rate": 2.4210125618576326e-06, + "loss": 1.3352, + "step": 1663 + }, + { + "epoch": 1.70929635336415, + "grad_norm": 2.440121650695801, + "learning_rate": 2.4191092500951656e-06, + "loss": 1.3644, + "step": 1664 + }, + { + "epoch": 1.7103235747303542, + "grad_norm": 2.5501232147216797, + "learning_rate": 2.417205938332699e-06, + "loss": 1.4073, + "step": 1665 + }, + { + "epoch": 1.7113507960965588, + "grad_norm": 2.4315030574798584, + "learning_rate": 2.4153026265702323e-06, + "loss": 1.3137, + "step": 1666 + }, + { + "epoch": 1.7123780174627632, + "grad_norm": 2.3719727993011475, + "learning_rate": 2.4133993148077657e-06, + "loss": 1.3484, + "step": 1667 + }, + { + "epoch": 1.7134052388289676, + "grad_norm": 2.3800482749938965, + "learning_rate": 2.411496003045299e-06, + "loss": 1.4106, + "step": 1668 + }, + { + "epoch": 1.7144324601951721, + "grad_norm": 2.4554190635681152, + "learning_rate": 2.4095926912828324e-06, + "loss": 1.3603, + "step": 1669 + }, + { + "epoch": 1.7154596815613765, + "grad_norm": 2.2984702587127686, + "learning_rate": 2.4076893795203658e-06, + "loss": 1.3739, + "step": 1670 + }, + { + "epoch": 1.7154596815613765, + "eval_loss": 1.752260446548462, + "eval_runtime": 37.9208, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 1670 + }, + { + "epoch": 1.7164869029275809, + "grad_norm": 2.3967223167419434, + "learning_rate": 2.405786067757899e-06, + "loss": 1.3598, + "step": 1671 + }, + { + "epoch": 1.7175141242937855, + "grad_norm": 2.585935592651367, + "learning_rate": 2.403882755995432e-06, + "loss": 1.4489, + "step": 1672 + }, + { + "epoch": 1.7185413456599896, + "grad_norm": 2.4227662086486816, + "learning_rate": 2.4019794442329654e-06, + "loss": 1.3745, + "step": 1673 + }, + { + "epoch": 1.7195685670261942, + "grad_norm": 2.485344648361206, + "learning_rate": 2.4000761324704988e-06, + "loss": 1.3405, + "step": 1674 + }, + { + "epoch": 1.7205957883923986, + "grad_norm": 2.4055545330047607, + "learning_rate": 2.398172820708032e-06, + "loss": 1.3914, + "step": 1675 + }, + { + "epoch": 1.721623009758603, + "grad_norm": 2.3449783325195312, + "learning_rate": 2.3962695089455655e-06, + "loss": 1.3637, + "step": 1676 + }, + { + "epoch": 1.7226502311248075, + "grad_norm": 2.338590621948242, + "learning_rate": 2.3943661971830984e-06, + "loss": 1.3571, + "step": 1677 + }, + { + "epoch": 1.7236774524910117, + "grad_norm": 2.3274776935577393, + "learning_rate": 2.3924628854206322e-06, + "loss": 1.3632, + "step": 1678 + }, + { + "epoch": 1.7247046738572163, + "grad_norm": 2.3467724323272705, + "learning_rate": 2.3905595736581656e-06, + "loss": 1.3568, + "step": 1679 + }, + { + "epoch": 1.7257318952234206, + "grad_norm": 2.552039861679077, + "learning_rate": 2.388656261895699e-06, + "loss": 1.4099, + "step": 1680 + }, + { + "epoch": 1.7257318952234206, + "eval_loss": 1.745519995689392, + "eval_runtime": 37.9527, + "eval_samples_per_second": 26.349, + "eval_steps_per_second": 1.107, + "step": 1680 + }, + { + "epoch": 1.726759116589625, + "grad_norm": 2.5332300662994385, + "learning_rate": 2.386752950133232e-06, + "loss": 1.4017, + "step": 1681 + }, + { + "epoch": 1.7277863379558296, + "grad_norm": 2.4033122062683105, + "learning_rate": 2.3848496383707653e-06, + "loss": 1.3866, + "step": 1682 + }, + { + "epoch": 1.7288135593220337, + "grad_norm": 2.460071086883545, + "learning_rate": 2.3829463266082986e-06, + "loss": 1.3824, + "step": 1683 + }, + { + "epoch": 1.7298407806882383, + "grad_norm": 2.5089826583862305, + "learning_rate": 2.381043014845832e-06, + "loss": 1.398, + "step": 1684 + }, + { + "epoch": 1.7308680020544427, + "grad_norm": 2.419689893722534, + "learning_rate": 2.3791397030833653e-06, + "loss": 1.3852, + "step": 1685 + }, + { + "epoch": 1.731895223420647, + "grad_norm": 2.314080238342285, + "learning_rate": 2.3772363913208983e-06, + "loss": 1.3666, + "step": 1686 + }, + { + "epoch": 1.7329224447868516, + "grad_norm": 2.324474334716797, + "learning_rate": 2.3753330795584316e-06, + "loss": 1.3495, + "step": 1687 + }, + { + "epoch": 1.733949666153056, + "grad_norm": 2.5079402923583984, + "learning_rate": 2.3734297677959654e-06, + "loss": 1.3731, + "step": 1688 + }, + { + "epoch": 1.7349768875192604, + "grad_norm": 2.492154598236084, + "learning_rate": 2.3715264560334984e-06, + "loss": 1.413, + "step": 1689 + }, + { + "epoch": 1.736004108885465, + "grad_norm": 2.3400940895080566, + "learning_rate": 2.3696231442710317e-06, + "loss": 1.3952, + "step": 1690 + }, + { + "epoch": 1.736004108885465, + "eval_loss": 1.7412214279174805, + "eval_runtime": 37.9159, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 1690 + }, + { + "epoch": 1.7370313302516691, + "grad_norm": 2.3744688034057617, + "learning_rate": 2.367719832508565e-06, + "loss": 1.396, + "step": 1691 + }, + { + "epoch": 1.7380585516178737, + "grad_norm": 2.4762585163116455, + "learning_rate": 2.3658165207460985e-06, + "loss": 1.4113, + "step": 1692 + }, + { + "epoch": 1.739085772984078, + "grad_norm": 2.361391305923462, + "learning_rate": 2.363913208983632e-06, + "loss": 1.3962, + "step": 1693 + }, + { + "epoch": 1.7401129943502824, + "grad_norm": 2.4838762283325195, + "learning_rate": 2.362009897221165e-06, + "loss": 1.4098, + "step": 1694 + }, + { + "epoch": 1.741140215716487, + "grad_norm": 2.4860999584198, + "learning_rate": 2.360106585458698e-06, + "loss": 1.4006, + "step": 1695 + }, + { + "epoch": 1.7421674370826912, + "grad_norm": 2.3834593296051025, + "learning_rate": 2.3582032736962315e-06, + "loss": 1.3445, + "step": 1696 + }, + { + "epoch": 1.7431946584488958, + "grad_norm": 2.5383975505828857, + "learning_rate": 2.356299961933765e-06, + "loss": 1.3849, + "step": 1697 + }, + { + "epoch": 1.7442218798151001, + "grad_norm": 2.414038896560669, + "learning_rate": 2.354396650171298e-06, + "loss": 1.356, + "step": 1698 + }, + { + "epoch": 1.7452491011813045, + "grad_norm": 2.3932273387908936, + "learning_rate": 2.3524933384088316e-06, + "loss": 1.3422, + "step": 1699 + }, + { + "epoch": 1.746276322547509, + "grad_norm": 2.5618491172790527, + "learning_rate": 2.350590026646365e-06, + "loss": 1.4324, + "step": 1700 + }, + { + "epoch": 1.746276322547509, + "eval_loss": 1.7560412883758545, + "eval_runtime": 37.9455, + "eval_samples_per_second": 26.354, + "eval_steps_per_second": 1.107, + "step": 1700 + }, + { + "epoch": 1.7473035439137135, + "grad_norm": 2.450718879699707, + "learning_rate": 2.3486867148838983e-06, + "loss": 1.3631, + "step": 1701 + }, + { + "epoch": 1.7483307652799178, + "grad_norm": 2.4136962890625, + "learning_rate": 2.3467834031214317e-06, + "loss": 1.3528, + "step": 1702 + }, + { + "epoch": 1.7493579866461222, + "grad_norm": 2.4310078620910645, + "learning_rate": 2.3448800913589646e-06, + "loss": 1.369, + "step": 1703 + }, + { + "epoch": 1.7503852080123266, + "grad_norm": 2.5222883224487305, + "learning_rate": 2.342976779596498e-06, + "loss": 1.3897, + "step": 1704 + }, + { + "epoch": 1.7514124293785311, + "grad_norm": 2.3903377056121826, + "learning_rate": 2.3410734678340313e-06, + "loss": 1.3526, + "step": 1705 + }, + { + "epoch": 1.7524396507447355, + "grad_norm": 2.3869972229003906, + "learning_rate": 2.3391701560715647e-06, + "loss": 1.4091, + "step": 1706 + }, + { + "epoch": 1.7534668721109399, + "grad_norm": 2.3368723392486572, + "learning_rate": 2.337266844309098e-06, + "loss": 1.3671, + "step": 1707 + }, + { + "epoch": 1.7544940934771445, + "grad_norm": 2.4195168018341064, + "learning_rate": 2.3353635325466314e-06, + "loss": 1.3701, + "step": 1708 + }, + { + "epoch": 1.7555213148433486, + "grad_norm": 2.4312965869903564, + "learning_rate": 2.3334602207841648e-06, + "loss": 1.3795, + "step": 1709 + }, + { + "epoch": 1.7565485362095532, + "grad_norm": 2.582592725753784, + "learning_rate": 2.331556909021698e-06, + "loss": 1.3743, + "step": 1710 + }, + { + "epoch": 1.7565485362095532, + "eval_loss": 1.7604093551635742, + "eval_runtime": 37.6917, + "eval_samples_per_second": 26.531, + "eval_steps_per_second": 1.114, + "step": 1710 + }, + { + "epoch": 1.7575757575757576, + "grad_norm": 2.4194061756134033, + "learning_rate": 2.3296535972592315e-06, + "loss": 1.4049, + "step": 1711 + }, + { + "epoch": 1.758602978941962, + "grad_norm": 2.4093074798583984, + "learning_rate": 2.3277502854967644e-06, + "loss": 1.352, + "step": 1712 + }, + { + "epoch": 1.7596302003081665, + "grad_norm": 2.509335517883301, + "learning_rate": 2.325846973734298e-06, + "loss": 1.4221, + "step": 1713 + }, + { + "epoch": 1.7606574216743707, + "grad_norm": 2.422558069229126, + "learning_rate": 2.323943661971831e-06, + "loss": 1.3508, + "step": 1714 + }, + { + "epoch": 1.7616846430405753, + "grad_norm": 2.3949813842773438, + "learning_rate": 2.3220403502093645e-06, + "loss": 1.4146, + "step": 1715 + }, + { + "epoch": 1.7627118644067796, + "grad_norm": 2.47383189201355, + "learning_rate": 2.320137038446898e-06, + "loss": 1.413, + "step": 1716 + }, + { + "epoch": 1.763739085772984, + "grad_norm": 2.403531312942505, + "learning_rate": 2.318233726684431e-06, + "loss": 1.4181, + "step": 1717 + }, + { + "epoch": 1.7647663071391886, + "grad_norm": 2.463175058364868, + "learning_rate": 2.3163304149219646e-06, + "loss": 1.4121, + "step": 1718 + }, + { + "epoch": 1.765793528505393, + "grad_norm": 2.4082279205322266, + "learning_rate": 2.314427103159498e-06, + "loss": 1.3844, + "step": 1719 + }, + { + "epoch": 1.7668207498715973, + "grad_norm": 2.2674102783203125, + "learning_rate": 2.312523791397031e-06, + "loss": 1.3647, + "step": 1720 + }, + { + "epoch": 1.7668207498715973, + "eval_loss": 1.7476056814193726, + "eval_runtime": 37.8818, + "eval_samples_per_second": 26.398, + "eval_steps_per_second": 1.109, + "step": 1720 + }, + { + "epoch": 1.767847971237802, + "grad_norm": 2.2334558963775635, + "learning_rate": 2.3106204796345643e-06, + "loss": 1.3425, + "step": 1721 + }, + { + "epoch": 1.768875192604006, + "grad_norm": 2.3458974361419678, + "learning_rate": 2.3087171678720976e-06, + "loss": 1.3707, + "step": 1722 + }, + { + "epoch": 1.7699024139702106, + "grad_norm": 2.437530040740967, + "learning_rate": 2.306813856109631e-06, + "loss": 1.422, + "step": 1723 + }, + { + "epoch": 1.770929635336415, + "grad_norm": 2.467456817626953, + "learning_rate": 2.3049105443471644e-06, + "loss": 1.3795, + "step": 1724 + }, + { + "epoch": 1.7719568567026194, + "grad_norm": 2.485546827316284, + "learning_rate": 2.3030072325846973e-06, + "loss": 1.3648, + "step": 1725 + }, + { + "epoch": 1.772984078068824, + "grad_norm": 2.4228715896606445, + "learning_rate": 2.3011039208222307e-06, + "loss": 1.3508, + "step": 1726 + }, + { + "epoch": 1.774011299435028, + "grad_norm": 2.3490238189697266, + "learning_rate": 2.299200609059764e-06, + "loss": 1.3587, + "step": 1727 + }, + { + "epoch": 1.7750385208012327, + "grad_norm": 2.333501100540161, + "learning_rate": 2.297297297297298e-06, + "loss": 1.3593, + "step": 1728 + }, + { + "epoch": 1.776065742167437, + "grad_norm": 2.459859609603882, + "learning_rate": 2.2953939855348308e-06, + "loss": 1.3333, + "step": 1729 + }, + { + "epoch": 1.7770929635336414, + "grad_norm": 2.476691961288452, + "learning_rate": 2.293490673772364e-06, + "loss": 1.4137, + "step": 1730 + }, + { + "epoch": 1.7770929635336414, + "eval_loss": 1.7561018466949463, + "eval_runtime": 37.9778, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1730 + }, + { + "epoch": 1.778120184899846, + "grad_norm": 2.443805694580078, + "learning_rate": 2.2915873620098975e-06, + "loss": 1.3981, + "step": 1731 + }, + { + "epoch": 1.7791474062660502, + "grad_norm": 2.4164700508117676, + "learning_rate": 2.289684050247431e-06, + "loss": 1.392, + "step": 1732 + }, + { + "epoch": 1.7801746276322548, + "grad_norm": 2.3005855083465576, + "learning_rate": 2.287780738484964e-06, + "loss": 1.3603, + "step": 1733 + }, + { + "epoch": 1.7812018489984591, + "grad_norm": 2.395570755004883, + "learning_rate": 2.285877426722497e-06, + "loss": 1.383, + "step": 1734 + }, + { + "epoch": 1.7822290703646635, + "grad_norm": 2.4991118907928467, + "learning_rate": 2.2839741149600305e-06, + "loss": 1.4121, + "step": 1735 + }, + { + "epoch": 1.783256291730868, + "grad_norm": 2.420830726623535, + "learning_rate": 2.282070803197564e-06, + "loss": 1.3997, + "step": 1736 + }, + { + "epoch": 1.7842835130970724, + "grad_norm": 2.3576571941375732, + "learning_rate": 2.2801674914350972e-06, + "loss": 1.3825, + "step": 1737 + }, + { + "epoch": 1.7853107344632768, + "grad_norm": 2.355811595916748, + "learning_rate": 2.2782641796726306e-06, + "loss": 1.3716, + "step": 1738 + }, + { + "epoch": 1.7863379558294814, + "grad_norm": 2.3596901893615723, + "learning_rate": 2.276360867910164e-06, + "loss": 1.4064, + "step": 1739 + }, + { + "epoch": 1.7873651771956856, + "grad_norm": 2.301748514175415, + "learning_rate": 2.2744575561476973e-06, + "loss": 1.3358, + "step": 1740 + }, + { + "epoch": 1.7873651771956856, + "eval_loss": 1.7705498933792114, + "eval_runtime": 37.9787, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1740 + }, + { + "epoch": 1.7883923985618901, + "grad_norm": 2.3325891494750977, + "learning_rate": 2.2725542443852307e-06, + "loss": 1.3749, + "step": 1741 + }, + { + "epoch": 1.7894196199280945, + "grad_norm": 2.3880693912506104, + "learning_rate": 2.2706509326227636e-06, + "loss": 1.4153, + "step": 1742 + }, + { + "epoch": 1.7904468412942989, + "grad_norm": 2.3157029151916504, + "learning_rate": 2.268747620860297e-06, + "loss": 1.3899, + "step": 1743 + }, + { + "epoch": 1.7914740626605035, + "grad_norm": 2.4292151927948, + "learning_rate": 2.2668443090978303e-06, + "loss": 1.3804, + "step": 1744 + }, + { + "epoch": 1.7925012840267076, + "grad_norm": 2.4811818599700928, + "learning_rate": 2.2649409973353637e-06, + "loss": 1.4036, + "step": 1745 + }, + { + "epoch": 1.7935285053929122, + "grad_norm": 2.260420799255371, + "learning_rate": 2.263037685572897e-06, + "loss": 1.3616, + "step": 1746 + }, + { + "epoch": 1.7945557267591166, + "grad_norm": 2.262769937515259, + "learning_rate": 2.26113437381043e-06, + "loss": 1.351, + "step": 1747 + }, + { + "epoch": 1.795582948125321, + "grad_norm": 2.481602191925049, + "learning_rate": 2.259231062047964e-06, + "loss": 1.3816, + "step": 1748 + }, + { + "epoch": 1.7966101694915255, + "grad_norm": 2.584700345993042, + "learning_rate": 2.257327750285497e-06, + "loss": 1.4026, + "step": 1749 + }, + { + "epoch": 1.7976373908577299, + "grad_norm": 2.3472707271575928, + "learning_rate": 2.2554244385230305e-06, + "loss": 1.37, + "step": 1750 + }, + { + "epoch": 1.7976373908577299, + "eval_loss": 1.7779643535614014, + "eval_runtime": 38.0042, + "eval_samples_per_second": 26.313, + "eval_steps_per_second": 1.105, + "step": 1750 + }, + { + "epoch": 1.7986646122239343, + "grad_norm": 2.3616456985473633, + "learning_rate": 2.2535211267605635e-06, + "loss": 1.3713, + "step": 1751 + }, + { + "epoch": 1.7996918335901386, + "grad_norm": 2.4747135639190674, + "learning_rate": 2.251617814998097e-06, + "loss": 1.3797, + "step": 1752 + }, + { + "epoch": 1.800719054956343, + "grad_norm": 2.396944284439087, + "learning_rate": 2.24971450323563e-06, + "loss": 1.3081, + "step": 1753 + }, + { + "epoch": 1.8017462763225476, + "grad_norm": 2.5272505283355713, + "learning_rate": 2.2478111914731636e-06, + "loss": 1.3888, + "step": 1754 + }, + { + "epoch": 1.802773497688752, + "grad_norm": 2.319617748260498, + "learning_rate": 2.245907879710697e-06, + "loss": 1.2698, + "step": 1755 + }, + { + "epoch": 1.8038007190549563, + "grad_norm": 2.4994258880615234, + "learning_rate": 2.24400456794823e-06, + "loss": 1.4026, + "step": 1756 + }, + { + "epoch": 1.804827940421161, + "grad_norm": 2.3463289737701416, + "learning_rate": 2.2421012561857632e-06, + "loss": 1.3682, + "step": 1757 + }, + { + "epoch": 1.805855161787365, + "grad_norm": 2.3547732830047607, + "learning_rate": 2.240197944423297e-06, + "loss": 1.3486, + "step": 1758 + }, + { + "epoch": 1.8068823831535696, + "grad_norm": 2.381570816040039, + "learning_rate": 2.23829463266083e-06, + "loss": 1.3531, + "step": 1759 + }, + { + "epoch": 1.807909604519774, + "grad_norm": 2.2834391593933105, + "learning_rate": 2.2363913208983633e-06, + "loss": 1.3467, + "step": 1760 + }, + { + "epoch": 1.807909604519774, + "eval_loss": 1.7836403846740723, + "eval_runtime": 37.966, + "eval_samples_per_second": 26.339, + "eval_steps_per_second": 1.106, + "step": 1760 + }, + { + "epoch": 1.8089368258859784, + "grad_norm": 2.3574178218841553, + "learning_rate": 2.2344880091358967e-06, + "loss": 1.3867, + "step": 1761 + }, + { + "epoch": 1.809964047252183, + "grad_norm": 2.3440022468566895, + "learning_rate": 2.23258469737343e-06, + "loss": 1.3652, + "step": 1762 + }, + { + "epoch": 1.810991268618387, + "grad_norm": 2.4115090370178223, + "learning_rate": 2.2306813856109634e-06, + "loss": 1.4059, + "step": 1763 + }, + { + "epoch": 1.8120184899845917, + "grad_norm": 2.3989944458007812, + "learning_rate": 2.2287780738484963e-06, + "loss": 1.3572, + "step": 1764 + }, + { + "epoch": 1.813045711350796, + "grad_norm": 2.394892930984497, + "learning_rate": 2.2268747620860297e-06, + "loss": 1.433, + "step": 1765 + }, + { + "epoch": 1.8140729327170004, + "grad_norm": 2.5118515491485596, + "learning_rate": 2.224971450323563e-06, + "loss": 1.3714, + "step": 1766 + }, + { + "epoch": 1.815100154083205, + "grad_norm": 2.392695426940918, + "learning_rate": 2.2230681385610964e-06, + "loss": 1.3061, + "step": 1767 + }, + { + "epoch": 1.8161273754494094, + "grad_norm": 2.4015321731567383, + "learning_rate": 2.2211648267986298e-06, + "loss": 1.3681, + "step": 1768 + }, + { + "epoch": 1.8171545968156138, + "grad_norm": 2.3518455028533936, + "learning_rate": 2.219261515036163e-06, + "loss": 1.3549, + "step": 1769 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 2.40037202835083, + "learning_rate": 2.2173582032736965e-06, + "loss": 1.3882, + "step": 1770 + }, + { + "epoch": 1.8181818181818183, + "eval_loss": 1.7858384847640991, + "eval_runtime": 37.9787, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 1770 + }, + { + "epoch": 1.8192090395480225, + "grad_norm": 2.4189655780792236, + "learning_rate": 2.21545489151123e-06, + "loss": 1.4019, + "step": 1771 + }, + { + "epoch": 1.820236260914227, + "grad_norm": 2.4965226650238037, + "learning_rate": 2.2135515797487632e-06, + "loss": 1.3703, + "step": 1772 + }, + { + "epoch": 1.8212634822804314, + "grad_norm": 2.4330036640167236, + "learning_rate": 2.211648267986296e-06, + "loss": 1.3689, + "step": 1773 + }, + { + "epoch": 1.8222907036466358, + "grad_norm": 2.3654327392578125, + "learning_rate": 2.2097449562238295e-06, + "loss": 1.3677, + "step": 1774 + }, + { + "epoch": 1.8233179250128404, + "grad_norm": 2.4047489166259766, + "learning_rate": 2.207841644461363e-06, + "loss": 1.3626, + "step": 1775 + }, + { + "epoch": 1.8243451463790445, + "grad_norm": 2.563310146331787, + "learning_rate": 2.2059383326988963e-06, + "loss": 1.424, + "step": 1776 + }, + { + "epoch": 1.8253723677452491, + "grad_norm": 2.506540536880493, + "learning_rate": 2.2040350209364296e-06, + "loss": 1.3727, + "step": 1777 + }, + { + "epoch": 1.8263995891114535, + "grad_norm": 2.271742105484009, + "learning_rate": 2.2021317091739626e-06, + "loss": 1.3412, + "step": 1778 + }, + { + "epoch": 1.8274268104776579, + "grad_norm": 2.3172881603240967, + "learning_rate": 2.2002283974114963e-06, + "loss": 1.3812, + "step": 1779 + }, + { + "epoch": 1.8284540318438625, + "grad_norm": 2.3927204608917236, + "learning_rate": 2.1983250856490297e-06, + "loss": 1.3148, + "step": 1780 + }, + { + "epoch": 1.8284540318438625, + "eval_loss": 1.7755651473999023, + "eval_runtime": 37.9896, + "eval_samples_per_second": 26.323, + "eval_steps_per_second": 1.106, + "step": 1780 + }, + { + "epoch": 1.8294812532100666, + "grad_norm": 2.391972303390503, + "learning_rate": 2.1964217738865626e-06, + "loss": 1.3553, + "step": 1781 + }, + { + "epoch": 1.8305084745762712, + "grad_norm": 2.3727471828460693, + "learning_rate": 2.194518462124096e-06, + "loss": 1.377, + "step": 1782 + }, + { + "epoch": 1.8315356959424756, + "grad_norm": 2.5614147186279297, + "learning_rate": 2.1926151503616294e-06, + "loss": 1.4188, + "step": 1783 + }, + { + "epoch": 1.83256291730868, + "grad_norm": 2.344714641571045, + "learning_rate": 2.1907118385991627e-06, + "loss": 1.3877, + "step": 1784 + }, + { + "epoch": 1.8335901386748845, + "grad_norm": 2.34287428855896, + "learning_rate": 2.188808526836696e-06, + "loss": 1.351, + "step": 1785 + }, + { + "epoch": 1.8346173600410889, + "grad_norm": 2.343554735183716, + "learning_rate": 2.1869052150742295e-06, + "loss": 1.382, + "step": 1786 + }, + { + "epoch": 1.8356445814072933, + "grad_norm": 2.2418034076690674, + "learning_rate": 2.1850019033117624e-06, + "loss": 1.3062, + "step": 1787 + }, + { + "epoch": 1.8366718027734978, + "grad_norm": 2.4358294010162354, + "learning_rate": 2.1830985915492958e-06, + "loss": 1.3681, + "step": 1788 + }, + { + "epoch": 1.837699024139702, + "grad_norm": 2.460731029510498, + "learning_rate": 2.1811952797868295e-06, + "loss": 1.3791, + "step": 1789 + }, + { + "epoch": 1.8387262455059066, + "grad_norm": 2.4244463443756104, + "learning_rate": 2.1792919680243625e-06, + "loss": 1.3528, + "step": 1790 + }, + { + "epoch": 1.8387262455059066, + "eval_loss": 1.764306664466858, + "eval_runtime": 37.9906, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 1790 + }, + { + "epoch": 1.839753466872111, + "grad_norm": 2.3808743953704834, + "learning_rate": 2.177388656261896e-06, + "loss": 1.3882, + "step": 1791 + }, + { + "epoch": 1.8407806882383153, + "grad_norm": 2.443690538406372, + "learning_rate": 2.1754853444994292e-06, + "loss": 1.4542, + "step": 1792 + }, + { + "epoch": 1.84180790960452, + "grad_norm": 2.3968520164489746, + "learning_rate": 2.1735820327369626e-06, + "loss": 1.2815, + "step": 1793 + }, + { + "epoch": 1.842835130970724, + "grad_norm": 2.5139951705932617, + "learning_rate": 2.171678720974496e-06, + "loss": 1.3986, + "step": 1794 + }, + { + "epoch": 1.8438623523369286, + "grad_norm": 2.3268849849700928, + "learning_rate": 2.169775409212029e-06, + "loss": 1.3923, + "step": 1795 + }, + { + "epoch": 1.844889573703133, + "grad_norm": 2.424268960952759, + "learning_rate": 2.1678720974495622e-06, + "loss": 1.4046, + "step": 1796 + }, + { + "epoch": 1.8459167950693374, + "grad_norm": 2.3144543170928955, + "learning_rate": 2.1659687856870956e-06, + "loss": 1.391, + "step": 1797 + }, + { + "epoch": 1.846944016435542, + "grad_norm": 2.4157726764678955, + "learning_rate": 2.164065473924629e-06, + "loss": 1.4084, + "step": 1798 + }, + { + "epoch": 1.8479712378017463, + "grad_norm": 2.3993160724639893, + "learning_rate": 2.1621621621621623e-06, + "loss": 1.3372, + "step": 1799 + }, + { + "epoch": 1.8489984591679507, + "grad_norm": 2.3267669677734375, + "learning_rate": 2.1602588503996957e-06, + "loss": 1.3899, + "step": 1800 + }, + { + "epoch": 1.8489984591679507, + "eval_loss": 1.7611773014068604, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 1800 + }, + { + "epoch": 1.850025680534155, + "grad_norm": 2.488994836807251, + "learning_rate": 2.158355538637229e-06, + "loss": 1.3647, + "step": 1801 + }, + { + "epoch": 1.8510529019003594, + "grad_norm": 2.32352876663208, + "learning_rate": 2.1564522268747624e-06, + "loss": 1.3615, + "step": 1802 + }, + { + "epoch": 1.852080123266564, + "grad_norm": 2.667987585067749, + "learning_rate": 2.1545489151122958e-06, + "loss": 1.4388, + "step": 1803 + }, + { + "epoch": 1.8531073446327684, + "grad_norm": 2.4570565223693848, + "learning_rate": 2.1526456033498287e-06, + "loss": 1.3939, + "step": 1804 + }, + { + "epoch": 1.8541345659989728, + "grad_norm": 2.4037978649139404, + "learning_rate": 2.150742291587362e-06, + "loss": 1.3834, + "step": 1805 + }, + { + "epoch": 1.8551617873651773, + "grad_norm": 2.42080020904541, + "learning_rate": 2.1488389798248954e-06, + "loss": 1.3758, + "step": 1806 + }, + { + "epoch": 1.8561890087313815, + "grad_norm": 2.3361966609954834, + "learning_rate": 2.146935668062429e-06, + "loss": 1.4024, + "step": 1807 + }, + { + "epoch": 1.857216230097586, + "grad_norm": 2.251089334487915, + "learning_rate": 2.145032356299962e-06, + "loss": 1.3365, + "step": 1808 + }, + { + "epoch": 1.8582434514637904, + "grad_norm": 2.4046289920806885, + "learning_rate": 2.1431290445374955e-06, + "loss": 1.3831, + "step": 1809 + }, + { + "epoch": 1.8592706728299948, + "grad_norm": 2.2613766193389893, + "learning_rate": 2.141225732775029e-06, + "loss": 1.3423, + "step": 1810 + }, + { + "epoch": 1.8592706728299948, + "eval_loss": 1.752854347229004, + "eval_runtime": 37.6469, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 1810 + }, + { + "epoch": 1.8602978941961994, + "grad_norm": 2.377535581588745, + "learning_rate": 2.1393224210125623e-06, + "loss": 1.3796, + "step": 1811 + }, + { + "epoch": 1.8613251155624035, + "grad_norm": 2.3959462642669678, + "learning_rate": 2.137419109250095e-06, + "loss": 1.3664, + "step": 1812 + }, + { + "epoch": 1.8623523369286081, + "grad_norm": 2.5317859649658203, + "learning_rate": 2.1355157974876286e-06, + "loss": 1.418, + "step": 1813 + }, + { + "epoch": 1.8633795582948125, + "grad_norm": 2.5266010761260986, + "learning_rate": 2.133612485725162e-06, + "loss": 1.4117, + "step": 1814 + }, + { + "epoch": 1.8644067796610169, + "grad_norm": 2.3911185264587402, + "learning_rate": 2.1317091739626953e-06, + "loss": 1.3206, + "step": 1815 + }, + { + "epoch": 1.8654340010272215, + "grad_norm": 2.358596086502075, + "learning_rate": 2.1298058622002286e-06, + "loss": 1.3475, + "step": 1816 + }, + { + "epoch": 1.8664612223934258, + "grad_norm": 2.348175048828125, + "learning_rate": 2.1279025504377616e-06, + "loss": 1.3769, + "step": 1817 + }, + { + "epoch": 1.8674884437596302, + "grad_norm": 2.417470932006836, + "learning_rate": 2.125999238675295e-06, + "loss": 1.3667, + "step": 1818 + }, + { + "epoch": 1.8685156651258348, + "grad_norm": 2.5871050357818604, + "learning_rate": 2.1240959269128287e-06, + "loss": 1.4139, + "step": 1819 + }, + { + "epoch": 1.869542886492039, + "grad_norm": 2.384218692779541, + "learning_rate": 2.122192615150362e-06, + "loss": 1.3755, + "step": 1820 + }, + { + "epoch": 1.869542886492039, + "eval_loss": 1.7497111558914185, + "eval_runtime": 37.7655, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 1820 + }, + { + "epoch": 1.8705701078582435, + "grad_norm": 2.354877233505249, + "learning_rate": 2.120289303387895e-06, + "loss": 1.3174, + "step": 1821 + }, + { + "epoch": 1.8715973292244479, + "grad_norm": 2.4020509719848633, + "learning_rate": 2.1183859916254284e-06, + "loss": 1.3699, + "step": 1822 + }, + { + "epoch": 1.8726245505906522, + "grad_norm": 2.2876627445220947, + "learning_rate": 2.1164826798629618e-06, + "loss": 1.4053, + "step": 1823 + }, + { + "epoch": 1.8736517719568568, + "grad_norm": 2.2526464462280273, + "learning_rate": 2.114579368100495e-06, + "loss": 1.3569, + "step": 1824 + }, + { + "epoch": 1.874678993323061, + "grad_norm": 2.340773344039917, + "learning_rate": 2.1126760563380285e-06, + "loss": 1.3439, + "step": 1825 + }, + { + "epoch": 1.8757062146892656, + "grad_norm": 2.315812110900879, + "learning_rate": 2.1107727445755614e-06, + "loss": 1.3129, + "step": 1826 + }, + { + "epoch": 1.87673343605547, + "grad_norm": 2.480487108230591, + "learning_rate": 2.1088694328130948e-06, + "loss": 1.4372, + "step": 1827 + }, + { + "epoch": 1.8777606574216743, + "grad_norm": 2.267662286758423, + "learning_rate": 2.106966121050628e-06, + "loss": 1.3429, + "step": 1828 + }, + { + "epoch": 1.878787878787879, + "grad_norm": 2.383023977279663, + "learning_rate": 2.1050628092881615e-06, + "loss": 1.3963, + "step": 1829 + }, + { + "epoch": 1.879815100154083, + "grad_norm": 2.4784996509552, + "learning_rate": 2.103159497525695e-06, + "loss": 1.4405, + "step": 1830 + }, + { + "epoch": 1.879815100154083, + "eval_loss": 1.7609645128250122, + "eval_runtime": 37.904, + "eval_samples_per_second": 26.382, + "eval_steps_per_second": 1.108, + "step": 1830 + }, + { + "epoch": 1.8808423215202876, + "grad_norm": 2.359614133834839, + "learning_rate": 2.1012561857632282e-06, + "loss": 1.3577, + "step": 1831 + }, + { + "epoch": 1.881869542886492, + "grad_norm": 2.452862501144409, + "learning_rate": 2.0993528740007616e-06, + "loss": 1.4082, + "step": 1832 + }, + { + "epoch": 1.8828967642526964, + "grad_norm": 2.384011745452881, + "learning_rate": 2.097449562238295e-06, + "loss": 1.4175, + "step": 1833 + }, + { + "epoch": 1.883923985618901, + "grad_norm": 2.4906508922576904, + "learning_rate": 2.095546250475828e-06, + "loss": 1.3693, + "step": 1834 + }, + { + "epoch": 1.8849512069851053, + "grad_norm": 2.510040521621704, + "learning_rate": 2.0936429387133613e-06, + "loss": 1.3703, + "step": 1835 + }, + { + "epoch": 1.8859784283513097, + "grad_norm": 2.3438405990600586, + "learning_rate": 2.0917396269508946e-06, + "loss": 1.4311, + "step": 1836 + }, + { + "epoch": 1.8870056497175143, + "grad_norm": 2.532226800918579, + "learning_rate": 2.089836315188428e-06, + "loss": 1.3986, + "step": 1837 + }, + { + "epoch": 1.8880328710837184, + "grad_norm": 2.4368748664855957, + "learning_rate": 2.0879330034259614e-06, + "loss": 1.4025, + "step": 1838 + }, + { + "epoch": 1.889060092449923, + "grad_norm": 2.4487054347991943, + "learning_rate": 2.0860296916634947e-06, + "loss": 1.3562, + "step": 1839 + }, + { + "epoch": 1.8900873138161274, + "grad_norm": 2.6636273860931396, + "learning_rate": 2.084126379901028e-06, + "loss": 1.4384, + "step": 1840 + }, + { + "epoch": 1.8900873138161274, + "eval_loss": 1.7675000429153442, + "eval_runtime": 37.9371, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 1840 + }, + { + "epoch": 1.8911145351823317, + "grad_norm": 2.497738838195801, + "learning_rate": 2.0822230681385614e-06, + "loss": 1.3462, + "step": 1841 + }, + { + "epoch": 1.8921417565485363, + "grad_norm": 2.4419288635253906, + "learning_rate": 2.080319756376095e-06, + "loss": 1.3974, + "step": 1842 + }, + { + "epoch": 1.8931689779147405, + "grad_norm": 2.3671085834503174, + "learning_rate": 2.0784164446136277e-06, + "loss": 1.3917, + "step": 1843 + }, + { + "epoch": 1.894196199280945, + "grad_norm": 2.3837625980377197, + "learning_rate": 2.076513132851161e-06, + "loss": 1.3663, + "step": 1844 + }, + { + "epoch": 1.8952234206471494, + "grad_norm": 2.590935468673706, + "learning_rate": 2.0746098210886945e-06, + "loss": 1.3962, + "step": 1845 + }, + { + "epoch": 1.8962506420133538, + "grad_norm": 2.2917118072509766, + "learning_rate": 2.072706509326228e-06, + "loss": 1.3641, + "step": 1846 + }, + { + "epoch": 1.8972778633795584, + "grad_norm": 2.4366025924682617, + "learning_rate": 2.070803197563761e-06, + "loss": 1.3749, + "step": 1847 + }, + { + "epoch": 1.8983050847457628, + "grad_norm": 2.385152816772461, + "learning_rate": 2.068899885801294e-06, + "loss": 1.3655, + "step": 1848 + }, + { + "epoch": 1.8993323061119671, + "grad_norm": 2.473550796508789, + "learning_rate": 2.066996574038828e-06, + "loss": 1.3917, + "step": 1849 + }, + { + "epoch": 1.9003595274781715, + "grad_norm": 2.360003709793091, + "learning_rate": 2.0650932622763613e-06, + "loss": 1.4089, + "step": 1850 + }, + { + "epoch": 1.9003595274781715, + "eval_loss": 1.7553730010986328, + "eval_runtime": 37.9412, + "eval_samples_per_second": 26.357, + "eval_steps_per_second": 1.107, + "step": 1850 + }, + { + "epoch": 1.9013867488443759, + "grad_norm": 2.3891544342041016, + "learning_rate": 2.0631899505138942e-06, + "loss": 1.3033, + "step": 1851 + }, + { + "epoch": 1.9024139702105805, + "grad_norm": 2.330413341522217, + "learning_rate": 2.0612866387514276e-06, + "loss": 1.3446, + "step": 1852 + }, + { + "epoch": 1.9034411915767848, + "grad_norm": 2.3061635494232178, + "learning_rate": 2.059383326988961e-06, + "loss": 1.3872, + "step": 1853 + }, + { + "epoch": 1.9044684129429892, + "grad_norm": 2.4351041316986084, + "learning_rate": 2.0574800152264943e-06, + "loss": 1.362, + "step": 1854 + }, + { + "epoch": 1.9054956343091938, + "grad_norm": 2.447535991668701, + "learning_rate": 2.0555767034640277e-06, + "loss": 1.3417, + "step": 1855 + }, + { + "epoch": 1.906522855675398, + "grad_norm": 2.369133472442627, + "learning_rate": 2.053673391701561e-06, + "loss": 1.3181, + "step": 1856 + }, + { + "epoch": 1.9075500770416025, + "grad_norm": 2.3961751461029053, + "learning_rate": 2.051770079939094e-06, + "loss": 1.3913, + "step": 1857 + }, + { + "epoch": 1.9085772984078069, + "grad_norm": 2.40244460105896, + "learning_rate": 2.0498667681766273e-06, + "loss": 1.3447, + "step": 1858 + }, + { + "epoch": 1.9096045197740112, + "grad_norm": 2.4332613945007324, + "learning_rate": 2.047963456414161e-06, + "loss": 1.3427, + "step": 1859 + }, + { + "epoch": 1.9106317411402158, + "grad_norm": 2.3410897254943848, + "learning_rate": 2.046060144651694e-06, + "loss": 1.3697, + "step": 1860 + }, + { + "epoch": 1.9106317411402158, + "eval_loss": 1.74896240234375, + "eval_runtime": 37.9315, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 1860 + }, + { + "epoch": 1.91165896250642, + "grad_norm": 2.3965389728546143, + "learning_rate": 2.0441568328892274e-06, + "loss": 1.3752, + "step": 1861 + }, + { + "epoch": 1.9126861838726246, + "grad_norm": 2.3173274993896484, + "learning_rate": 2.0422535211267608e-06, + "loss": 1.3807, + "step": 1862 + }, + { + "epoch": 1.913713405238829, + "grad_norm": 2.2883224487304688, + "learning_rate": 2.040350209364294e-06, + "loss": 1.3551, + "step": 1863 + }, + { + "epoch": 1.9147406266050333, + "grad_norm": 2.2770445346832275, + "learning_rate": 2.0384468976018275e-06, + "loss": 1.3405, + "step": 1864 + }, + { + "epoch": 1.915767847971238, + "grad_norm": 2.3694701194763184, + "learning_rate": 2.0365435858393604e-06, + "loss": 1.3538, + "step": 1865 + }, + { + "epoch": 1.9167950693374423, + "grad_norm": 2.4186787605285645, + "learning_rate": 2.034640274076894e-06, + "loss": 1.353, + "step": 1866 + }, + { + "epoch": 1.9178222907036466, + "grad_norm": 2.4059510231018066, + "learning_rate": 2.032736962314427e-06, + "loss": 1.4149, + "step": 1867 + }, + { + "epoch": 1.9188495120698512, + "grad_norm": 2.5529062747955322, + "learning_rate": 2.0308336505519605e-06, + "loss": 1.3669, + "step": 1868 + }, + { + "epoch": 1.9198767334360554, + "grad_norm": 2.3925650119781494, + "learning_rate": 2.028930338789494e-06, + "loss": 1.4178, + "step": 1869 + }, + { + "epoch": 1.92090395480226, + "grad_norm": 2.5203816890716553, + "learning_rate": 2.0270270270270273e-06, + "loss": 1.3418, + "step": 1870 + }, + { + "epoch": 1.92090395480226, + "eval_loss": 1.7473065853118896, + "eval_runtime": 37.9634, + "eval_samples_per_second": 26.341, + "eval_steps_per_second": 1.106, + "step": 1870 + }, + { + "epoch": 1.9219311761684643, + "grad_norm": 2.4792747497558594, + "learning_rate": 2.0251237152645606e-06, + "loss": 1.3611, + "step": 1871 + }, + { + "epoch": 1.9229583975346687, + "grad_norm": 2.357645273208618, + "learning_rate": 2.023220403502094e-06, + "loss": 1.3433, + "step": 1872 + }, + { + "epoch": 1.9239856189008733, + "grad_norm": 2.284149408340454, + "learning_rate": 2.0213170917396273e-06, + "loss": 1.3157, + "step": 1873 + }, + { + "epoch": 1.9250128402670774, + "grad_norm": 2.3971638679504395, + "learning_rate": 2.0194137799771603e-06, + "loss": 1.3435, + "step": 1874 + }, + { + "epoch": 1.926040061633282, + "grad_norm": 2.605325222015381, + "learning_rate": 2.0175104682146937e-06, + "loss": 1.3619, + "step": 1875 + }, + { + "epoch": 1.9270672829994864, + "grad_norm": 2.3957643508911133, + "learning_rate": 2.015607156452227e-06, + "loss": 1.3078, + "step": 1876 + }, + { + "epoch": 1.9280945043656907, + "grad_norm": 2.3887429237365723, + "learning_rate": 2.0137038446897604e-06, + "loss": 1.3447, + "step": 1877 + }, + { + "epoch": 1.9291217257318953, + "grad_norm": 2.31253719329834, + "learning_rate": 2.0118005329272937e-06, + "loss": 1.3467, + "step": 1878 + }, + { + "epoch": 1.9301489470980995, + "grad_norm": 2.3084523677825928, + "learning_rate": 2.0098972211648267e-06, + "loss": 1.3466, + "step": 1879 + }, + { + "epoch": 1.931176168464304, + "grad_norm": 2.3024210929870605, + "learning_rate": 2.0079939094023605e-06, + "loss": 1.3525, + "step": 1880 + }, + { + "epoch": 1.931176168464304, + "eval_loss": 1.7498610019683838, + "eval_runtime": 37.9363, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 1880 + }, + { + "epoch": 1.9322033898305084, + "grad_norm": 2.342660665512085, + "learning_rate": 2.006090597639894e-06, + "loss": 1.4013, + "step": 1881 + }, + { + "epoch": 1.9332306111967128, + "grad_norm": 2.4165151119232178, + "learning_rate": 2.0041872858774268e-06, + "loss": 1.4154, + "step": 1882 + }, + { + "epoch": 1.9342578325629174, + "grad_norm": 2.3674707412719727, + "learning_rate": 2.00228397411496e-06, + "loss": 1.3903, + "step": 1883 + }, + { + "epoch": 1.9352850539291218, + "grad_norm": 2.2478187084198, + "learning_rate": 2.0003806623524935e-06, + "loss": 1.3708, + "step": 1884 + }, + { + "epoch": 1.9363122752953261, + "grad_norm": 2.328073740005493, + "learning_rate": 1.998477350590027e-06, + "loss": 1.3013, + "step": 1885 + }, + { + "epoch": 1.9373394966615307, + "grad_norm": 2.4772889614105225, + "learning_rate": 1.9965740388275602e-06, + "loss": 1.3818, + "step": 1886 + }, + { + "epoch": 1.9383667180277349, + "grad_norm": 2.456451892852783, + "learning_rate": 1.994670727065093e-06, + "loss": 1.3806, + "step": 1887 + }, + { + "epoch": 1.9393939393939394, + "grad_norm": 2.325561046600342, + "learning_rate": 1.9927674153026265e-06, + "loss": 1.3262, + "step": 1888 + }, + { + "epoch": 1.9404211607601438, + "grad_norm": 2.447673797607422, + "learning_rate": 1.99086410354016e-06, + "loss": 1.3776, + "step": 1889 + }, + { + "epoch": 1.9414483821263482, + "grad_norm": 2.4231903553009033, + "learning_rate": 1.9889607917776937e-06, + "loss": 1.3593, + "step": 1890 + }, + { + "epoch": 1.9414483821263482, + "eval_loss": 1.7612230777740479, + "eval_runtime": 37.9427, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 1890 + }, + { + "epoch": 1.9424756034925528, + "grad_norm": 2.5399789810180664, + "learning_rate": 1.9870574800152266e-06, + "loss": 1.3737, + "step": 1891 + }, + { + "epoch": 1.943502824858757, + "grad_norm": 2.3155572414398193, + "learning_rate": 1.98515416825276e-06, + "loss": 1.388, + "step": 1892 + }, + { + "epoch": 1.9445300462249615, + "grad_norm": 2.33693265914917, + "learning_rate": 1.9832508564902933e-06, + "loss": 1.3734, + "step": 1893 + }, + { + "epoch": 1.9455572675911659, + "grad_norm": 2.2426419258117676, + "learning_rate": 1.9813475447278267e-06, + "loss": 1.3011, + "step": 1894 + }, + { + "epoch": 1.9465844889573702, + "grad_norm": 2.4306788444519043, + "learning_rate": 1.97944423296536e-06, + "loss": 1.3721, + "step": 1895 + }, + { + "epoch": 1.9476117103235748, + "grad_norm": 2.329252004623413, + "learning_rate": 1.977540921202893e-06, + "loss": 1.363, + "step": 1896 + }, + { + "epoch": 1.9486389316897792, + "grad_norm": 2.3523757457733154, + "learning_rate": 1.9756376094404264e-06, + "loss": 1.3485, + "step": 1897 + }, + { + "epoch": 1.9496661530559836, + "grad_norm": 2.3240344524383545, + "learning_rate": 1.9737342976779597e-06, + "loss": 1.3258, + "step": 1898 + }, + { + "epoch": 1.950693374422188, + "grad_norm": 2.4068586826324463, + "learning_rate": 1.971830985915493e-06, + "loss": 1.406, + "step": 1899 + }, + { + "epoch": 1.9517205957883923, + "grad_norm": 2.385835886001587, + "learning_rate": 1.9699276741530264e-06, + "loss": 1.3641, + "step": 1900 + }, + { + "epoch": 1.9517205957883923, + "eval_loss": 1.747411847114563, + "eval_runtime": 37.9419, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 1900 + }, + { + "epoch": 1.9527478171545969, + "grad_norm": 2.3512096405029297, + "learning_rate": 1.96802436239056e-06, + "loss": 1.3647, + "step": 1901 + }, + { + "epoch": 1.9537750385208013, + "grad_norm": 2.3813812732696533, + "learning_rate": 1.966121050628093e-06, + "loss": 1.3422, + "step": 1902 + }, + { + "epoch": 1.9548022598870056, + "grad_norm": 2.404499053955078, + "learning_rate": 1.9642177388656265e-06, + "loss": 1.3918, + "step": 1903 + }, + { + "epoch": 1.9558294812532102, + "grad_norm": 2.4429967403411865, + "learning_rate": 1.9623144271031595e-06, + "loss": 1.4049, + "step": 1904 + }, + { + "epoch": 1.9568567026194144, + "grad_norm": 2.505833864212036, + "learning_rate": 1.960411115340693e-06, + "loss": 1.4112, + "step": 1905 + }, + { + "epoch": 1.957883923985619, + "grad_norm": 2.4441304206848145, + "learning_rate": 1.958507803578226e-06, + "loss": 1.3648, + "step": 1906 + }, + { + "epoch": 1.9589111453518233, + "grad_norm": 2.4598183631896973, + "learning_rate": 1.9566044918157596e-06, + "loss": 1.3879, + "step": 1907 + }, + { + "epoch": 1.9599383667180277, + "grad_norm": 2.3862075805664062, + "learning_rate": 1.954701180053293e-06, + "loss": 1.3831, + "step": 1908 + }, + { + "epoch": 1.9609655880842323, + "grad_norm": 2.434634208679199, + "learning_rate": 1.952797868290826e-06, + "loss": 1.3499, + "step": 1909 + }, + { + "epoch": 1.9619928094504364, + "grad_norm": 2.4674196243286133, + "learning_rate": 1.9508945565283596e-06, + "loss": 1.3941, + "step": 1910 + }, + { + "epoch": 1.9619928094504364, + "eval_loss": 1.734555959701538, + "eval_runtime": 37.6492, + "eval_samples_per_second": 26.561, + "eval_steps_per_second": 1.116, + "step": 1910 + }, + { + "epoch": 1.963020030816641, + "grad_norm": 2.336597442626953, + "learning_rate": 1.948991244765893e-06, + "loss": 1.3731, + "step": 1911 + }, + { + "epoch": 1.9640472521828454, + "grad_norm": 2.4431259632110596, + "learning_rate": 1.9470879330034264e-06, + "loss": 1.3926, + "step": 1912 + }, + { + "epoch": 1.9650744735490497, + "grad_norm": 2.4182162284851074, + "learning_rate": 1.9451846212409593e-06, + "loss": 1.3291, + "step": 1913 + }, + { + "epoch": 1.9661016949152543, + "grad_norm": 2.382519483566284, + "learning_rate": 1.9432813094784927e-06, + "loss": 1.3891, + "step": 1914 + }, + { + "epoch": 1.9671289162814587, + "grad_norm": 2.3186089992523193, + "learning_rate": 1.941377997716026e-06, + "loss": 1.3206, + "step": 1915 + }, + { + "epoch": 1.968156137647663, + "grad_norm": 2.32810378074646, + "learning_rate": 1.9394746859535594e-06, + "loss": 1.3564, + "step": 1916 + }, + { + "epoch": 1.9691833590138677, + "grad_norm": 2.3164830207824707, + "learning_rate": 1.9375713741910928e-06, + "loss": 1.3154, + "step": 1917 + }, + { + "epoch": 1.9702105803800718, + "grad_norm": 2.3495121002197266, + "learning_rate": 1.9356680624286257e-06, + "loss": 1.3533, + "step": 1918 + }, + { + "epoch": 1.9712378017462764, + "grad_norm": 2.307860851287842, + "learning_rate": 1.933764750666159e-06, + "loss": 1.3142, + "step": 1919 + }, + { + "epoch": 1.9722650231124808, + "grad_norm": 2.2913570404052734, + "learning_rate": 1.931861438903693e-06, + "loss": 1.3672, + "step": 1920 + }, + { + "epoch": 1.9722650231124808, + "eval_loss": 1.7303361892700195, + "eval_runtime": 37.7597, + "eval_samples_per_second": 26.483, + "eval_steps_per_second": 1.112, + "step": 1920 + }, + { + "epoch": 1.9732922444786851, + "grad_norm": 2.3222689628601074, + "learning_rate": 1.929958127141226e-06, + "loss": 1.3317, + "step": 1921 + }, + { + "epoch": 1.9743194658448897, + "grad_norm": 2.2708919048309326, + "learning_rate": 1.928054815378759e-06, + "loss": 1.3518, + "step": 1922 + }, + { + "epoch": 1.9753466872110939, + "grad_norm": 2.4408884048461914, + "learning_rate": 1.9261515036162925e-06, + "loss": 1.4332, + "step": 1923 + }, + { + "epoch": 1.9763739085772984, + "grad_norm": 2.469506025314331, + "learning_rate": 1.924248191853826e-06, + "loss": 1.3927, + "step": 1924 + }, + { + "epoch": 1.9774011299435028, + "grad_norm": 2.3409810066223145, + "learning_rate": 1.9223448800913592e-06, + "loss": 1.3692, + "step": 1925 + }, + { + "epoch": 1.9784283513097072, + "grad_norm": 2.293306589126587, + "learning_rate": 1.9204415683288926e-06, + "loss": 1.3491, + "step": 1926 + }, + { + "epoch": 1.9794555726759118, + "grad_norm": 2.428431272506714, + "learning_rate": 1.9185382565664255e-06, + "loss": 1.381, + "step": 1927 + }, + { + "epoch": 1.980482794042116, + "grad_norm": 2.4767544269561768, + "learning_rate": 1.916634944803959e-06, + "loss": 1.3953, + "step": 1928 + }, + { + "epoch": 1.9815100154083205, + "grad_norm": 2.44821834564209, + "learning_rate": 1.9147316330414923e-06, + "loss": 1.3746, + "step": 1929 + }, + { + "epoch": 1.9825372367745249, + "grad_norm": 2.349822759628296, + "learning_rate": 1.9128283212790256e-06, + "loss": 1.3619, + "step": 1930 + }, + { + "epoch": 1.9825372367745249, + "eval_loss": 1.7362234592437744, + "eval_runtime": 37.8754, + "eval_samples_per_second": 26.402, + "eval_steps_per_second": 1.109, + "step": 1930 + }, + { + "epoch": 1.9835644581407292, + "grad_norm": 2.5379276275634766, + "learning_rate": 1.910925009516559e-06, + "loss": 1.4051, + "step": 1931 + }, + { + "epoch": 1.9845916795069338, + "grad_norm": 2.329698324203491, + "learning_rate": 1.9090216977540924e-06, + "loss": 1.3693, + "step": 1932 + }, + { + "epoch": 1.9856189008731382, + "grad_norm": 2.424672842025757, + "learning_rate": 1.9071183859916257e-06, + "loss": 1.4073, + "step": 1933 + }, + { + "epoch": 1.9866461222393426, + "grad_norm": 2.480088949203491, + "learning_rate": 1.9052150742291589e-06, + "loss": 1.3816, + "step": 1934 + }, + { + "epoch": 1.9876733436055471, + "grad_norm": 2.4967830181121826, + "learning_rate": 1.9033117624666922e-06, + "loss": 1.3804, + "step": 1935 + }, + { + "epoch": 1.9887005649717513, + "grad_norm": 2.337944984436035, + "learning_rate": 1.9014084507042254e-06, + "loss": 1.3452, + "step": 1936 + }, + { + "epoch": 1.9897277863379559, + "grad_norm": 2.4424421787261963, + "learning_rate": 1.8995051389417587e-06, + "loss": 1.3382, + "step": 1937 + }, + { + "epoch": 1.9907550077041603, + "grad_norm": 2.2587194442749023, + "learning_rate": 1.8976018271792921e-06, + "loss": 1.3229, + "step": 1938 + }, + { + "epoch": 1.9917822290703646, + "grad_norm": 2.4167850017547607, + "learning_rate": 1.8956985154168253e-06, + "loss": 1.3828, + "step": 1939 + }, + { + "epoch": 1.9928094504365692, + "grad_norm": 2.397630214691162, + "learning_rate": 1.8937952036543588e-06, + "loss": 1.3499, + "step": 1940 + }, + { + "epoch": 1.9928094504365692, + "eval_loss": 1.7392780780792236, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 1940 + }, + { + "epoch": 1.9938366718027734, + "grad_norm": 2.3320090770721436, + "learning_rate": 1.8918918918918922e-06, + "loss": 1.3157, + "step": 1941 + }, + { + "epoch": 1.994863893168978, + "grad_norm": 2.4160780906677246, + "learning_rate": 1.8899885801294253e-06, + "loss": 1.3693, + "step": 1942 + }, + { + "epoch": 1.9958911145351823, + "grad_norm": 2.4204070568084717, + "learning_rate": 1.8880852683669587e-06, + "loss": 1.3787, + "step": 1943 + }, + { + "epoch": 1.9969183359013867, + "grad_norm": 2.3747220039367676, + "learning_rate": 1.886181956604492e-06, + "loss": 1.3449, + "step": 1944 + }, + { + "epoch": 1.9979455572675913, + "grad_norm": 2.454822063446045, + "learning_rate": 1.8842786448420252e-06, + "loss": 1.3478, + "step": 1945 + }, + { + "epoch": 1.9989727786337956, + "grad_norm": 2.3999180793762207, + "learning_rate": 1.8823753330795586e-06, + "loss": 1.3786, + "step": 1946 + }, + { + "epoch": 2.0, + "grad_norm": 2.295583963394165, + "learning_rate": 1.8804720213170917e-06, + "loss": 1.2788, + "step": 1947 + }, + { + "epoch": 2.0010272213662046, + "grad_norm": 2.5762782096862793, + "learning_rate": 1.878568709554625e-06, + "loss": 1.2981, + "step": 1948 + }, + { + "epoch": 2.0020544427324087, + "grad_norm": 2.7369227409362793, + "learning_rate": 1.8766653977921585e-06, + "loss": 1.3315, + "step": 1949 + }, + { + "epoch": 2.0030816640986133, + "grad_norm": 2.5014007091522217, + "learning_rate": 1.874762086029692e-06, + "loss": 1.3112, + "step": 1950 + }, + { + "epoch": 2.0030816640986133, + "eval_loss": 1.7656217813491821, + "eval_runtime": 37.9186, + "eval_samples_per_second": 26.372, + "eval_steps_per_second": 1.108, + "step": 1950 + }, + { + "epoch": 2.0041088854648175, + "grad_norm": 2.7794134616851807, + "learning_rate": 1.8728587742672252e-06, + "loss": 1.3328, + "step": 1951 + }, + { + "epoch": 2.005136106831022, + "grad_norm": 2.9862377643585205, + "learning_rate": 1.8709554625047585e-06, + "loss": 1.3467, + "step": 1952 + }, + { + "epoch": 2.0061633281972266, + "grad_norm": 2.8717260360717773, + "learning_rate": 1.8690521507422917e-06, + "loss": 1.2869, + "step": 1953 + }, + { + "epoch": 2.007190549563431, + "grad_norm": 2.6782584190368652, + "learning_rate": 1.867148838979825e-06, + "loss": 1.3081, + "step": 1954 + }, + { + "epoch": 2.0082177709296354, + "grad_norm": 2.804760456085205, + "learning_rate": 1.8652455272173584e-06, + "loss": 1.3089, + "step": 1955 + }, + { + "epoch": 2.00924499229584, + "grad_norm": 2.8827545642852783, + "learning_rate": 1.8633422154548916e-06, + "loss": 1.352, + "step": 1956 + }, + { + "epoch": 2.010272213662044, + "grad_norm": 2.6346447467803955, + "learning_rate": 1.861438903692425e-06, + "loss": 1.2582, + "step": 1957 + }, + { + "epoch": 2.0112994350282487, + "grad_norm": 2.690293073654175, + "learning_rate": 1.859535591929958e-06, + "loss": 1.2807, + "step": 1958 + }, + { + "epoch": 2.012326656394453, + "grad_norm": 2.8933660984039307, + "learning_rate": 1.8576322801674915e-06, + "loss": 1.3553, + "step": 1959 + }, + { + "epoch": 2.0133538777606574, + "grad_norm": 2.6514511108398438, + "learning_rate": 1.855728968405025e-06, + "loss": 1.2971, + "step": 1960 + }, + { + "epoch": 2.0133538777606574, + "eval_loss": 1.7628012895584106, + "eval_runtime": 37.9132, + "eval_samples_per_second": 26.376, + "eval_steps_per_second": 1.108, + "step": 1960 + }, + { + "epoch": 2.014381099126862, + "grad_norm": 2.8454601764678955, + "learning_rate": 1.8538256566425584e-06, + "loss": 1.3132, + "step": 1961 + }, + { + "epoch": 2.015408320493066, + "grad_norm": 2.715981960296631, + "learning_rate": 1.8519223448800915e-06, + "loss": 1.3503, + "step": 1962 + }, + { + "epoch": 2.0164355418592708, + "grad_norm": 2.688493251800537, + "learning_rate": 1.850019033117625e-06, + "loss": 1.2966, + "step": 1963 + }, + { + "epoch": 2.017462763225475, + "grad_norm": 2.6956987380981445, + "learning_rate": 1.848115721355158e-06, + "loss": 1.2969, + "step": 1964 + }, + { + "epoch": 2.0184899845916795, + "grad_norm": 2.697744607925415, + "learning_rate": 1.8462124095926914e-06, + "loss": 1.3207, + "step": 1965 + }, + { + "epoch": 2.019517205957884, + "grad_norm": 2.7628865242004395, + "learning_rate": 1.8443090978302248e-06, + "loss": 1.3311, + "step": 1966 + }, + { + "epoch": 2.0205444273240882, + "grad_norm": 2.801499366760254, + "learning_rate": 1.842405786067758e-06, + "loss": 1.3411, + "step": 1967 + }, + { + "epoch": 2.021571648690293, + "grad_norm": 2.6009418964385986, + "learning_rate": 1.8405024743052913e-06, + "loss": 1.2985, + "step": 1968 + }, + { + "epoch": 2.022598870056497, + "grad_norm": 2.772385358810425, + "learning_rate": 1.8385991625428247e-06, + "loss": 1.3407, + "step": 1969 + }, + { + "epoch": 2.0236260914227016, + "grad_norm": 2.558302164077759, + "learning_rate": 1.8366958507803578e-06, + "loss": 1.2814, + "step": 1970 + }, + { + "epoch": 2.0236260914227016, + "eval_loss": 1.7421122789382935, + "eval_runtime": 37.9255, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 1970 + }, + { + "epoch": 2.024653312788906, + "grad_norm": 2.813371181488037, + "learning_rate": 1.8347925390178914e-06, + "loss": 1.3309, + "step": 1971 + }, + { + "epoch": 2.0256805341551103, + "grad_norm": 2.742785692214966, + "learning_rate": 1.8328892272554247e-06, + "loss": 1.2711, + "step": 1972 + }, + { + "epoch": 2.026707755521315, + "grad_norm": 2.889450788497925, + "learning_rate": 1.8309859154929579e-06, + "loss": 1.3309, + "step": 1973 + }, + { + "epoch": 2.0277349768875195, + "grad_norm": 2.919278860092163, + "learning_rate": 1.8290826037304913e-06, + "loss": 1.3168, + "step": 1974 + }, + { + "epoch": 2.0287621982537236, + "grad_norm": 2.6912894248962402, + "learning_rate": 1.8271792919680246e-06, + "loss": 1.29, + "step": 1975 + }, + { + "epoch": 2.029789419619928, + "grad_norm": 2.584214687347412, + "learning_rate": 1.8252759802055578e-06, + "loss": 1.2873, + "step": 1976 + }, + { + "epoch": 2.0308166409861323, + "grad_norm": 2.6445112228393555, + "learning_rate": 1.8233726684430911e-06, + "loss": 1.3335, + "step": 1977 + }, + { + "epoch": 2.031843862352337, + "grad_norm": 2.6441738605499268, + "learning_rate": 1.8214693566806243e-06, + "loss": 1.2795, + "step": 1978 + }, + { + "epoch": 2.0328710837185415, + "grad_norm": 2.7365334033966064, + "learning_rate": 1.8195660449181576e-06, + "loss": 1.3142, + "step": 1979 + }, + { + "epoch": 2.0338983050847457, + "grad_norm": 2.7914040088653564, + "learning_rate": 1.817662733155691e-06, + "loss": 1.3175, + "step": 1980 + }, + { + "epoch": 2.0338983050847457, + "eval_loss": 1.7680878639221191, + "eval_runtime": 37.9012, + "eval_samples_per_second": 26.384, + "eval_steps_per_second": 1.108, + "step": 1980 + }, + { + "epoch": 2.0349255264509503, + "grad_norm": 2.6883442401885986, + "learning_rate": 1.8157594213932244e-06, + "loss": 1.3157, + "step": 1981 + }, + { + "epoch": 2.0359527478171544, + "grad_norm": 2.7459399700164795, + "learning_rate": 1.8138561096307577e-06, + "loss": 1.33, + "step": 1982 + }, + { + "epoch": 2.036979969183359, + "grad_norm": 2.620150327682495, + "learning_rate": 1.811952797868291e-06, + "loss": 1.2664, + "step": 1983 + }, + { + "epoch": 2.0380071905495636, + "grad_norm": 2.6702964305877686, + "learning_rate": 1.8100494861058242e-06, + "loss": 1.3174, + "step": 1984 + }, + { + "epoch": 2.0390344119157677, + "grad_norm": 2.6688029766082764, + "learning_rate": 1.8081461743433576e-06, + "loss": 1.3224, + "step": 1985 + }, + { + "epoch": 2.0400616332819723, + "grad_norm": 2.7634220123291016, + "learning_rate": 1.806242862580891e-06, + "loss": 1.2804, + "step": 1986 + }, + { + "epoch": 2.0410888546481765, + "grad_norm": 2.8583552837371826, + "learning_rate": 1.8043395508184241e-06, + "loss": 1.3242, + "step": 1987 + }, + { + "epoch": 2.042116076014381, + "grad_norm": 2.776536703109741, + "learning_rate": 1.8024362390559575e-06, + "loss": 1.3101, + "step": 1988 + }, + { + "epoch": 2.0431432973805856, + "grad_norm": 2.8194644451141357, + "learning_rate": 1.8005329272934906e-06, + "loss": 1.2875, + "step": 1989 + }, + { + "epoch": 2.04417051874679, + "grad_norm": 3.063290596008301, + "learning_rate": 1.798629615531024e-06, + "loss": 1.365, + "step": 1990 + }, + { + "epoch": 2.04417051874679, + "eval_loss": 1.7815766334533691, + "eval_runtime": 37.9473, + "eval_samples_per_second": 26.352, + "eval_steps_per_second": 1.107, + "step": 1990 + }, + { + "epoch": 2.0451977401129944, + "grad_norm": 2.8954226970672607, + "learning_rate": 1.7967263037685576e-06, + "loss": 1.3109, + "step": 1991 + }, + { + "epoch": 2.046224961479199, + "grad_norm": 2.9319002628326416, + "learning_rate": 1.794822992006091e-06, + "loss": 1.2983, + "step": 1992 + }, + { + "epoch": 2.047252182845403, + "grad_norm": 2.920145034790039, + "learning_rate": 1.792919680243624e-06, + "loss": 1.3571, + "step": 1993 + }, + { + "epoch": 2.0482794042116077, + "grad_norm": 2.6473276615142822, + "learning_rate": 1.7910163684811574e-06, + "loss": 1.2752, + "step": 1994 + }, + { + "epoch": 2.049306625577812, + "grad_norm": 2.6453239917755127, + "learning_rate": 1.7891130567186906e-06, + "loss": 1.311, + "step": 1995 + }, + { + "epoch": 2.0503338469440164, + "grad_norm": 2.6123287677764893, + "learning_rate": 1.787209744956224e-06, + "loss": 1.2817, + "step": 1996 + }, + { + "epoch": 2.051361068310221, + "grad_norm": 2.6542904376983643, + "learning_rate": 1.7853064331937573e-06, + "loss": 1.3067, + "step": 1997 + }, + { + "epoch": 2.052388289676425, + "grad_norm": 2.799492835998535, + "learning_rate": 1.7834031214312905e-06, + "loss": 1.3436, + "step": 1998 + }, + { + "epoch": 2.0534155110426298, + "grad_norm": 2.6220905780792236, + "learning_rate": 1.7814998096688238e-06, + "loss": 1.2819, + "step": 1999 + }, + { + "epoch": 2.054442732408834, + "grad_norm": 2.691195487976074, + "learning_rate": 1.779596497906357e-06, + "loss": 1.3099, + "step": 2000 + }, + { + "epoch": 2.054442732408834, + "eval_loss": 1.770631194114685, + "eval_runtime": 37.9652, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 2000 + }, + { + "epoch": 2.0554699537750385, + "grad_norm": 2.8053042888641357, + "learning_rate": 1.7776931861438906e-06, + "loss": 1.3599, + "step": 2001 + }, + { + "epoch": 2.056497175141243, + "grad_norm": 2.784608840942383, + "learning_rate": 1.775789874381424e-06, + "loss": 1.3527, + "step": 2002 + }, + { + "epoch": 2.0575243965074472, + "grad_norm": 2.8933472633361816, + "learning_rate": 1.7738865626189573e-06, + "loss": 1.3363, + "step": 2003 + }, + { + "epoch": 2.058551617873652, + "grad_norm": 2.691600799560547, + "learning_rate": 1.7719832508564904e-06, + "loss": 1.2678, + "step": 2004 + }, + { + "epoch": 2.0595788392398564, + "grad_norm": 2.627678394317627, + "learning_rate": 1.7700799390940238e-06, + "loss": 1.3002, + "step": 2005 + }, + { + "epoch": 2.0606060606060606, + "grad_norm": 2.724674940109253, + "learning_rate": 1.768176627331557e-06, + "loss": 1.3237, + "step": 2006 + }, + { + "epoch": 2.061633281972265, + "grad_norm": 2.59698224067688, + "learning_rate": 1.7662733155690903e-06, + "loss": 1.3165, + "step": 2007 + }, + { + "epoch": 2.0626605033384693, + "grad_norm": 2.7686543464660645, + "learning_rate": 1.7643700038066237e-06, + "loss": 1.3478, + "step": 2008 + }, + { + "epoch": 2.063687724704674, + "grad_norm": 2.6561412811279297, + "learning_rate": 1.7624666920441568e-06, + "loss": 1.2851, + "step": 2009 + }, + { + "epoch": 2.0647149460708785, + "grad_norm": 2.722731351852417, + "learning_rate": 1.7605633802816902e-06, + "loss": 1.2726, + "step": 2010 + }, + { + "epoch": 2.0647149460708785, + "eval_loss": 1.7728410959243774, + "eval_runtime": 37.6541, + "eval_samples_per_second": 26.558, + "eval_steps_per_second": 1.115, + "step": 2010 + }, + { + "epoch": 2.0657421674370826, + "grad_norm": 2.910609483718872, + "learning_rate": 1.7586600685192238e-06, + "loss": 1.3603, + "step": 2011 + }, + { + "epoch": 2.066769388803287, + "grad_norm": 2.9910264015197754, + "learning_rate": 1.756756756756757e-06, + "loss": 1.3705, + "step": 2012 + }, + { + "epoch": 2.0677966101694913, + "grad_norm": 2.6477038860321045, + "learning_rate": 1.7548534449942903e-06, + "loss": 1.3372, + "step": 2013 + }, + { + "epoch": 2.068823831535696, + "grad_norm": 2.667626142501831, + "learning_rate": 1.7529501332318236e-06, + "loss": 1.3068, + "step": 2014 + }, + { + "epoch": 2.0698510529019005, + "grad_norm": 2.7895913124084473, + "learning_rate": 1.7510468214693568e-06, + "loss": 1.2813, + "step": 2015 + }, + { + "epoch": 2.0708782742681047, + "grad_norm": 2.5626792907714844, + "learning_rate": 1.7491435097068902e-06, + "loss": 1.2852, + "step": 2016 + }, + { + "epoch": 2.0719054956343093, + "grad_norm": 2.71805739402771, + "learning_rate": 1.7472401979444233e-06, + "loss": 1.3203, + "step": 2017 + }, + { + "epoch": 2.0729327170005134, + "grad_norm": 2.6200578212738037, + "learning_rate": 1.7453368861819567e-06, + "loss": 1.2617, + "step": 2018 + }, + { + "epoch": 2.073959938366718, + "grad_norm": 2.7212417125701904, + "learning_rate": 1.74343357441949e-06, + "loss": 1.3082, + "step": 2019 + }, + { + "epoch": 2.0749871597329226, + "grad_norm": 2.983325958251953, + "learning_rate": 1.7415302626570232e-06, + "loss": 1.3699, + "step": 2020 + }, + { + "epoch": 2.0749871597329226, + "eval_loss": 1.756274700164795, + "eval_runtime": 37.7628, + "eval_samples_per_second": 26.481, + "eval_steps_per_second": 1.112, + "step": 2020 + }, + { + "epoch": 2.0760143810991267, + "grad_norm": 2.6345975399017334, + "learning_rate": 1.7396269508945568e-06, + "loss": 1.3024, + "step": 2021 + }, + { + "epoch": 2.0770416024653313, + "grad_norm": 2.670807123184204, + "learning_rate": 1.7377236391320901e-06, + "loss": 1.2737, + "step": 2022 + }, + { + "epoch": 2.078068823831536, + "grad_norm": 2.788457155227661, + "learning_rate": 1.7358203273696233e-06, + "loss": 1.3459, + "step": 2023 + }, + { + "epoch": 2.07909604519774, + "grad_norm": 2.740338087081909, + "learning_rate": 1.7339170156071566e-06, + "loss": 1.3019, + "step": 2024 + }, + { + "epoch": 2.0801232665639446, + "grad_norm": 2.605855941772461, + "learning_rate": 1.73201370384469e-06, + "loss": 1.3096, + "step": 2025 + }, + { + "epoch": 2.081150487930149, + "grad_norm": 2.640096664428711, + "learning_rate": 1.7301103920822231e-06, + "loss": 1.3168, + "step": 2026 + }, + { + "epoch": 2.0821777092963534, + "grad_norm": 2.7602169513702393, + "learning_rate": 1.7282070803197565e-06, + "loss": 1.3121, + "step": 2027 + }, + { + "epoch": 2.083204930662558, + "grad_norm": 2.658534526824951, + "learning_rate": 1.7263037685572897e-06, + "loss": 1.3071, + "step": 2028 + }, + { + "epoch": 2.084232152028762, + "grad_norm": 2.7146220207214355, + "learning_rate": 1.724400456794823e-06, + "loss": 1.3336, + "step": 2029 + }, + { + "epoch": 2.0852593733949667, + "grad_norm": 2.694387197494507, + "learning_rate": 1.7224971450323564e-06, + "loss": 1.3013, + "step": 2030 + }, + { + "epoch": 2.0852593733949667, + "eval_loss": 1.7371397018432617, + "eval_runtime": 37.848, + "eval_samples_per_second": 26.421, + "eval_steps_per_second": 1.11, + "step": 2030 + }, + { + "epoch": 2.086286594761171, + "grad_norm": 2.7298598289489746, + "learning_rate": 1.72059383326989e-06, + "loss": 1.3048, + "step": 2031 + }, + { + "epoch": 2.0873138161273754, + "grad_norm": 2.9527862071990967, + "learning_rate": 1.7186905215074231e-06, + "loss": 1.3561, + "step": 2032 + }, + { + "epoch": 2.08834103749358, + "grad_norm": 2.7970757484436035, + "learning_rate": 1.7167872097449565e-06, + "loss": 1.2897, + "step": 2033 + }, + { + "epoch": 2.089368258859784, + "grad_norm": 2.7579336166381836, + "learning_rate": 1.7148838979824896e-06, + "loss": 1.3092, + "step": 2034 + }, + { + "epoch": 2.0903954802259888, + "grad_norm": 2.8146934509277344, + "learning_rate": 1.712980586220023e-06, + "loss": 1.3056, + "step": 2035 + }, + { + "epoch": 2.091422701592193, + "grad_norm": 2.812089443206787, + "learning_rate": 1.7110772744575563e-06, + "loss": 1.2924, + "step": 2036 + }, + { + "epoch": 2.0924499229583975, + "grad_norm": 2.7205023765563965, + "learning_rate": 1.7091739626950895e-06, + "loss": 1.3058, + "step": 2037 + }, + { + "epoch": 2.093477144324602, + "grad_norm": 2.692681312561035, + "learning_rate": 1.7072706509326229e-06, + "loss": 1.3336, + "step": 2038 + }, + { + "epoch": 2.0945043656908062, + "grad_norm": 2.8189735412597656, + "learning_rate": 1.7053673391701562e-06, + "loss": 1.3171, + "step": 2039 + }, + { + "epoch": 2.095531587057011, + "grad_norm": 2.664710760116577, + "learning_rate": 1.7034640274076894e-06, + "loss": 1.285, + "step": 2040 + }, + { + "epoch": 2.095531587057011, + "eval_loss": 1.7483630180358887, + "eval_runtime": 37.9016, + "eval_samples_per_second": 26.384, + "eval_steps_per_second": 1.108, + "step": 2040 + }, + { + "epoch": 2.0965588084232154, + "grad_norm": 2.664710760116577, + "learning_rate": 1.7034640274076894e-06, + "loss": 1.35, + "step": 2041 + }, + { + "epoch": 2.0975860297894195, + "grad_norm": 2.6864893436431885, + "learning_rate": 1.701560715645223e-06, + "loss": 1.2723, + "step": 2042 + }, + { + "epoch": 2.098613251155624, + "grad_norm": 2.751150608062744, + "learning_rate": 1.6996574038827563e-06, + "loss": 1.3154, + "step": 2043 + }, + { + "epoch": 2.0996404725218283, + "grad_norm": 2.7600581645965576, + "learning_rate": 1.6977540921202895e-06, + "loss": 1.3043, + "step": 2044 + }, + { + "epoch": 2.100667693888033, + "grad_norm": 2.7397212982177734, + "learning_rate": 1.6958507803578228e-06, + "loss": 1.3183, + "step": 2045 + }, + { + "epoch": 2.1016949152542375, + "grad_norm": 2.7808914184570312, + "learning_rate": 1.693947468595356e-06, + "loss": 1.3367, + "step": 2046 + }, + { + "epoch": 2.1027221366204416, + "grad_norm": 2.534454584121704, + "learning_rate": 1.6920441568328893e-06, + "loss": 1.2638, + "step": 2047 + }, + { + "epoch": 2.103749357986646, + "grad_norm": 2.6857175827026367, + "learning_rate": 1.6901408450704227e-06, + "loss": 1.2955, + "step": 2048 + }, + { + "epoch": 2.1047765793528503, + "grad_norm": 2.689323663711548, + "learning_rate": 1.6882375333079559e-06, + "loss": 1.3482, + "step": 2049 + }, + { + "epoch": 2.105803800719055, + "grad_norm": 2.7330009937286377, + "learning_rate": 1.6863342215454892e-06, + "loss": 1.3405, + "step": 2050 + }, + { + "epoch": 2.105803800719055, + "eval_loss": 1.757278323173523, + "eval_runtime": 37.9263, + "eval_samples_per_second": 26.367, + "eval_steps_per_second": 1.107, + "step": 2050 + }, + { + "epoch": 2.1068310220852595, + "grad_norm": 2.847092628479004, + "learning_rate": 1.6844309097830226e-06, + "loss": 1.3338, + "step": 2051 + }, + { + "epoch": 2.1078582434514637, + "grad_norm": 2.8156185150146484, + "learning_rate": 1.682527598020556e-06, + "loss": 1.3446, + "step": 2052 + }, + { + "epoch": 2.1088854648176683, + "grad_norm": 2.665828227996826, + "learning_rate": 1.6806242862580893e-06, + "loss": 1.2962, + "step": 2053 + }, + { + "epoch": 2.109912686183873, + "grad_norm": 2.6218175888061523, + "learning_rate": 1.6787209744956227e-06, + "loss": 1.2798, + "step": 2054 + }, + { + "epoch": 2.110939907550077, + "grad_norm": 2.622291326522827, + "learning_rate": 1.6768176627331558e-06, + "loss": 1.3243, + "step": 2055 + }, + { + "epoch": 2.1119671289162816, + "grad_norm": 2.7171177864074707, + "learning_rate": 1.6749143509706892e-06, + "loss": 1.2919, + "step": 2056 + }, + { + "epoch": 2.1129943502824857, + "grad_norm": 2.655845880508423, + "learning_rate": 1.6730110392082225e-06, + "loss": 1.3152, + "step": 2057 + }, + { + "epoch": 2.1140215716486903, + "grad_norm": 2.7783730030059814, + "learning_rate": 1.6711077274457557e-06, + "loss": 1.2694, + "step": 2058 + }, + { + "epoch": 2.115048793014895, + "grad_norm": 2.819627046585083, + "learning_rate": 1.669204415683289e-06, + "loss": 1.3335, + "step": 2059 + }, + { + "epoch": 2.116076014381099, + "grad_norm": 2.817685127258301, + "learning_rate": 1.6673011039208222e-06, + "loss": 1.3172, + "step": 2060 + }, + { + "epoch": 2.116076014381099, + "eval_loss": 1.7683359384536743, + "eval_runtime": 37.8941, + "eval_samples_per_second": 26.389, + "eval_steps_per_second": 1.108, + "step": 2060 + }, + { + "epoch": 2.1171032357473036, + "grad_norm": 2.9053642749786377, + "learning_rate": 1.6653977921583556e-06, + "loss": 1.3356, + "step": 2061 + }, + { + "epoch": 2.118130457113508, + "grad_norm": 2.6939432621002197, + "learning_rate": 1.663494480395889e-06, + "loss": 1.2915, + "step": 2062 + }, + { + "epoch": 2.1191576784797124, + "grad_norm": 2.8870749473571777, + "learning_rate": 1.6615911686334225e-06, + "loss": 1.3188, + "step": 2063 + }, + { + "epoch": 2.120184899845917, + "grad_norm": 2.6949965953826904, + "learning_rate": 1.6596878568709557e-06, + "loss": 1.2984, + "step": 2064 + }, + { + "epoch": 2.121212121212121, + "grad_norm": 2.6951143741607666, + "learning_rate": 1.657784545108489e-06, + "loss": 1.3328, + "step": 2065 + }, + { + "epoch": 2.1222393425783257, + "grad_norm": 2.6718437671661377, + "learning_rate": 1.6558812333460222e-06, + "loss": 1.3327, + "step": 2066 + }, + { + "epoch": 2.12326656394453, + "grad_norm": 2.7002549171447754, + "learning_rate": 1.6539779215835555e-06, + "loss": 1.2941, + "step": 2067 + }, + { + "epoch": 2.1242937853107344, + "grad_norm": 2.695836305618286, + "learning_rate": 1.652074609821089e-06, + "loss": 1.301, + "step": 2068 + }, + { + "epoch": 2.125321006676939, + "grad_norm": 2.7710680961608887, + "learning_rate": 1.650171298058622e-06, + "loss": 1.3084, + "step": 2069 + }, + { + "epoch": 2.126348228043143, + "grad_norm": 2.7416014671325684, + "learning_rate": 1.6482679862961554e-06, + "loss": 1.2984, + "step": 2070 + }, + { + "epoch": 2.126348228043143, + "eval_loss": 1.7566925287246704, + "eval_runtime": 37.9783, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 2070 + }, + { + "epoch": 2.1273754494093478, + "grad_norm": 2.570422887802124, + "learning_rate": 1.6463646745336886e-06, + "loss": 1.2817, + "step": 2071 + }, + { + "epoch": 2.128402670775552, + "grad_norm": 2.658071517944336, + "learning_rate": 1.644461362771222e-06, + "loss": 1.308, + "step": 2072 + }, + { + "epoch": 2.1294298921417565, + "grad_norm": 2.65232515335083, + "learning_rate": 1.6425580510087555e-06, + "loss": 1.2854, + "step": 2073 + }, + { + "epoch": 2.130457113507961, + "grad_norm": 2.6072998046875, + "learning_rate": 1.6406547392462889e-06, + "loss": 1.2836, + "step": 2074 + }, + { + "epoch": 2.131484334874165, + "grad_norm": 2.6997647285461426, + "learning_rate": 1.638751427483822e-06, + "loss": 1.3258, + "step": 2075 + }, + { + "epoch": 2.13251155624037, + "grad_norm": 2.5479812622070312, + "learning_rate": 1.6368481157213554e-06, + "loss": 1.2987, + "step": 2076 + }, + { + "epoch": 2.1335387776065744, + "grad_norm": 2.801560878753662, + "learning_rate": 1.6349448039588885e-06, + "loss": 1.3542, + "step": 2077 + }, + { + "epoch": 2.1345659989727785, + "grad_norm": 2.6710598468780518, + "learning_rate": 1.6330414921964219e-06, + "loss": 1.2832, + "step": 2078 + }, + { + "epoch": 2.135593220338983, + "grad_norm": 2.8505167961120605, + "learning_rate": 1.6311381804339553e-06, + "loss": 1.327, + "step": 2079 + }, + { + "epoch": 2.1366204417051873, + "grad_norm": 2.6584553718566895, + "learning_rate": 1.6292348686714884e-06, + "loss": 1.3003, + "step": 2080 + }, + { + "epoch": 2.1366204417051873, + "eval_loss": 1.7610517740249634, + "eval_runtime": 37.9139, + "eval_samples_per_second": 26.376, + "eval_steps_per_second": 1.108, + "step": 2080 + }, + { + "epoch": 2.137647663071392, + "grad_norm": 2.7930376529693604, + "learning_rate": 1.6273315569090218e-06, + "loss": 1.3759, + "step": 2081 + }, + { + "epoch": 2.1386748844375965, + "grad_norm": 2.6514129638671875, + "learning_rate": 1.625428245146555e-06, + "loss": 1.3512, + "step": 2082 + }, + { + "epoch": 2.1397021058038006, + "grad_norm": 2.626838445663452, + "learning_rate": 1.6235249333840885e-06, + "loss": 1.3074, + "step": 2083 + }, + { + "epoch": 2.140729327170005, + "grad_norm": 2.741288661956787, + "learning_rate": 1.6216216216216219e-06, + "loss": 1.3327, + "step": 2084 + }, + { + "epoch": 2.1417565485362093, + "grad_norm": 2.7371652126312256, + "learning_rate": 1.6197183098591552e-06, + "loss": 1.3193, + "step": 2085 + }, + { + "epoch": 2.142783769902414, + "grad_norm": 2.631516218185425, + "learning_rate": 1.6178149980966884e-06, + "loss": 1.2899, + "step": 2086 + }, + { + "epoch": 2.1438109912686185, + "grad_norm": 2.7853667736053467, + "learning_rate": 1.6159116863342217e-06, + "loss": 1.356, + "step": 2087 + }, + { + "epoch": 2.1448382126348227, + "grad_norm": 2.598886489868164, + "learning_rate": 1.6140083745717549e-06, + "loss": 1.2885, + "step": 2088 + }, + { + "epoch": 2.1458654340010272, + "grad_norm": 2.700669288635254, + "learning_rate": 1.6121050628092882e-06, + "loss": 1.3471, + "step": 2089 + }, + { + "epoch": 2.146892655367232, + "grad_norm": 2.6174018383026123, + "learning_rate": 1.6102017510468216e-06, + "loss": 1.2635, + "step": 2090 + }, + { + "epoch": 2.146892655367232, + "eval_loss": 1.764804482460022, + "eval_runtime": 37.8935, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 2090 + }, + { + "epoch": 2.147919876733436, + "grad_norm": 2.8135688304901123, + "learning_rate": 1.6082984392843548e-06, + "loss": 1.3178, + "step": 2091 + }, + { + "epoch": 2.1489470980996406, + "grad_norm": 2.8416011333465576, + "learning_rate": 1.6063951275218881e-06, + "loss": 1.3652, + "step": 2092 + }, + { + "epoch": 2.1499743194658447, + "grad_norm": 2.780001640319824, + "learning_rate": 1.6044918157594217e-06, + "loss": 1.2921, + "step": 2093 + }, + { + "epoch": 2.1510015408320493, + "grad_norm": 2.634460926055908, + "learning_rate": 1.6025885039969548e-06, + "loss": 1.3224, + "step": 2094 + }, + { + "epoch": 2.152028762198254, + "grad_norm": 2.6783623695373535, + "learning_rate": 1.6006851922344882e-06, + "loss": 1.3297, + "step": 2095 + }, + { + "epoch": 2.153055983564458, + "grad_norm": 2.7781169414520264, + "learning_rate": 1.5987818804720216e-06, + "loss": 1.3324, + "step": 2096 + }, + { + "epoch": 2.1540832049306626, + "grad_norm": 2.587174415588379, + "learning_rate": 1.5968785687095547e-06, + "loss": 1.2801, + "step": 2097 + }, + { + "epoch": 2.1551104262968668, + "grad_norm": 2.654083728790283, + "learning_rate": 1.594975256947088e-06, + "loss": 1.3323, + "step": 2098 + }, + { + "epoch": 2.1561376476630714, + "grad_norm": 2.6695969104766846, + "learning_rate": 1.5930719451846212e-06, + "loss": 1.289, + "step": 2099 + }, + { + "epoch": 2.157164869029276, + "grad_norm": 2.7060039043426514, + "learning_rate": 1.5911686334221546e-06, + "loss": 1.3425, + "step": 2100 + }, + { + "epoch": 2.157164869029276, + "eval_loss": 1.7727104425430298, + "eval_runtime": 37.9001, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 2100 + }, + { + "epoch": 2.15819209039548, + "grad_norm": 2.536756753921509, + "learning_rate": 1.589265321659688e-06, + "loss": 1.2538, + "step": 2101 + }, + { + "epoch": 2.1592193117616847, + "grad_norm": 2.655916213989258, + "learning_rate": 1.5873620098972211e-06, + "loss": 1.2925, + "step": 2102 + }, + { + "epoch": 2.1602465331278893, + "grad_norm": 2.686228036880493, + "learning_rate": 1.5854586981347547e-06, + "loss": 1.3386, + "step": 2103 + }, + { + "epoch": 2.1612737544940934, + "grad_norm": 2.794640302658081, + "learning_rate": 1.583555386372288e-06, + "loss": 1.348, + "step": 2104 + }, + { + "epoch": 2.162300975860298, + "grad_norm": 2.6761186122894287, + "learning_rate": 1.5816520746098212e-06, + "loss": 1.3029, + "step": 2105 + }, + { + "epoch": 2.163328197226502, + "grad_norm": 2.5585744380950928, + "learning_rate": 1.5797487628473546e-06, + "loss": 1.2802, + "step": 2106 + }, + { + "epoch": 2.1643554185927067, + "grad_norm": 2.6511101722717285, + "learning_rate": 1.577845451084888e-06, + "loss": 1.3048, + "step": 2107 + }, + { + "epoch": 2.1653826399589113, + "grad_norm": 2.601940870285034, + "learning_rate": 1.575942139322421e-06, + "loss": 1.2433, + "step": 2108 + }, + { + "epoch": 2.1664098613251155, + "grad_norm": 2.676440954208374, + "learning_rate": 1.5740388275599544e-06, + "loss": 1.3296, + "step": 2109 + }, + { + "epoch": 2.16743708269132, + "grad_norm": 2.7463326454162598, + "learning_rate": 1.5721355157974876e-06, + "loss": 1.2931, + "step": 2110 + }, + { + "epoch": 2.16743708269132, + "eval_loss": 1.7691689729690552, + "eval_runtime": 37.6043, + "eval_samples_per_second": 26.593, + "eval_steps_per_second": 1.117, + "step": 2110 + }, + { + "epoch": 2.168464304057524, + "grad_norm": 2.895216464996338, + "learning_rate": 1.570232204035021e-06, + "loss": 1.3091, + "step": 2111 + }, + { + "epoch": 2.169491525423729, + "grad_norm": 2.765704870223999, + "learning_rate": 1.5683288922725543e-06, + "loss": 1.3197, + "step": 2112 + }, + { + "epoch": 2.1705187467899334, + "grad_norm": 2.750681161880493, + "learning_rate": 1.5664255805100879e-06, + "loss": 1.3016, + "step": 2113 + }, + { + "epoch": 2.1715459681561375, + "grad_norm": 2.6807963848114014, + "learning_rate": 1.564522268747621e-06, + "loss": 1.3176, + "step": 2114 + }, + { + "epoch": 2.172573189522342, + "grad_norm": 2.618666410446167, + "learning_rate": 1.5626189569851544e-06, + "loss": 1.2391, + "step": 2115 + }, + { + "epoch": 2.1736004108885463, + "grad_norm": 2.745319128036499, + "learning_rate": 1.5607156452226875e-06, + "loss": 1.3264, + "step": 2116 + }, + { + "epoch": 2.174627632254751, + "grad_norm": 2.787904977798462, + "learning_rate": 1.558812333460221e-06, + "loss": 1.319, + "step": 2117 + }, + { + "epoch": 2.1756548536209555, + "grad_norm": 2.6901752948760986, + "learning_rate": 1.5569090216977543e-06, + "loss": 1.3216, + "step": 2118 + }, + { + "epoch": 2.1766820749871596, + "grad_norm": 2.723172426223755, + "learning_rate": 1.5550057099352874e-06, + "loss": 1.3241, + "step": 2119 + }, + { + "epoch": 2.177709296353364, + "grad_norm": 2.6102678775787354, + "learning_rate": 1.5531023981728208e-06, + "loss": 1.2892, + "step": 2120 + }, + { + "epoch": 2.177709296353364, + "eval_loss": 1.7888449430465698, + "eval_runtime": 37.7503, + "eval_samples_per_second": 26.49, + "eval_steps_per_second": 1.113, + "step": 2120 + }, + { + "epoch": 2.1787365177195683, + "grad_norm": 2.653026819229126, + "learning_rate": 1.5511990864103542e-06, + "loss": 1.322, + "step": 2121 + }, + { + "epoch": 2.179763739085773, + "grad_norm": 2.806363344192505, + "learning_rate": 1.5492957746478873e-06, + "loss": 1.3412, + "step": 2122 + }, + { + "epoch": 2.1807909604519775, + "grad_norm": 2.682968854904175, + "learning_rate": 1.5473924628854209e-06, + "loss": 1.2953, + "step": 2123 + }, + { + "epoch": 2.1818181818181817, + "grad_norm": 2.7165629863739014, + "learning_rate": 1.5454891511229542e-06, + "loss": 1.312, + "step": 2124 + }, + { + "epoch": 2.1828454031843862, + "grad_norm": 2.8055427074432373, + "learning_rate": 1.5435858393604874e-06, + "loss": 1.3318, + "step": 2125 + }, + { + "epoch": 2.183872624550591, + "grad_norm": 2.731459617614746, + "learning_rate": 1.5416825275980208e-06, + "loss": 1.3059, + "step": 2126 + }, + { + "epoch": 2.184899845916795, + "grad_norm": 2.6671993732452393, + "learning_rate": 1.539779215835554e-06, + "loss": 1.2945, + "step": 2127 + }, + { + "epoch": 2.1859270672829996, + "grad_norm": 2.8046534061431885, + "learning_rate": 1.5378759040730873e-06, + "loss": 1.3437, + "step": 2128 + }, + { + "epoch": 2.1869542886492037, + "grad_norm": 2.717656135559082, + "learning_rate": 1.5359725923106206e-06, + "loss": 1.3221, + "step": 2129 + }, + { + "epoch": 2.1879815100154083, + "grad_norm": 2.746224880218506, + "learning_rate": 1.5340692805481538e-06, + "loss": 1.2857, + "step": 2130 + }, + { + "epoch": 2.1879815100154083, + "eval_loss": 1.777937889099121, + "eval_runtime": 37.8481, + "eval_samples_per_second": 26.421, + "eval_steps_per_second": 1.11, + "step": 2130 + }, + { + "epoch": 2.189008731381613, + "grad_norm": 2.6477184295654297, + "learning_rate": 1.5321659687856871e-06, + "loss": 1.2811, + "step": 2131 + }, + { + "epoch": 2.190035952747817, + "grad_norm": 2.664757251739502, + "learning_rate": 1.5302626570232205e-06, + "loss": 1.2542, + "step": 2132 + }, + { + "epoch": 2.1910631741140216, + "grad_norm": 2.8133883476257324, + "learning_rate": 1.5283593452607539e-06, + "loss": 1.3485, + "step": 2133 + }, + { + "epoch": 2.1920903954802258, + "grad_norm": 2.723867893218994, + "learning_rate": 1.5264560334982872e-06, + "loss": 1.2974, + "step": 2134 + }, + { + "epoch": 2.1931176168464304, + "grad_norm": 2.776700496673584, + "learning_rate": 1.5245527217358206e-06, + "loss": 1.3219, + "step": 2135 + }, + { + "epoch": 2.194144838212635, + "grad_norm": 2.547466993331909, + "learning_rate": 1.5226494099733537e-06, + "loss": 1.2923, + "step": 2136 + }, + { + "epoch": 2.195172059578839, + "grad_norm": 2.7371773719787598, + "learning_rate": 1.520746098210887e-06, + "loss": 1.3466, + "step": 2137 + }, + { + "epoch": 2.1961992809450437, + "grad_norm": 2.4609546661376953, + "learning_rate": 1.5188427864484205e-06, + "loss": 1.2865, + "step": 2138 + }, + { + "epoch": 2.1972265023112483, + "grad_norm": 2.6336286067962646, + "learning_rate": 1.5169394746859536e-06, + "loss": 1.3071, + "step": 2139 + }, + { + "epoch": 2.1982537236774524, + "grad_norm": 2.6983938217163086, + "learning_rate": 1.515036162923487e-06, + "loss": 1.3216, + "step": 2140 + }, + { + "epoch": 2.1982537236774524, + "eval_loss": 1.7778681516647339, + "eval_runtime": 37.873, + "eval_samples_per_second": 26.404, + "eval_steps_per_second": 1.109, + "step": 2140 + }, + { + "epoch": 2.199280945043657, + "grad_norm": 2.8055241107940674, + "learning_rate": 1.5131328511610201e-06, + "loss": 1.3517, + "step": 2141 + }, + { + "epoch": 2.200308166409861, + "grad_norm": 2.760460615158081, + "learning_rate": 1.5112295393985535e-06, + "loss": 1.3242, + "step": 2142 + }, + { + "epoch": 2.2013353877760657, + "grad_norm": 2.574289560317993, + "learning_rate": 1.509326227636087e-06, + "loss": 1.2595, + "step": 2143 + }, + { + "epoch": 2.2023626091422703, + "grad_norm": 2.827829360961914, + "learning_rate": 1.5074229158736204e-06, + "loss": 1.3167, + "step": 2144 + }, + { + "epoch": 2.2033898305084745, + "grad_norm": 2.7533328533172607, + "learning_rate": 1.5055196041111536e-06, + "loss": 1.3161, + "step": 2145 + }, + { + "epoch": 2.204417051874679, + "grad_norm": 2.865976095199585, + "learning_rate": 1.503616292348687e-06, + "loss": 1.3444, + "step": 2146 + }, + { + "epoch": 2.205444273240883, + "grad_norm": 2.5228123664855957, + "learning_rate": 1.50171298058622e-06, + "loss": 1.2566, + "step": 2147 + }, + { + "epoch": 2.206471494607088, + "grad_norm": 2.682549238204956, + "learning_rate": 1.4998096688237535e-06, + "loss": 1.3284, + "step": 2148 + }, + { + "epoch": 2.2074987159732924, + "grad_norm": 2.8158624172210693, + "learning_rate": 1.4979063570612868e-06, + "loss": 1.3148, + "step": 2149 + }, + { + "epoch": 2.2085259373394965, + "grad_norm": 2.7644388675689697, + "learning_rate": 1.49600304529882e-06, + "loss": 1.2914, + "step": 2150 + }, + { + "epoch": 2.2085259373394965, + "eval_loss": 1.7767399549484253, + "eval_runtime": 37.8771, + "eval_samples_per_second": 26.401, + "eval_steps_per_second": 1.109, + "step": 2150 + }, + { + "epoch": 2.209553158705701, + "grad_norm": 2.7370104789733887, + "learning_rate": 1.4940997335363533e-06, + "loss": 1.3165, + "step": 2151 + }, + { + "epoch": 2.2105803800719057, + "grad_norm": 2.692795753479004, + "learning_rate": 1.4921964217738865e-06, + "loss": 1.3038, + "step": 2152 + }, + { + "epoch": 2.21160760143811, + "grad_norm": 2.8083877563476562, + "learning_rate": 1.49029311001142e-06, + "loss": 1.3315, + "step": 2153 + }, + { + "epoch": 2.2126348228043144, + "grad_norm": 2.816497325897217, + "learning_rate": 1.4883897982489534e-06, + "loss": 1.3319, + "step": 2154 + }, + { + "epoch": 2.2136620441705186, + "grad_norm": 2.7338058948516846, + "learning_rate": 1.4864864864864868e-06, + "loss": 1.3057, + "step": 2155 + }, + { + "epoch": 2.214689265536723, + "grad_norm": 2.631532669067383, + "learning_rate": 1.48458317472402e-06, + "loss": 1.3123, + "step": 2156 + }, + { + "epoch": 2.2157164869029278, + "grad_norm": 2.511723279953003, + "learning_rate": 1.4826798629615533e-06, + "loss": 1.2714, + "step": 2157 + }, + { + "epoch": 2.216743708269132, + "grad_norm": 2.6720175743103027, + "learning_rate": 1.4807765511990864e-06, + "loss": 1.2806, + "step": 2158 + }, + { + "epoch": 2.2177709296353365, + "grad_norm": 2.6232240200042725, + "learning_rate": 1.4788732394366198e-06, + "loss": 1.3057, + "step": 2159 + }, + { + "epoch": 2.2187981510015407, + "grad_norm": 2.7055530548095703, + "learning_rate": 1.4769699276741532e-06, + "loss": 1.2896, + "step": 2160 + }, + { + "epoch": 2.2187981510015407, + "eval_loss": 1.780350923538208, + "eval_runtime": 37.9284, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 2160 + }, + { + "epoch": 2.2198253723677452, + "grad_norm": 2.7227275371551514, + "learning_rate": 1.4750666159116863e-06, + "loss": 1.2688, + "step": 2161 + }, + { + "epoch": 2.22085259373395, + "grad_norm": 2.741649866104126, + "learning_rate": 1.4731633041492197e-06, + "loss": 1.3318, + "step": 2162 + }, + { + "epoch": 2.221879815100154, + "grad_norm": 2.901906967163086, + "learning_rate": 1.4712599923867528e-06, + "loss": 1.3502, + "step": 2163 + }, + { + "epoch": 2.2229070364663586, + "grad_norm": 2.6306073665618896, + "learning_rate": 1.4693566806242864e-06, + "loss": 1.2753, + "step": 2164 + }, + { + "epoch": 2.2239342578325627, + "grad_norm": 2.806844472885132, + "learning_rate": 1.4674533688618198e-06, + "loss": 1.3006, + "step": 2165 + }, + { + "epoch": 2.2249614791987673, + "grad_norm": 2.754244089126587, + "learning_rate": 1.4655500570993531e-06, + "loss": 1.3036, + "step": 2166 + }, + { + "epoch": 2.225988700564972, + "grad_norm": 2.7528069019317627, + "learning_rate": 1.4636467453368863e-06, + "loss": 1.3551, + "step": 2167 + }, + { + "epoch": 2.227015921931176, + "grad_norm": 2.9147045612335205, + "learning_rate": 1.4617434335744197e-06, + "loss": 1.3444, + "step": 2168 + }, + { + "epoch": 2.2280431432973806, + "grad_norm": 2.7143068313598633, + "learning_rate": 1.4598401218119528e-06, + "loss": 1.3333, + "step": 2169 + }, + { + "epoch": 2.2290703646635848, + "grad_norm": 2.7245426177978516, + "learning_rate": 1.4579368100494862e-06, + "loss": 1.32, + "step": 2170 + }, + { + "epoch": 2.2290703646635848, + "eval_loss": 1.7687398195266724, + "eval_runtime": 37.931, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 2170 + }, + { + "epoch": 2.2300975860297894, + "grad_norm": 2.614847421646118, + "learning_rate": 1.4560334982870195e-06, + "loss": 1.316, + "step": 2171 + }, + { + "epoch": 2.231124807395994, + "grad_norm": 2.6008236408233643, + "learning_rate": 1.4541301865245527e-06, + "loss": 1.3271, + "step": 2172 + }, + { + "epoch": 2.232152028762198, + "grad_norm": 2.6043577194213867, + "learning_rate": 1.452226874762086e-06, + "loss": 1.2838, + "step": 2173 + }, + { + "epoch": 2.2331792501284027, + "grad_norm": 2.800450325012207, + "learning_rate": 1.4503235629996196e-06, + "loss": 1.3542, + "step": 2174 + }, + { + "epoch": 2.2342064714946073, + "grad_norm": 2.6808807849884033, + "learning_rate": 1.4484202512371528e-06, + "loss": 1.2573, + "step": 2175 + }, + { + "epoch": 2.2352336928608114, + "grad_norm": 3.005553722381592, + "learning_rate": 1.4465169394746861e-06, + "loss": 1.3598, + "step": 2176 + }, + { + "epoch": 2.236260914227016, + "grad_norm": 2.8075437545776367, + "learning_rate": 1.4446136277122195e-06, + "loss": 1.2687, + "step": 2177 + }, + { + "epoch": 2.23728813559322, + "grad_norm": 2.828524589538574, + "learning_rate": 1.4427103159497526e-06, + "loss": 1.2906, + "step": 2178 + }, + { + "epoch": 2.2383153569594247, + "grad_norm": 2.7086875438690186, + "learning_rate": 1.440807004187286e-06, + "loss": 1.3199, + "step": 2179 + }, + { + "epoch": 2.2393425783256293, + "grad_norm": 2.8826253414154053, + "learning_rate": 1.4389036924248192e-06, + "loss": 1.3414, + "step": 2180 + }, + { + "epoch": 2.2393425783256293, + "eval_loss": 1.7614880800247192, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 2180 + }, + { + "epoch": 2.2403697996918335, + "grad_norm": 2.8024802207946777, + "learning_rate": 1.4370003806623525e-06, + "loss": 1.3127, + "step": 2181 + }, + { + "epoch": 2.241397021058038, + "grad_norm": 2.7175004482269287, + "learning_rate": 1.4350970688998859e-06, + "loss": 1.3589, + "step": 2182 + }, + { + "epoch": 2.242424242424242, + "grad_norm": 2.7450461387634277, + "learning_rate": 1.433193757137419e-06, + "loss": 1.2934, + "step": 2183 + }, + { + "epoch": 2.243451463790447, + "grad_norm": 2.627926826477051, + "learning_rate": 1.4312904453749526e-06, + "loss": 1.2605, + "step": 2184 + }, + { + "epoch": 2.2444786851566514, + "grad_norm": 2.7466957569122314, + "learning_rate": 1.429387133612486e-06, + "loss": 1.2993, + "step": 2185 + }, + { + "epoch": 2.2455059065228555, + "grad_norm": 2.8719801902770996, + "learning_rate": 1.4274838218500191e-06, + "loss": 1.3251, + "step": 2186 + }, + { + "epoch": 2.24653312788906, + "grad_norm": 3.052253007888794, + "learning_rate": 1.4255805100875525e-06, + "loss": 1.3638, + "step": 2187 + }, + { + "epoch": 2.2475603492552647, + "grad_norm": 2.860090970993042, + "learning_rate": 1.4236771983250858e-06, + "loss": 1.3121, + "step": 2188 + }, + { + "epoch": 2.248587570621469, + "grad_norm": 2.656618356704712, + "learning_rate": 1.421773886562619e-06, + "loss": 1.2857, + "step": 2189 + }, + { + "epoch": 2.2496147919876734, + "grad_norm": 2.6252174377441406, + "learning_rate": 1.4198705748001524e-06, + "loss": 1.3094, + "step": 2190 + }, + { + "epoch": 2.2496147919876734, + "eval_loss": 1.7660510540008545, + "eval_runtime": 37.9085, + "eval_samples_per_second": 26.379, + "eval_steps_per_second": 1.108, + "step": 2190 + }, + { + "epoch": 2.2506420133538776, + "grad_norm": 2.610424757003784, + "learning_rate": 1.4179672630376855e-06, + "loss": 1.2809, + "step": 2191 + }, + { + "epoch": 2.251669234720082, + "grad_norm": 2.5957748889923096, + "learning_rate": 1.4160639512752189e-06, + "loss": 1.3009, + "step": 2192 + }, + { + "epoch": 2.2526964560862868, + "grad_norm": 2.6947720050811768, + "learning_rate": 1.4141606395127522e-06, + "loss": 1.2697, + "step": 2193 + }, + { + "epoch": 2.253723677452491, + "grad_norm": 2.5192782878875732, + "learning_rate": 1.4122573277502858e-06, + "loss": 1.3245, + "step": 2194 + }, + { + "epoch": 2.2547508988186955, + "grad_norm": 2.7121758460998535, + "learning_rate": 1.410354015987819e-06, + "loss": 1.3012, + "step": 2195 + }, + { + "epoch": 2.2557781201848996, + "grad_norm": 2.8090286254882812, + "learning_rate": 1.4084507042253523e-06, + "loss": 1.3411, + "step": 2196 + }, + { + "epoch": 2.2568053415511042, + "grad_norm": 2.6683340072631836, + "learning_rate": 1.4065473924628855e-06, + "loss": 1.3059, + "step": 2197 + }, + { + "epoch": 2.257832562917309, + "grad_norm": 2.782543659210205, + "learning_rate": 1.4046440807004188e-06, + "loss": 1.2978, + "step": 2198 + }, + { + "epoch": 2.258859784283513, + "grad_norm": 2.693094253540039, + "learning_rate": 1.4027407689379522e-06, + "loss": 1.2785, + "step": 2199 + }, + { + "epoch": 2.2598870056497176, + "grad_norm": 2.7780873775482178, + "learning_rate": 1.4008374571754854e-06, + "loss": 1.3178, + "step": 2200 + }, + { + "epoch": 2.2598870056497176, + "eval_loss": 1.7748886346817017, + "eval_runtime": 37.9075, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 2200 + }, + { + "epoch": 2.260914227015922, + "grad_norm": 2.670476198196411, + "learning_rate": 1.3989341454130187e-06, + "loss": 1.3095, + "step": 2201 + }, + { + "epoch": 2.2619414483821263, + "grad_norm": 2.7604143619537354, + "learning_rate": 1.397030833650552e-06, + "loss": 1.3342, + "step": 2202 + }, + { + "epoch": 2.262968669748331, + "grad_norm": 2.7799160480499268, + "learning_rate": 1.3951275218880852e-06, + "loss": 1.3734, + "step": 2203 + }, + { + "epoch": 2.263995891114535, + "grad_norm": 2.5651214122772217, + "learning_rate": 1.3932242101256188e-06, + "loss": 1.2775, + "step": 2204 + }, + { + "epoch": 2.2650231124807396, + "grad_norm": 2.8254587650299072, + "learning_rate": 1.3913208983631522e-06, + "loss": 1.3148, + "step": 2205 + }, + { + "epoch": 2.266050333846944, + "grad_norm": 2.694624662399292, + "learning_rate": 1.3894175866006853e-06, + "loss": 1.2702, + "step": 2206 + }, + { + "epoch": 2.2670775552131484, + "grad_norm": 2.789055824279785, + "learning_rate": 1.3875142748382187e-06, + "loss": 1.2516, + "step": 2207 + }, + { + "epoch": 2.268104776579353, + "grad_norm": 2.848062753677368, + "learning_rate": 1.385610963075752e-06, + "loss": 1.3162, + "step": 2208 + }, + { + "epoch": 2.269131997945557, + "grad_norm": 2.939532518386841, + "learning_rate": 1.3837076513132852e-06, + "loss": 1.3428, + "step": 2209 + }, + { + "epoch": 2.2701592193117617, + "grad_norm": 2.9084253311157227, + "learning_rate": 1.3818043395508186e-06, + "loss": 1.3277, + "step": 2210 + }, + { + "epoch": 2.2701592193117617, + "eval_loss": 1.786512017250061, + "eval_runtime": 37.685, + "eval_samples_per_second": 26.536, + "eval_steps_per_second": 1.115, + "step": 2210 + }, + { + "epoch": 2.2711864406779663, + "grad_norm": 2.6281025409698486, + "learning_rate": 1.3799010277883517e-06, + "loss": 1.3128, + "step": 2211 + }, + { + "epoch": 2.2722136620441704, + "grad_norm": 2.748809337615967, + "learning_rate": 1.377997716025885e-06, + "loss": 1.3114, + "step": 2212 + }, + { + "epoch": 2.273240883410375, + "grad_norm": 2.5523533821105957, + "learning_rate": 1.3760944042634184e-06, + "loss": 1.3401, + "step": 2213 + }, + { + "epoch": 2.2742681047765796, + "grad_norm": 2.6466875076293945, + "learning_rate": 1.3741910925009518e-06, + "loss": 1.3338, + "step": 2214 + }, + { + "epoch": 2.2752953261427837, + "grad_norm": 2.7305641174316406, + "learning_rate": 1.3722877807384852e-06, + "loss": 1.3321, + "step": 2215 + }, + { + "epoch": 2.2763225475089883, + "grad_norm": 2.6950535774230957, + "learning_rate": 1.3703844689760185e-06, + "loss": 1.3053, + "step": 2216 + }, + { + "epoch": 2.2773497688751925, + "grad_norm": 2.5823051929473877, + "learning_rate": 1.3684811572135517e-06, + "loss": 1.2622, + "step": 2217 + }, + { + "epoch": 2.278376990241397, + "grad_norm": 2.891983985900879, + "learning_rate": 1.366577845451085e-06, + "loss": 1.3346, + "step": 2218 + }, + { + "epoch": 2.279404211607601, + "grad_norm": 2.8419189453125, + "learning_rate": 1.3646745336886184e-06, + "loss": 1.3336, + "step": 2219 + }, + { + "epoch": 2.280431432973806, + "grad_norm": 2.632622480392456, + "learning_rate": 1.3627712219261515e-06, + "loss": 1.2895, + "step": 2220 + }, + { + "epoch": 2.280431432973806, + "eval_loss": 1.7908138036727905, + "eval_runtime": 37.8557, + "eval_samples_per_second": 26.416, + "eval_steps_per_second": 1.109, + "step": 2220 + }, + { + "epoch": 2.2814586543400104, + "grad_norm": 2.6359457969665527, + "learning_rate": 1.360867910163685e-06, + "loss": 1.3073, + "step": 2221 + }, + { + "epoch": 2.2824858757062145, + "grad_norm": 2.6809372901916504, + "learning_rate": 1.358964598401218e-06, + "loss": 1.3131, + "step": 2222 + }, + { + "epoch": 2.283513097072419, + "grad_norm": 2.579667806625366, + "learning_rate": 1.3570612866387514e-06, + "loss": 1.2958, + "step": 2223 + }, + { + "epoch": 2.2845403184386237, + "grad_norm": 2.8612430095672607, + "learning_rate": 1.355157974876285e-06, + "loss": 1.3478, + "step": 2224 + }, + { + "epoch": 2.285567539804828, + "grad_norm": 2.728100061416626, + "learning_rate": 1.3532546631138184e-06, + "loss": 1.3238, + "step": 2225 + }, + { + "epoch": 2.2865947611710324, + "grad_norm": 2.665339469909668, + "learning_rate": 1.3513513513513515e-06, + "loss": 1.2984, + "step": 2226 + }, + { + "epoch": 2.2876219825372366, + "grad_norm": 2.677293062210083, + "learning_rate": 1.3494480395888849e-06, + "loss": 1.3052, + "step": 2227 + }, + { + "epoch": 2.288649203903441, + "grad_norm": 2.8480992317199707, + "learning_rate": 1.347544727826418e-06, + "loss": 1.2887, + "step": 2228 + }, + { + "epoch": 2.2896764252696458, + "grad_norm": 2.6627776622772217, + "learning_rate": 1.3456414160639514e-06, + "loss": 1.3243, + "step": 2229 + }, + { + "epoch": 2.29070364663585, + "grad_norm": 2.711862087249756, + "learning_rate": 1.3437381043014847e-06, + "loss": 1.3112, + "step": 2230 + }, + { + "epoch": 2.29070364663585, + "eval_loss": 1.781941533088684, + "eval_runtime": 37.9343, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 2230 + }, + { + "epoch": 2.2917308680020545, + "grad_norm": 2.6429977416992188, + "learning_rate": 1.341834792539018e-06, + "loss": 1.2945, + "step": 2231 + }, + { + "epoch": 2.2927580893682586, + "grad_norm": 2.557021379470825, + "learning_rate": 1.3399314807765513e-06, + "loss": 1.32, + "step": 2232 + }, + { + "epoch": 2.2937853107344632, + "grad_norm": 2.5697803497314453, + "learning_rate": 1.3380281690140844e-06, + "loss": 1.3161, + "step": 2233 + }, + { + "epoch": 2.294812532100668, + "grad_norm": 2.5973010063171387, + "learning_rate": 1.336124857251618e-06, + "loss": 1.2822, + "step": 2234 + }, + { + "epoch": 2.295839753466872, + "grad_norm": 2.622098207473755, + "learning_rate": 1.3342215454891513e-06, + "loss": 1.2853, + "step": 2235 + }, + { + "epoch": 2.2968669748330766, + "grad_norm": 2.75445294380188, + "learning_rate": 1.3323182337266847e-06, + "loss": 1.3009, + "step": 2236 + }, + { + "epoch": 2.297894196199281, + "grad_norm": 2.8000597953796387, + "learning_rate": 1.3304149219642179e-06, + "loss": 1.3547, + "step": 2237 + }, + { + "epoch": 2.2989214175654853, + "grad_norm": 2.7219057083129883, + "learning_rate": 1.3285116102017512e-06, + "loss": 1.3128, + "step": 2238 + }, + { + "epoch": 2.29994863893169, + "grad_norm": 2.5688209533691406, + "learning_rate": 1.3266082984392844e-06, + "loss": 1.3077, + "step": 2239 + }, + { + "epoch": 2.300975860297894, + "grad_norm": 2.629265546798706, + "learning_rate": 1.3247049866768177e-06, + "loss": 1.2929, + "step": 2240 + }, + { + "epoch": 2.300975860297894, + "eval_loss": 1.7657849788665771, + "eval_runtime": 37.9938, + "eval_samples_per_second": 26.32, + "eval_steps_per_second": 1.105, + "step": 2240 + }, + { + "epoch": 2.3020030816640986, + "grad_norm": 2.6893677711486816, + "learning_rate": 1.322801674914351e-06, + "loss": 1.2815, + "step": 2241 + }, + { + "epoch": 2.303030303030303, + "grad_norm": 2.6499500274658203, + "learning_rate": 1.3208983631518843e-06, + "loss": 1.3231, + "step": 2242 + }, + { + "epoch": 2.3040575243965074, + "grad_norm": 2.6495258808135986, + "learning_rate": 1.3189950513894176e-06, + "loss": 1.3084, + "step": 2243 + }, + { + "epoch": 2.305084745762712, + "grad_norm": 2.6610732078552246, + "learning_rate": 1.3170917396269512e-06, + "loss": 1.3182, + "step": 2244 + }, + { + "epoch": 2.306111967128916, + "grad_norm": 2.8128137588500977, + "learning_rate": 1.3151884278644843e-06, + "loss": 1.327, + "step": 2245 + }, + { + "epoch": 2.3071391884951207, + "grad_norm": 2.7301197052001953, + "learning_rate": 1.3132851161020177e-06, + "loss": 1.2781, + "step": 2246 + }, + { + "epoch": 2.3081664098613253, + "grad_norm": 2.9533233642578125, + "learning_rate": 1.311381804339551e-06, + "loss": 1.3728, + "step": 2247 + }, + { + "epoch": 2.3091936312275294, + "grad_norm": 2.6429927349090576, + "learning_rate": 1.3094784925770842e-06, + "loss": 1.2931, + "step": 2248 + }, + { + "epoch": 2.310220852593734, + "grad_norm": 2.649885416030884, + "learning_rate": 1.3075751808146176e-06, + "loss": 1.3158, + "step": 2249 + }, + { + "epoch": 2.3112480739599386, + "grad_norm": 2.6279773712158203, + "learning_rate": 1.3056718690521507e-06, + "loss": 1.2875, + "step": 2250 + }, + { + "epoch": 2.3112480739599386, + "eval_loss": 1.777553915977478, + "eval_runtime": 38.0475, + "eval_samples_per_second": 26.283, + "eval_steps_per_second": 1.104, + "step": 2250 + }, + { + "epoch": 2.3122752953261427, + "grad_norm": 2.6415603160858154, + "learning_rate": 1.303768557289684e-06, + "loss": 1.289, + "step": 2251 + }, + { + "epoch": 2.3133025166923473, + "grad_norm": 2.586108684539795, + "learning_rate": 1.3018652455272175e-06, + "loss": 1.2985, + "step": 2252 + }, + { + "epoch": 2.3143297380585515, + "grad_norm": 2.592040538787842, + "learning_rate": 1.2999619337647506e-06, + "loss": 1.2964, + "step": 2253 + }, + { + "epoch": 2.315356959424756, + "grad_norm": 2.6993772983551025, + "learning_rate": 1.2980586220022842e-06, + "loss": 1.3257, + "step": 2254 + }, + { + "epoch": 2.3163841807909606, + "grad_norm": 2.7006278038024902, + "learning_rate": 1.2961553102398175e-06, + "loss": 1.3463, + "step": 2255 + }, + { + "epoch": 2.317411402157165, + "grad_norm": 2.749055862426758, + "learning_rate": 1.2942519984773507e-06, + "loss": 1.3183, + "step": 2256 + }, + { + "epoch": 2.3184386235233694, + "grad_norm": 2.8093061447143555, + "learning_rate": 1.292348686714884e-06, + "loss": 1.2923, + "step": 2257 + }, + { + "epoch": 2.3194658448895735, + "grad_norm": 2.6441662311553955, + "learning_rate": 1.2904453749524174e-06, + "loss": 1.3332, + "step": 2258 + }, + { + "epoch": 2.320493066255778, + "grad_norm": 2.588270425796509, + "learning_rate": 1.2885420631899506e-06, + "loss": 1.2563, + "step": 2259 + }, + { + "epoch": 2.3215202876219827, + "grad_norm": 2.659374952316284, + "learning_rate": 1.286638751427484e-06, + "loss": 1.2948, + "step": 2260 + }, + { + "epoch": 2.3215202876219827, + "eval_loss": 1.7904608249664307, + "eval_runtime": 38.0335, + "eval_samples_per_second": 26.293, + "eval_steps_per_second": 1.104, + "step": 2260 + }, + { + "epoch": 2.322547508988187, + "grad_norm": 2.6419613361358643, + "learning_rate": 1.284735439665017e-06, + "loss": 1.3183, + "step": 2261 + }, + { + "epoch": 2.3235747303543914, + "grad_norm": 2.7432727813720703, + "learning_rate": 1.2828321279025504e-06, + "loss": 1.3301, + "step": 2262 + }, + { + "epoch": 2.324601951720596, + "grad_norm": 2.7030932903289795, + "learning_rate": 1.2809288161400838e-06, + "loss": 1.3091, + "step": 2263 + }, + { + "epoch": 2.3256291730868, + "grad_norm": 2.621903896331787, + "learning_rate": 1.279025504377617e-06, + "loss": 1.2925, + "step": 2264 + }, + { + "epoch": 2.3266563944530048, + "grad_norm": 2.811548948287964, + "learning_rate": 1.2771221926151505e-06, + "loss": 1.3613, + "step": 2265 + }, + { + "epoch": 2.327683615819209, + "grad_norm": 2.543182134628296, + "learning_rate": 1.275218880852684e-06, + "loss": 1.2973, + "step": 2266 + }, + { + "epoch": 2.3287108371854135, + "grad_norm": 2.6845998764038086, + "learning_rate": 1.273315569090217e-06, + "loss": 1.2844, + "step": 2267 + }, + { + "epoch": 2.3297380585516176, + "grad_norm": 2.812823534011841, + "learning_rate": 1.2714122573277504e-06, + "loss": 1.3607, + "step": 2268 + }, + { + "epoch": 2.3307652799178222, + "grad_norm": 3.091813325881958, + "learning_rate": 1.2695089455652838e-06, + "loss": 1.3648, + "step": 2269 + }, + { + "epoch": 2.331792501284027, + "grad_norm": 2.7097294330596924, + "learning_rate": 1.267605633802817e-06, + "loss": 1.326, + "step": 2270 + }, + { + "epoch": 2.331792501284027, + "eval_loss": 1.791210412979126, + "eval_runtime": 38.0649, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 2270 + }, + { + "epoch": 2.332819722650231, + "grad_norm": 2.68989896774292, + "learning_rate": 1.2657023220403503e-06, + "loss": 1.3186, + "step": 2271 + }, + { + "epoch": 2.3338469440164356, + "grad_norm": 2.7425694465637207, + "learning_rate": 1.2637990102778836e-06, + "loss": 1.3014, + "step": 2272 + }, + { + "epoch": 2.33487416538264, + "grad_norm": 2.641493797302246, + "learning_rate": 1.2618956985154168e-06, + "loss": 1.3023, + "step": 2273 + }, + { + "epoch": 2.3359013867488443, + "grad_norm": 2.6478893756866455, + "learning_rate": 1.2599923867529502e-06, + "loss": 1.3508, + "step": 2274 + }, + { + "epoch": 2.336928608115049, + "grad_norm": 2.744187831878662, + "learning_rate": 1.2580890749904837e-06, + "loss": 1.3164, + "step": 2275 + }, + { + "epoch": 2.337955829481253, + "grad_norm": 2.823047161102295, + "learning_rate": 1.2561857632280169e-06, + "loss": 1.3418, + "step": 2276 + }, + { + "epoch": 2.3389830508474576, + "grad_norm": 2.9462826251983643, + "learning_rate": 1.2542824514655502e-06, + "loss": 1.3739, + "step": 2277 + }, + { + "epoch": 2.340010272213662, + "grad_norm": 2.7174856662750244, + "learning_rate": 1.2523791397030834e-06, + "loss": 1.3254, + "step": 2278 + }, + { + "epoch": 2.3410374935798663, + "grad_norm": 2.6504011154174805, + "learning_rate": 1.2504758279406168e-06, + "loss": 1.306, + "step": 2279 + }, + { + "epoch": 2.342064714946071, + "grad_norm": 2.701664447784424, + "learning_rate": 1.2485725161781501e-06, + "loss": 1.3302, + "step": 2280 + }, + { + "epoch": 2.342064714946071, + "eval_loss": 1.770766019821167, + "eval_runtime": 38.0551, + "eval_samples_per_second": 26.278, + "eval_steps_per_second": 1.104, + "step": 2280 + }, + { + "epoch": 2.343091936312275, + "grad_norm": 2.7651360034942627, + "learning_rate": 1.2466692044156833e-06, + "loss": 1.4107, + "step": 2281 + }, + { + "epoch": 2.3441191576784797, + "grad_norm": 2.60256028175354, + "learning_rate": 1.2447658926532168e-06, + "loss": 1.2769, + "step": 2282 + }, + { + "epoch": 2.3451463790446843, + "grad_norm": 2.6077115535736084, + "learning_rate": 1.24286258089075e-06, + "loss": 1.3148, + "step": 2283 + }, + { + "epoch": 2.3461736004108884, + "grad_norm": 2.6873903274536133, + "learning_rate": 1.2409592691282834e-06, + "loss": 1.2795, + "step": 2284 + }, + { + "epoch": 2.347200821777093, + "grad_norm": 2.9477450847625732, + "learning_rate": 1.2390559573658165e-06, + "loss": 1.3257, + "step": 2285 + }, + { + "epoch": 2.3482280431432976, + "grad_norm": 2.799549102783203, + "learning_rate": 1.2371526456033499e-06, + "loss": 1.3274, + "step": 2286 + }, + { + "epoch": 2.3492552645095017, + "grad_norm": 2.7725980281829834, + "learning_rate": 1.2352493338408832e-06, + "loss": 1.2803, + "step": 2287 + }, + { + "epoch": 2.3502824858757063, + "grad_norm": 3.027376413345337, + "learning_rate": 1.2333460220784166e-06, + "loss": 1.3907, + "step": 2288 + }, + { + "epoch": 2.3513097072419105, + "grad_norm": 2.7917747497558594, + "learning_rate": 1.23144271031595e-06, + "loss": 1.3458, + "step": 2289 + }, + { + "epoch": 2.352336928608115, + "grad_norm": 2.7375993728637695, + "learning_rate": 1.2295393985534831e-06, + "loss": 1.3496, + "step": 2290 + }, + { + "epoch": 2.352336928608115, + "eval_loss": 1.7680095434188843, + "eval_runtime": 38.0425, + "eval_samples_per_second": 26.286, + "eval_steps_per_second": 1.104, + "step": 2290 + }, + { + "epoch": 2.3533641499743196, + "grad_norm": 2.593547821044922, + "learning_rate": 1.2276360867910165e-06, + "loss": 1.2395, + "step": 2291 + }, + { + "epoch": 2.354391371340524, + "grad_norm": 2.7914979457855225, + "learning_rate": 1.2257327750285498e-06, + "loss": 1.3629, + "step": 2292 + }, + { + "epoch": 2.3554185927067284, + "grad_norm": 2.802846670150757, + "learning_rate": 1.2238294632660832e-06, + "loss": 1.2883, + "step": 2293 + }, + { + "epoch": 2.3564458140729325, + "grad_norm": 2.5035150051116943, + "learning_rate": 1.2219261515036164e-06, + "loss": 1.2894, + "step": 2294 + }, + { + "epoch": 2.357473035439137, + "grad_norm": 2.745405435562134, + "learning_rate": 1.2200228397411497e-06, + "loss": 1.3327, + "step": 2295 + }, + { + "epoch": 2.3585002568053417, + "grad_norm": 2.9135396480560303, + "learning_rate": 1.2181195279786829e-06, + "loss": 1.3636, + "step": 2296 + }, + { + "epoch": 2.359527478171546, + "grad_norm": 2.8527722358703613, + "learning_rate": 1.2162162162162164e-06, + "loss": 1.3059, + "step": 2297 + }, + { + "epoch": 2.3605546995377504, + "grad_norm": 2.641160249710083, + "learning_rate": 1.2143129044537496e-06, + "loss": 1.295, + "step": 2298 + }, + { + "epoch": 2.361581920903955, + "grad_norm": 2.6421735286712646, + "learning_rate": 1.212409592691283e-06, + "loss": 1.3052, + "step": 2299 + }, + { + "epoch": 2.362609142270159, + "grad_norm": 2.6424500942230225, + "learning_rate": 1.2105062809288163e-06, + "loss": 1.3108, + "step": 2300 + }, + { + "epoch": 2.362609142270159, + "eval_loss": 1.765913724899292, + "eval_runtime": 38.0324, + "eval_samples_per_second": 26.293, + "eval_steps_per_second": 1.104, + "step": 2300 + }, + { + "epoch": 2.3636363636363638, + "grad_norm": 2.814763307571411, + "learning_rate": 1.2086029691663495e-06, + "loss": 1.3234, + "step": 2301 + }, + { + "epoch": 2.364663585002568, + "grad_norm": 2.5212297439575195, + "learning_rate": 1.2066996574038828e-06, + "loss": 1.2656, + "step": 2302 + }, + { + "epoch": 2.3656908063687725, + "grad_norm": 2.8822665214538574, + "learning_rate": 1.2047963456414162e-06, + "loss": 1.3229, + "step": 2303 + }, + { + "epoch": 2.366718027734977, + "grad_norm": 2.6476027965545654, + "learning_rate": 1.2028930338789496e-06, + "loss": 1.3194, + "step": 2304 + }, + { + "epoch": 2.3677452491011812, + "grad_norm": 2.634824514389038, + "learning_rate": 1.2009897221164827e-06, + "loss": 1.2621, + "step": 2305 + }, + { + "epoch": 2.368772470467386, + "grad_norm": 2.7401492595672607, + "learning_rate": 1.199086410354016e-06, + "loss": 1.3255, + "step": 2306 + }, + { + "epoch": 2.36979969183359, + "grad_norm": 2.6971330642700195, + "learning_rate": 1.1971830985915492e-06, + "loss": 1.3003, + "step": 2307 + }, + { + "epoch": 2.3708269131997946, + "grad_norm": 2.7432987689971924, + "learning_rate": 1.1952797868290828e-06, + "loss": 1.2978, + "step": 2308 + }, + { + "epoch": 2.371854134565999, + "grad_norm": 2.7680413722991943, + "learning_rate": 1.193376475066616e-06, + "loss": 1.3168, + "step": 2309 + }, + { + "epoch": 2.3728813559322033, + "grad_norm": 2.7269768714904785, + "learning_rate": 1.1914731633041493e-06, + "loss": 1.3253, + "step": 2310 + }, + { + "epoch": 2.3728813559322033, + "eval_loss": 1.7663168907165527, + "eval_runtime": 37.7171, + "eval_samples_per_second": 26.513, + "eval_steps_per_second": 1.114, + "step": 2310 + }, + { + "epoch": 2.373908577298408, + "grad_norm": 2.6725943088531494, + "learning_rate": 1.1895698515416827e-06, + "loss": 1.2929, + "step": 2311 + }, + { + "epoch": 2.3749357986646125, + "grad_norm": 2.5516929626464844, + "learning_rate": 1.1876665397792158e-06, + "loss": 1.2936, + "step": 2312 + }, + { + "epoch": 2.3759630200308166, + "grad_norm": 2.7443480491638184, + "learning_rate": 1.1857632280167492e-06, + "loss": 1.3228, + "step": 2313 + }, + { + "epoch": 2.376990241397021, + "grad_norm": 2.7275967597961426, + "learning_rate": 1.1838599162542825e-06, + "loss": 1.2625, + "step": 2314 + }, + { + "epoch": 2.3780174627632253, + "grad_norm": 2.8085899353027344, + "learning_rate": 1.181956604491816e-06, + "loss": 1.3307, + "step": 2315 + }, + { + "epoch": 2.37904468412943, + "grad_norm": 2.6398675441741943, + "learning_rate": 1.180053292729349e-06, + "loss": 1.3144, + "step": 2316 + }, + { + "epoch": 2.380071905495634, + "grad_norm": 2.800874948501587, + "learning_rate": 1.1781499809668824e-06, + "loss": 1.3722, + "step": 2317 + }, + { + "epoch": 2.3810991268618387, + "grad_norm": 2.65503191947937, + "learning_rate": 1.1762466692044158e-06, + "loss": 1.3142, + "step": 2318 + }, + { + "epoch": 2.3821263482280433, + "grad_norm": 2.7648980617523193, + "learning_rate": 1.1743433574419491e-06, + "loss": 1.3473, + "step": 2319 + }, + { + "epoch": 2.3831535695942474, + "grad_norm": 2.656358242034912, + "learning_rate": 1.1724400456794823e-06, + "loss": 1.2912, + "step": 2320 + }, + { + "epoch": 2.3831535695942474, + "eval_loss": 1.7656192779541016, + "eval_runtime": 37.8498, + "eval_samples_per_second": 26.42, + "eval_steps_per_second": 1.11, + "step": 2320 + }, + { + "epoch": 2.384180790960452, + "grad_norm": 2.813964605331421, + "learning_rate": 1.1705367339170157e-06, + "loss": 1.3367, + "step": 2321 + }, + { + "epoch": 2.3852080123266566, + "grad_norm": 2.7210686206817627, + "learning_rate": 1.168633422154549e-06, + "loss": 1.3138, + "step": 2322 + }, + { + "epoch": 2.3862352336928607, + "grad_norm": 2.65240216255188, + "learning_rate": 1.1667301103920824e-06, + "loss": 1.2812, + "step": 2323 + }, + { + "epoch": 2.3872624550590653, + "grad_norm": 2.672790288925171, + "learning_rate": 1.1648267986296157e-06, + "loss": 1.2484, + "step": 2324 + }, + { + "epoch": 2.3882896764252695, + "grad_norm": 2.737819194793701, + "learning_rate": 1.162923486867149e-06, + "loss": 1.3027, + "step": 2325 + }, + { + "epoch": 2.389316897791474, + "grad_norm": 2.662963390350342, + "learning_rate": 1.1610201751046823e-06, + "loss": 1.3115, + "step": 2326 + }, + { + "epoch": 2.3903441191576786, + "grad_norm": 2.83809232711792, + "learning_rate": 1.1591168633422154e-06, + "loss": 1.3755, + "step": 2327 + }, + { + "epoch": 2.391371340523883, + "grad_norm": 2.722337484359741, + "learning_rate": 1.157213551579749e-06, + "loss": 1.3285, + "step": 2328 + }, + { + "epoch": 2.3923985618900874, + "grad_norm": 2.699765205383301, + "learning_rate": 1.1553102398172821e-06, + "loss": 1.3057, + "step": 2329 + }, + { + "epoch": 2.3934257832562915, + "grad_norm": 2.5885703563690186, + "learning_rate": 1.1534069280548155e-06, + "loss": 1.2963, + "step": 2330 + }, + { + "epoch": 2.3934257832562915, + "eval_loss": 1.7690107822418213, + "eval_runtime": 37.9639, + "eval_samples_per_second": 26.341, + "eval_steps_per_second": 1.106, + "step": 2330 + }, + { + "epoch": 2.394453004622496, + "grad_norm": 2.732694387435913, + "learning_rate": 1.1515036162923487e-06, + "loss": 1.3144, + "step": 2331 + }, + { + "epoch": 2.3954802259887007, + "grad_norm": 2.6117560863494873, + "learning_rate": 1.149600304529882e-06, + "loss": 1.283, + "step": 2332 + }, + { + "epoch": 2.396507447354905, + "grad_norm": 2.6792120933532715, + "learning_rate": 1.1476969927674154e-06, + "loss": 1.2959, + "step": 2333 + }, + { + "epoch": 2.3975346687211094, + "grad_norm": 2.7566049098968506, + "learning_rate": 1.1457936810049487e-06, + "loss": 1.309, + "step": 2334 + }, + { + "epoch": 2.398561890087314, + "grad_norm": 2.9152936935424805, + "learning_rate": 1.143890369242482e-06, + "loss": 1.3432, + "step": 2335 + }, + { + "epoch": 2.399589111453518, + "grad_norm": 2.6911866664886475, + "learning_rate": 1.1419870574800153e-06, + "loss": 1.2709, + "step": 2336 + }, + { + "epoch": 2.4006163328197228, + "grad_norm": 2.67596697807312, + "learning_rate": 1.1400837457175486e-06, + "loss": 1.2974, + "step": 2337 + }, + { + "epoch": 2.401643554185927, + "grad_norm": 2.5220348834991455, + "learning_rate": 1.138180433955082e-06, + "loss": 1.2691, + "step": 2338 + }, + { + "epoch": 2.4026707755521315, + "grad_norm": 2.7079286575317383, + "learning_rate": 1.1362771221926153e-06, + "loss": 1.3438, + "step": 2339 + }, + { + "epoch": 2.403697996918336, + "grad_norm": 2.5796804428100586, + "learning_rate": 1.1343738104301485e-06, + "loss": 1.3093, + "step": 2340 + }, + { + "epoch": 2.403697996918336, + "eval_loss": 1.7833822965621948, + "eval_runtime": 38.0179, + "eval_samples_per_second": 26.303, + "eval_steps_per_second": 1.105, + "step": 2340 + }, + { + "epoch": 2.4047252182845402, + "grad_norm": 2.533360242843628, + "learning_rate": 1.1324704986676819e-06, + "loss": 1.2876, + "step": 2341 + }, + { + "epoch": 2.405752439650745, + "grad_norm": 2.8825149536132812, + "learning_rate": 1.130567186905215e-06, + "loss": 1.3615, + "step": 2342 + }, + { + "epoch": 2.406779661016949, + "grad_norm": 2.5276525020599365, + "learning_rate": 1.1286638751427486e-06, + "loss": 1.2974, + "step": 2343 + }, + { + "epoch": 2.4078068823831535, + "grad_norm": 2.660332202911377, + "learning_rate": 1.1267605633802817e-06, + "loss": 1.3046, + "step": 2344 + }, + { + "epoch": 2.408834103749358, + "grad_norm": 2.81973934173584, + "learning_rate": 1.124857251617815e-06, + "loss": 1.329, + "step": 2345 + }, + { + "epoch": 2.4098613251155623, + "grad_norm": 2.6832475662231445, + "learning_rate": 1.1229539398553485e-06, + "loss": 1.3621, + "step": 2346 + }, + { + "epoch": 2.410888546481767, + "grad_norm": 2.5246076583862305, + "learning_rate": 1.1210506280928816e-06, + "loss": 1.3043, + "step": 2347 + }, + { + "epoch": 2.4119157678479715, + "grad_norm": 2.889202117919922, + "learning_rate": 1.119147316330415e-06, + "loss": 1.3412, + "step": 2348 + }, + { + "epoch": 2.4129429892141756, + "grad_norm": 2.761444091796875, + "learning_rate": 1.1172440045679483e-06, + "loss": 1.3069, + "step": 2349 + }, + { + "epoch": 2.41397021058038, + "grad_norm": 2.7676186561584473, + "learning_rate": 1.1153406928054817e-06, + "loss": 1.3057, + "step": 2350 + }, + { + "epoch": 2.41397021058038, + "eval_loss": 1.7964602708816528, + "eval_runtime": 38.0018, + "eval_samples_per_second": 26.315, + "eval_steps_per_second": 1.105, + "step": 2350 + }, + { + "epoch": 2.4149974319465843, + "grad_norm": 2.761202096939087, + "learning_rate": 1.1134373810430148e-06, + "loss": 1.3355, + "step": 2351 + }, + { + "epoch": 2.416024653312789, + "grad_norm": 2.74330735206604, + "learning_rate": 1.1115340692805482e-06, + "loss": 1.3415, + "step": 2352 + }, + { + "epoch": 2.4170518746789935, + "grad_norm": 2.6388099193573, + "learning_rate": 1.1096307575180816e-06, + "loss": 1.314, + "step": 2353 + }, + { + "epoch": 2.4180790960451977, + "grad_norm": 2.6058437824249268, + "learning_rate": 1.107727445755615e-06, + "loss": 1.2745, + "step": 2354 + }, + { + "epoch": 2.4191063174114023, + "grad_norm": 2.674837350845337, + "learning_rate": 1.105824133993148e-06, + "loss": 1.2943, + "step": 2355 + }, + { + "epoch": 2.4201335387776064, + "grad_norm": 2.777331829071045, + "learning_rate": 1.1039208222306814e-06, + "loss": 1.3136, + "step": 2356 + }, + { + "epoch": 2.421160760143811, + "grad_norm": 2.7025792598724365, + "learning_rate": 1.1020175104682148e-06, + "loss": 1.3051, + "step": 2357 + }, + { + "epoch": 2.4221879815100156, + "grad_norm": 2.570185899734497, + "learning_rate": 1.1001141987057482e-06, + "loss": 1.2964, + "step": 2358 + }, + { + "epoch": 2.4232152028762197, + "grad_norm": 2.6053781509399414, + "learning_rate": 1.0982108869432813e-06, + "loss": 1.3073, + "step": 2359 + }, + { + "epoch": 2.4242424242424243, + "grad_norm": 2.680983543395996, + "learning_rate": 1.0963075751808147e-06, + "loss": 1.3415, + "step": 2360 + }, + { + "epoch": 2.4242424242424243, + "eval_loss": 1.7697505950927734, + "eval_runtime": 37.9457, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 2360 + }, + { + "epoch": 2.425269645608629, + "grad_norm": 2.558598518371582, + "learning_rate": 1.094404263418348e-06, + "loss": 1.3074, + "step": 2361 + }, + { + "epoch": 2.426296866974833, + "grad_norm": 2.6231720447540283, + "learning_rate": 1.0925009516558812e-06, + "loss": 1.3031, + "step": 2362 + }, + { + "epoch": 2.4273240883410376, + "grad_norm": 2.730914831161499, + "learning_rate": 1.0905976398934148e-06, + "loss": 1.3017, + "step": 2363 + }, + { + "epoch": 2.428351309707242, + "grad_norm": 2.82137393951416, + "learning_rate": 1.088694328130948e-06, + "loss": 1.3154, + "step": 2364 + }, + { + "epoch": 2.4293785310734464, + "grad_norm": 2.6977908611297607, + "learning_rate": 1.0867910163684813e-06, + "loss": 1.2774, + "step": 2365 + }, + { + "epoch": 2.4304057524396505, + "grad_norm": 2.784348964691162, + "learning_rate": 1.0848877046060144e-06, + "loss": 1.3035, + "step": 2366 + }, + { + "epoch": 2.431432973805855, + "grad_norm": 2.663001537322998, + "learning_rate": 1.0829843928435478e-06, + "loss": 1.3026, + "step": 2367 + }, + { + "epoch": 2.4324601951720597, + "grad_norm": 2.7743303775787354, + "learning_rate": 1.0810810810810812e-06, + "loss": 1.3085, + "step": 2368 + }, + { + "epoch": 2.433487416538264, + "grad_norm": 2.797743082046509, + "learning_rate": 1.0791777693186145e-06, + "loss": 1.3254, + "step": 2369 + }, + { + "epoch": 2.4345146379044684, + "grad_norm": 2.625260829925537, + "learning_rate": 1.0772744575561479e-06, + "loss": 1.3098, + "step": 2370 + }, + { + "epoch": 2.4345146379044684, + "eval_loss": 1.7611181735992432, + "eval_runtime": 37.9845, + "eval_samples_per_second": 26.327, + "eval_steps_per_second": 1.106, + "step": 2370 + }, + { + "epoch": 2.435541859270673, + "grad_norm": 2.580679178237915, + "learning_rate": 1.075371145793681e-06, + "loss": 1.2925, + "step": 2371 + }, + { + "epoch": 2.436569080636877, + "grad_norm": 2.655032157897949, + "learning_rate": 1.0734678340312144e-06, + "loss": 1.3122, + "step": 2372 + }, + { + "epoch": 2.4375963020030817, + "grad_norm": 2.6039280891418457, + "learning_rate": 1.0715645222687478e-06, + "loss": 1.3037, + "step": 2373 + }, + { + "epoch": 2.438623523369286, + "grad_norm": 2.532874822616577, + "learning_rate": 1.0696612105062811e-06, + "loss": 1.2921, + "step": 2374 + }, + { + "epoch": 2.4396507447354905, + "grad_norm": 2.71315336227417, + "learning_rate": 1.0677578987438143e-06, + "loss": 1.3249, + "step": 2375 + }, + { + "epoch": 2.440677966101695, + "grad_norm": 2.797964334487915, + "learning_rate": 1.0658545869813476e-06, + "loss": 1.2849, + "step": 2376 + }, + { + "epoch": 2.441705187467899, + "grad_norm": 2.8070008754730225, + "learning_rate": 1.0639512752188808e-06, + "loss": 1.2956, + "step": 2377 + }, + { + "epoch": 2.442732408834104, + "grad_norm": 2.6451804637908936, + "learning_rate": 1.0620479634564144e-06, + "loss": 1.3076, + "step": 2378 + }, + { + "epoch": 2.443759630200308, + "grad_norm": 2.6175758838653564, + "learning_rate": 1.0601446516939475e-06, + "loss": 1.2766, + "step": 2379 + }, + { + "epoch": 2.4447868515665125, + "grad_norm": 2.5483524799346924, + "learning_rate": 1.0582413399314809e-06, + "loss": 1.2777, + "step": 2380 + }, + { + "epoch": 2.4447868515665125, + "eval_loss": 1.7605587244033813, + "eval_runtime": 37.9393, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 2380 + }, + { + "epoch": 2.445814072932717, + "grad_norm": 2.6982779502868652, + "learning_rate": 1.0563380281690142e-06, + "loss": 1.2826, + "step": 2381 + }, + { + "epoch": 2.4468412942989213, + "grad_norm": 2.575270891189575, + "learning_rate": 1.0544347164065474e-06, + "loss": 1.2855, + "step": 2382 + }, + { + "epoch": 2.447868515665126, + "grad_norm": 2.7389488220214844, + "learning_rate": 1.0525314046440808e-06, + "loss": 1.3179, + "step": 2383 + }, + { + "epoch": 2.4488957370313305, + "grad_norm": 2.555800199508667, + "learning_rate": 1.0506280928816141e-06, + "loss": 1.2919, + "step": 2384 + }, + { + "epoch": 2.4499229583975346, + "grad_norm": 2.580599546432495, + "learning_rate": 1.0487247811191475e-06, + "loss": 1.2943, + "step": 2385 + }, + { + "epoch": 2.450950179763739, + "grad_norm": 2.763772964477539, + "learning_rate": 1.0468214693566806e-06, + "loss": 1.3054, + "step": 2386 + }, + { + "epoch": 2.4519774011299433, + "grad_norm": 2.7483937740325928, + "learning_rate": 1.044918157594214e-06, + "loss": 1.3428, + "step": 2387 + }, + { + "epoch": 2.453004622496148, + "grad_norm": 2.644343376159668, + "learning_rate": 1.0430148458317474e-06, + "loss": 1.299, + "step": 2388 + }, + { + "epoch": 2.4540318438623525, + "grad_norm": 2.6011316776275635, + "learning_rate": 1.0411115340692807e-06, + "loss": 1.271, + "step": 2389 + }, + { + "epoch": 2.4550590652285567, + "grad_norm": 2.8445608615875244, + "learning_rate": 1.0392082223068139e-06, + "loss": 1.3426, + "step": 2390 + }, + { + "epoch": 2.4550590652285567, + "eval_loss": 1.764119267463684, + "eval_runtime": 37.9678, + "eval_samples_per_second": 26.338, + "eval_steps_per_second": 1.106, + "step": 2390 + }, + { + "epoch": 2.4560862865947612, + "grad_norm": 2.7232890129089355, + "learning_rate": 1.0373049105443472e-06, + "loss": 1.3083, + "step": 2391 + }, + { + "epoch": 2.4571135079609654, + "grad_norm": 2.7028329372406006, + "learning_rate": 1.0354015987818806e-06, + "loss": 1.3206, + "step": 2392 + }, + { + "epoch": 2.45814072932717, + "grad_norm": 2.7255351543426514, + "learning_rate": 1.033498287019414e-06, + "loss": 1.2851, + "step": 2393 + }, + { + "epoch": 2.4591679506933746, + "grad_norm": 2.9048759937286377, + "learning_rate": 1.0315949752569471e-06, + "loss": 1.335, + "step": 2394 + }, + { + "epoch": 2.4601951720595787, + "grad_norm": 2.680952310562134, + "learning_rate": 1.0296916634944805e-06, + "loss": 1.2832, + "step": 2395 + }, + { + "epoch": 2.4612223934257833, + "grad_norm": 2.6702933311462402, + "learning_rate": 1.0277883517320138e-06, + "loss": 1.3055, + "step": 2396 + }, + { + "epoch": 2.462249614791988, + "grad_norm": 2.851332664489746, + "learning_rate": 1.025885039969547e-06, + "loss": 1.3422, + "step": 2397 + }, + { + "epoch": 2.463276836158192, + "grad_norm": 2.5233068466186523, + "learning_rate": 1.0239817282070806e-06, + "loss": 1.2412, + "step": 2398 + }, + { + "epoch": 2.4643040575243966, + "grad_norm": 2.6488661766052246, + "learning_rate": 1.0220784164446137e-06, + "loss": 1.2725, + "step": 2399 + }, + { + "epoch": 2.4653312788906008, + "grad_norm": 2.558196544647217, + "learning_rate": 1.020175104682147e-06, + "loss": 1.3043, + "step": 2400 + }, + { + "epoch": 2.4653312788906008, + "eval_loss": 1.7678213119506836, + "eval_runtime": 37.9906, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 2400 + }, + { + "epoch": 2.4663585002568054, + "grad_norm": 2.601803779602051, + "learning_rate": 1.0182717929196802e-06, + "loss": 1.2979, + "step": 2401 + }, + { + "epoch": 2.4673857216230095, + "grad_norm": 2.6790895462036133, + "learning_rate": 1.0163684811572136e-06, + "loss": 1.3215, + "step": 2402 + }, + { + "epoch": 2.468412942989214, + "grad_norm": 2.6814472675323486, + "learning_rate": 1.014465169394747e-06, + "loss": 1.2561, + "step": 2403 + }, + { + "epoch": 2.4694401643554187, + "grad_norm": 2.7114996910095215, + "learning_rate": 1.0125618576322803e-06, + "loss": 1.2966, + "step": 2404 + }, + { + "epoch": 2.470467385721623, + "grad_norm": 2.6394622325897217, + "learning_rate": 1.0106585458698137e-06, + "loss": 1.278, + "step": 2405 + }, + { + "epoch": 2.4714946070878274, + "grad_norm": 2.656283140182495, + "learning_rate": 1.0087552341073468e-06, + "loss": 1.2798, + "step": 2406 + }, + { + "epoch": 2.472521828454032, + "grad_norm": 2.719639539718628, + "learning_rate": 1.0068519223448802e-06, + "loss": 1.3656, + "step": 2407 + }, + { + "epoch": 2.473549049820236, + "grad_norm": 2.7754440307617188, + "learning_rate": 1.0049486105824133e-06, + "loss": 1.3384, + "step": 2408 + }, + { + "epoch": 2.4745762711864407, + "grad_norm": 2.6474952697753906, + "learning_rate": 1.003045298819947e-06, + "loss": 1.2859, + "step": 2409 + }, + { + "epoch": 2.4756034925526453, + "grad_norm": 2.6622884273529053, + "learning_rate": 1.00114198705748e-06, + "loss": 1.2753, + "step": 2410 + }, + { + "epoch": 2.4756034925526453, + "eval_loss": 1.7663556337356567, + "eval_runtime": 37.6695, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 2410 + }, + { + "epoch": 2.4766307139188495, + "grad_norm": 2.721719741821289, + "learning_rate": 9.992386752950134e-07, + "loss": 1.2807, + "step": 2411 + }, + { + "epoch": 2.477657935285054, + "grad_norm": 2.5972471237182617, + "learning_rate": 9.973353635325466e-07, + "loss": 1.269, + "step": 2412 + }, + { + "epoch": 2.478685156651258, + "grad_norm": 2.616769790649414, + "learning_rate": 9.9543205177008e-07, + "loss": 1.2789, + "step": 2413 + }, + { + "epoch": 2.479712378017463, + "grad_norm": 2.573221206665039, + "learning_rate": 9.935287400076133e-07, + "loss": 1.282, + "step": 2414 + }, + { + "epoch": 2.480739599383667, + "grad_norm": 2.6644439697265625, + "learning_rate": 9.916254282451467e-07, + "loss": 1.2676, + "step": 2415 + }, + { + "epoch": 2.4817668207498715, + "grad_norm": 2.6775522232055664, + "learning_rate": 9.8972211648268e-07, + "loss": 1.2999, + "step": 2416 + }, + { + "epoch": 2.482794042116076, + "grad_norm": 2.785191297531128, + "learning_rate": 9.878188047202132e-07, + "loss": 1.3202, + "step": 2417 + }, + { + "epoch": 2.4838212634822803, + "grad_norm": 2.7016355991363525, + "learning_rate": 9.859154929577465e-07, + "loss": 1.295, + "step": 2418 + }, + { + "epoch": 2.484848484848485, + "grad_norm": 2.6089084148406982, + "learning_rate": 9.8401218119528e-07, + "loss": 1.2846, + "step": 2419 + }, + { + "epoch": 2.4858757062146895, + "grad_norm": 2.8124489784240723, + "learning_rate": 9.821088694328133e-07, + "loss": 1.2883, + "step": 2420 + }, + { + "epoch": 2.4858757062146895, + "eval_loss": 1.7576653957366943, + "eval_runtime": 37.867, + "eval_samples_per_second": 26.408, + "eval_steps_per_second": 1.109, + "step": 2420 + }, + { + "epoch": 2.4869029275808936, + "grad_norm": 2.7730648517608643, + "learning_rate": 9.802055576703464e-07, + "loss": 1.3321, + "step": 2421 + }, + { + "epoch": 2.487930148947098, + "grad_norm": 2.725834846496582, + "learning_rate": 9.783022459078798e-07, + "loss": 1.2931, + "step": 2422 + }, + { + "epoch": 2.4889573703133023, + "grad_norm": 2.607725143432617, + "learning_rate": 9.76398934145413e-07, + "loss": 1.3002, + "step": 2423 + }, + { + "epoch": 2.489984591679507, + "grad_norm": 2.6080846786499023, + "learning_rate": 9.744956223829465e-07, + "loss": 1.3102, + "step": 2424 + }, + { + "epoch": 2.4910118130457115, + "grad_norm": 2.6172633171081543, + "learning_rate": 9.725923106204797e-07, + "loss": 1.3234, + "step": 2425 + }, + { + "epoch": 2.4920390344119157, + "grad_norm": 2.6781582832336426, + "learning_rate": 9.70688998858013e-07, + "loss": 1.3442, + "step": 2426 + }, + { + "epoch": 2.4930662557781202, + "grad_norm": 2.8291099071502686, + "learning_rate": 9.687856870955464e-07, + "loss": 1.3401, + "step": 2427 + }, + { + "epoch": 2.4940934771443244, + "grad_norm": 2.589317560195923, + "learning_rate": 9.668823753330795e-07, + "loss": 1.3241, + "step": 2428 + }, + { + "epoch": 2.495120698510529, + "grad_norm": 2.5891406536102295, + "learning_rate": 9.64979063570613e-07, + "loss": 1.2589, + "step": 2429 + }, + { + "epoch": 2.4961479198767336, + "grad_norm": 2.589101791381836, + "learning_rate": 9.630757518081463e-07, + "loss": 1.3036, + "step": 2430 + }, + { + "epoch": 2.4961479198767336, + "eval_loss": 1.7622722387313843, + "eval_runtime": 38.0143, + "eval_samples_per_second": 26.306, + "eval_steps_per_second": 1.105, + "step": 2430 + }, + { + "epoch": 2.4971751412429377, + "grad_norm": 2.75529146194458, + "learning_rate": 9.611724400456796e-07, + "loss": 1.2969, + "step": 2431 + }, + { + "epoch": 2.4982023626091423, + "grad_norm": 2.557976722717285, + "learning_rate": 9.592691282832128e-07, + "loss": 1.2881, + "step": 2432 + }, + { + "epoch": 2.499229583975347, + "grad_norm": 2.638200521469116, + "learning_rate": 9.573658165207461e-07, + "loss": 1.3007, + "step": 2433 + }, + { + "epoch": 2.500256805341551, + "grad_norm": 2.6940360069274902, + "learning_rate": 9.554625047582795e-07, + "loss": 1.316, + "step": 2434 + }, + { + "epoch": 2.5012840267077556, + "grad_norm": 2.517789602279663, + "learning_rate": 9.535591929958129e-07, + "loss": 1.2844, + "step": 2435 + }, + { + "epoch": 2.50231124807396, + "grad_norm": 2.5775527954101562, + "learning_rate": 9.516558812333461e-07, + "loss": 1.3084, + "step": 2436 + }, + { + "epoch": 2.5033384694401644, + "grad_norm": 2.605240821838379, + "learning_rate": 9.497525694708794e-07, + "loss": 1.3242, + "step": 2437 + }, + { + "epoch": 2.5043656908063685, + "grad_norm": 2.56835675239563, + "learning_rate": 9.478492577084126e-07, + "loss": 1.2799, + "step": 2438 + }, + { + "epoch": 2.505392912172573, + "grad_norm": 2.564911127090454, + "learning_rate": 9.459459459459461e-07, + "loss": 1.2987, + "step": 2439 + }, + { + "epoch": 2.5064201335387777, + "grad_norm": 2.6561388969421387, + "learning_rate": 9.440426341834794e-07, + "loss": 1.29, + "step": 2440 + }, + { + "epoch": 2.5064201335387777, + "eval_loss": 1.766339898109436, + "eval_runtime": 38.0448, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 2440 + }, + { + "epoch": 2.507447354904982, + "grad_norm": 2.5807695388793945, + "learning_rate": 9.421393224210126e-07, + "loss": 1.3058, + "step": 2441 + }, + { + "epoch": 2.5084745762711864, + "grad_norm": 2.6242141723632812, + "learning_rate": 9.402360106585459e-07, + "loss": 1.3012, + "step": 2442 + }, + { + "epoch": 2.509501797637391, + "grad_norm": 2.6174166202545166, + "learning_rate": 9.383326988960792e-07, + "loss": 1.2716, + "step": 2443 + }, + { + "epoch": 2.510529019003595, + "grad_norm": 2.5244762897491455, + "learning_rate": 9.364293871336126e-07, + "loss": 1.2625, + "step": 2444 + }, + { + "epoch": 2.5115562403697997, + "grad_norm": 2.61104679107666, + "learning_rate": 9.345260753711458e-07, + "loss": 1.3049, + "step": 2445 + }, + { + "epoch": 2.5125834617360043, + "grad_norm": 2.735579013824463, + "learning_rate": 9.326227636086792e-07, + "loss": 1.3159, + "step": 2446 + }, + { + "epoch": 2.5136106831022085, + "grad_norm": 2.5533411502838135, + "learning_rate": 9.307194518462125e-07, + "loss": 1.2611, + "step": 2447 + }, + { + "epoch": 2.514637904468413, + "grad_norm": 2.705458164215088, + "learning_rate": 9.288161400837457e-07, + "loss": 1.3754, + "step": 2448 + }, + { + "epoch": 2.515665125834617, + "grad_norm": 2.5326483249664307, + "learning_rate": 9.269128283212792e-07, + "loss": 1.272, + "step": 2449 + }, + { + "epoch": 2.516692347200822, + "grad_norm": 2.6174025535583496, + "learning_rate": 9.250095165588125e-07, + "loss": 1.3067, + "step": 2450 + }, + { + "epoch": 2.516692347200822, + "eval_loss": 1.7679108381271362, + "eval_runtime": 38.0209, + "eval_samples_per_second": 26.301, + "eval_steps_per_second": 1.105, + "step": 2450 + }, + { + "epoch": 2.517719568567026, + "grad_norm": 2.6596055030822754, + "learning_rate": 9.231062047963457e-07, + "loss": 1.3485, + "step": 2451 + }, + { + "epoch": 2.5187467899332305, + "grad_norm": 2.7891628742218018, + "learning_rate": 9.21202893033879e-07, + "loss": 1.3559, + "step": 2452 + }, + { + "epoch": 2.519774011299435, + "grad_norm": 2.560696840286255, + "learning_rate": 9.192995812714123e-07, + "loss": 1.3187, + "step": 2453 + }, + { + "epoch": 2.5208012326656393, + "grad_norm": 2.6536800861358643, + "learning_rate": 9.173962695089457e-07, + "loss": 1.3129, + "step": 2454 + }, + { + "epoch": 2.521828454031844, + "grad_norm": 2.6149590015411377, + "learning_rate": 9.154929577464789e-07, + "loss": 1.3111, + "step": 2455 + }, + { + "epoch": 2.5228556753980484, + "grad_norm": 2.672088861465454, + "learning_rate": 9.135896459840123e-07, + "loss": 1.3172, + "step": 2456 + }, + { + "epoch": 2.5238828967642526, + "grad_norm": 2.7675814628601074, + "learning_rate": 9.116863342215456e-07, + "loss": 1.3029, + "step": 2457 + }, + { + "epoch": 2.524910118130457, + "grad_norm": 2.931568145751953, + "learning_rate": 9.097830224590788e-07, + "loss": 1.3424, + "step": 2458 + }, + { + "epoch": 2.5259373394966618, + "grad_norm": 2.6001431941986084, + "learning_rate": 9.078797106966122e-07, + "loss": 1.2604, + "step": 2459 + }, + { + "epoch": 2.526964560862866, + "grad_norm": 2.8028078079223633, + "learning_rate": 9.059763989341455e-07, + "loss": 1.3386, + "step": 2460 + }, + { + "epoch": 2.526964560862866, + "eval_loss": 1.7729586362838745, + "eval_runtime": 38.0609, + "eval_samples_per_second": 26.274, + "eval_steps_per_second": 1.103, + "step": 2460 + }, + { + "epoch": 2.5279917822290705, + "grad_norm": 2.6733815670013428, + "learning_rate": 9.040730871716788e-07, + "loss": 1.2824, + "step": 2461 + }, + { + "epoch": 2.5290190035952747, + "grad_norm": 2.6084516048431396, + "learning_rate": 9.021697754092121e-07, + "loss": 1.3075, + "step": 2462 + }, + { + "epoch": 2.5300462249614792, + "grad_norm": 2.6287853717803955, + "learning_rate": 9.002664636467453e-07, + "loss": 1.3112, + "step": 2463 + }, + { + "epoch": 2.5310734463276834, + "grad_norm": 2.702615976333618, + "learning_rate": 8.983631518842788e-07, + "loss": 1.3282, + "step": 2464 + }, + { + "epoch": 2.532100667693888, + "grad_norm": 2.60542893409729, + "learning_rate": 8.96459840121812e-07, + "loss": 1.3386, + "step": 2465 + }, + { + "epoch": 2.5331278890600926, + "grad_norm": 2.6633386611938477, + "learning_rate": 8.945565283593453e-07, + "loss": 1.3019, + "step": 2466 + }, + { + "epoch": 2.5341551104262967, + "grad_norm": 2.5968427658081055, + "learning_rate": 8.926532165968787e-07, + "loss": 1.3085, + "step": 2467 + }, + { + "epoch": 2.5351823317925013, + "grad_norm": 2.746767044067383, + "learning_rate": 8.907499048344119e-07, + "loss": 1.3233, + "step": 2468 + }, + { + "epoch": 2.536209553158706, + "grad_norm": 2.589085578918457, + "learning_rate": 8.888465930719453e-07, + "loss": 1.258, + "step": 2469 + }, + { + "epoch": 2.53723677452491, + "grad_norm": 2.61661434173584, + "learning_rate": 8.869432813094786e-07, + "loss": 1.2664, + "step": 2470 + }, + { + "epoch": 2.53723677452491, + "eval_loss": 1.7717350721359253, + "eval_runtime": 38.0362, + "eval_samples_per_second": 26.291, + "eval_steps_per_second": 1.104, + "step": 2470 + }, + { + "epoch": 2.5382639958911146, + "grad_norm": 2.657177448272705, + "learning_rate": 8.850399695470119e-07, + "loss": 1.3052, + "step": 2471 + }, + { + "epoch": 2.539291217257319, + "grad_norm": 2.603573799133301, + "learning_rate": 8.831366577845452e-07, + "loss": 1.3235, + "step": 2472 + }, + { + "epoch": 2.5403184386235234, + "grad_norm": 2.6154067516326904, + "learning_rate": 8.812333460220784e-07, + "loss": 1.2742, + "step": 2473 + }, + { + "epoch": 2.541345659989728, + "grad_norm": 2.515085220336914, + "learning_rate": 8.793300342596119e-07, + "loss": 1.2535, + "step": 2474 + }, + { + "epoch": 2.542372881355932, + "grad_norm": 2.6949517726898193, + "learning_rate": 8.774267224971451e-07, + "loss": 1.3088, + "step": 2475 + }, + { + "epoch": 2.5434001027221367, + "grad_norm": 2.891838312149048, + "learning_rate": 8.755234107346784e-07, + "loss": 1.3416, + "step": 2476 + }, + { + "epoch": 2.544427324088341, + "grad_norm": 2.623758554458618, + "learning_rate": 8.736200989722117e-07, + "loss": 1.2775, + "step": 2477 + }, + { + "epoch": 2.5454545454545454, + "grad_norm": 2.65680193901062, + "learning_rate": 8.71716787209745e-07, + "loss": 1.2999, + "step": 2478 + }, + { + "epoch": 2.54648176682075, + "grad_norm": 2.85903000831604, + "learning_rate": 8.698134754472784e-07, + "loss": 1.3453, + "step": 2479 + }, + { + "epoch": 2.547508988186954, + "grad_norm": 2.7487850189208984, + "learning_rate": 8.679101636848116e-07, + "loss": 1.2432, + "step": 2480 + }, + { + "epoch": 2.547508988186954, + "eval_loss": 1.7745047807693481, + "eval_runtime": 37.9926, + "eval_samples_per_second": 26.321, + "eval_steps_per_second": 1.105, + "step": 2480 + }, + { + "epoch": 2.5485362095531587, + "grad_norm": 2.737975835800171, + "learning_rate": 8.66006851922345e-07, + "loss": 1.3116, + "step": 2481 + }, + { + "epoch": 2.5495634309193633, + "grad_norm": 2.6670281887054443, + "learning_rate": 8.641035401598783e-07, + "loss": 1.2989, + "step": 2482 + }, + { + "epoch": 2.5505906522855675, + "grad_norm": 2.6732940673828125, + "learning_rate": 8.622002283974115e-07, + "loss": 1.2907, + "step": 2483 + }, + { + "epoch": 2.551617873651772, + "grad_norm": 2.7414166927337646, + "learning_rate": 8.60296916634945e-07, + "loss": 1.3347, + "step": 2484 + }, + { + "epoch": 2.5526450950179767, + "grad_norm": 2.557168960571289, + "learning_rate": 8.583936048724782e-07, + "loss": 1.2939, + "step": 2485 + }, + { + "epoch": 2.553672316384181, + "grad_norm": 2.5460598468780518, + "learning_rate": 8.564902931100115e-07, + "loss": 1.285, + "step": 2486 + }, + { + "epoch": 2.554699537750385, + "grad_norm": 2.7185349464416504, + "learning_rate": 8.545869813475448e-07, + "loss": 1.3066, + "step": 2487 + }, + { + "epoch": 2.5557267591165895, + "grad_norm": 2.6216139793395996, + "learning_rate": 8.526836695850781e-07, + "loss": 1.3032, + "step": 2488 + }, + { + "epoch": 2.556753980482794, + "grad_norm": 2.7842869758605957, + "learning_rate": 8.507803578226115e-07, + "loss": 1.3279, + "step": 2489 + }, + { + "epoch": 2.5577812018489983, + "grad_norm": 2.6502983570098877, + "learning_rate": 8.488770460601447e-07, + "loss": 1.3119, + "step": 2490 + }, + { + "epoch": 2.5577812018489983, + "eval_loss": 1.7768635749816895, + "eval_runtime": 37.9867, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 2490 + }, + { + "epoch": 2.558808423215203, + "grad_norm": 2.695314645767212, + "learning_rate": 8.46973734297678e-07, + "loss": 1.2769, + "step": 2491 + }, + { + "epoch": 2.5598356445814074, + "grad_norm": 2.661893844604492, + "learning_rate": 8.450704225352114e-07, + "loss": 1.2892, + "step": 2492 + }, + { + "epoch": 2.5608628659476116, + "grad_norm": 2.686737537384033, + "learning_rate": 8.431671107727446e-07, + "loss": 1.2883, + "step": 2493 + }, + { + "epoch": 2.561890087313816, + "grad_norm": 2.661504030227661, + "learning_rate": 8.41263799010278e-07, + "loss": 1.3139, + "step": 2494 + }, + { + "epoch": 2.5629173086800208, + "grad_norm": 2.6665518283843994, + "learning_rate": 8.393604872478113e-07, + "loss": 1.2981, + "step": 2495 + }, + { + "epoch": 2.563944530046225, + "grad_norm": 2.722047805786133, + "learning_rate": 8.374571754853446e-07, + "loss": 1.2862, + "step": 2496 + }, + { + "epoch": 2.5649717514124295, + "grad_norm": 2.6358418464660645, + "learning_rate": 8.355538637228778e-07, + "loss": 1.2934, + "step": 2497 + }, + { + "epoch": 2.5659989727786336, + "grad_norm": 2.7137503623962402, + "learning_rate": 8.336505519604111e-07, + "loss": 1.2874, + "step": 2498 + }, + { + "epoch": 2.5670261941448382, + "grad_norm": 2.589345693588257, + "learning_rate": 8.317472401979445e-07, + "loss": 1.3229, + "step": 2499 + }, + { + "epoch": 2.5680534155110424, + "grad_norm": 2.597588062286377, + "learning_rate": 8.298439284354778e-07, + "loss": 1.3481, + "step": 2500 + }, + { + "epoch": 2.5680534155110424, + "eval_loss": 1.7751270532608032, + "eval_runtime": 37.9958, + "eval_samples_per_second": 26.319, + "eval_steps_per_second": 1.105, + "step": 2500 + }, + { + "epoch": 2.569080636877247, + "grad_norm": 2.7178173065185547, + "learning_rate": 8.279406166730111e-07, + "loss": 1.3318, + "step": 2501 + }, + { + "epoch": 2.5701078582434516, + "grad_norm": 2.6459460258483887, + "learning_rate": 8.260373049105444e-07, + "loss": 1.332, + "step": 2502 + }, + { + "epoch": 2.5711350796096557, + "grad_norm": 2.6359541416168213, + "learning_rate": 8.241339931480777e-07, + "loss": 1.311, + "step": 2503 + }, + { + "epoch": 2.5721623009758603, + "grad_norm": 2.706570863723755, + "learning_rate": 8.22230681385611e-07, + "loss": 1.3391, + "step": 2504 + }, + { + "epoch": 2.573189522342065, + "grad_norm": 2.591161012649536, + "learning_rate": 8.203273696231444e-07, + "loss": 1.3058, + "step": 2505 + }, + { + "epoch": 2.574216743708269, + "grad_norm": 2.595672607421875, + "learning_rate": 8.184240578606777e-07, + "loss": 1.2646, + "step": 2506 + }, + { + "epoch": 2.5752439650744736, + "grad_norm": 2.7604939937591553, + "learning_rate": 8.165207460982109e-07, + "loss": 1.3191, + "step": 2507 + }, + { + "epoch": 2.576271186440678, + "grad_norm": 2.6298060417175293, + "learning_rate": 8.146174343357442e-07, + "loss": 1.2735, + "step": 2508 + }, + { + "epoch": 2.5772984078068824, + "grad_norm": 2.7220308780670166, + "learning_rate": 8.127141225732775e-07, + "loss": 1.3164, + "step": 2509 + }, + { + "epoch": 2.578325629173087, + "grad_norm": 2.875459671020508, + "learning_rate": 8.108108108108109e-07, + "loss": 1.3236, + "step": 2510 + }, + { + "epoch": 2.578325629173087, + "eval_loss": 1.7866631746292114, + "eval_runtime": 37.6739, + "eval_samples_per_second": 26.544, + "eval_steps_per_second": 1.115, + "step": 2510 + }, + { + "epoch": 2.579352850539291, + "grad_norm": 2.747443675994873, + "learning_rate": 8.089074990483442e-07, + "loss": 1.3147, + "step": 2511 + }, + { + "epoch": 2.5803800719054957, + "grad_norm": 2.7403504848480225, + "learning_rate": 8.070041872858774e-07, + "loss": 1.3016, + "step": 2512 + }, + { + "epoch": 2.5814072932717, + "grad_norm": 2.700188159942627, + "learning_rate": 8.051008755234108e-07, + "loss": 1.3118, + "step": 2513 + }, + { + "epoch": 2.5824345146379044, + "grad_norm": 2.575113296508789, + "learning_rate": 8.031975637609441e-07, + "loss": 1.255, + "step": 2514 + }, + { + "epoch": 2.583461736004109, + "grad_norm": 2.7600820064544678, + "learning_rate": 8.012942519984774e-07, + "loss": 1.305, + "step": 2515 + }, + { + "epoch": 2.584488957370313, + "grad_norm": 2.5697903633117676, + "learning_rate": 7.993909402360108e-07, + "loss": 1.327, + "step": 2516 + }, + { + "epoch": 2.5855161787365177, + "grad_norm": 2.617521047592163, + "learning_rate": 7.97487628473544e-07, + "loss": 1.3196, + "step": 2517 + }, + { + "epoch": 2.5865434001027223, + "grad_norm": 2.708566427230835, + "learning_rate": 7.955843167110773e-07, + "loss": 1.2993, + "step": 2518 + }, + { + "epoch": 2.5875706214689265, + "grad_norm": 2.683959722518921, + "learning_rate": 7.936810049486106e-07, + "loss": 1.2748, + "step": 2519 + }, + { + "epoch": 2.588597842835131, + "grad_norm": 2.6184964179992676, + "learning_rate": 7.91777693186144e-07, + "loss": 1.2935, + "step": 2520 + }, + { + "epoch": 2.588597842835131, + "eval_loss": 1.7722079753875732, + "eval_runtime": 37.7975, + "eval_samples_per_second": 26.457, + "eval_steps_per_second": 1.111, + "step": 2520 + }, + { + "epoch": 2.5896250642013356, + "grad_norm": 2.538283586502075, + "learning_rate": 7.898743814236773e-07, + "loss": 1.3149, + "step": 2521 + }, + { + "epoch": 2.59065228556754, + "grad_norm": 2.5143439769744873, + "learning_rate": 7.879710696612105e-07, + "loss": 1.2907, + "step": 2522 + }, + { + "epoch": 2.5916795069337444, + "grad_norm": 2.5969698429107666, + "learning_rate": 7.860677578987438e-07, + "loss": 1.3575, + "step": 2523 + }, + { + "epoch": 2.5927067282999485, + "grad_norm": 2.617640733718872, + "learning_rate": 7.841644461362772e-07, + "loss": 1.3152, + "step": 2524 + }, + { + "epoch": 2.593733949666153, + "grad_norm": 2.649030923843384, + "learning_rate": 7.822611343738105e-07, + "loss": 1.2949, + "step": 2525 + }, + { + "epoch": 2.5947611710323573, + "grad_norm": 2.694195032119751, + "learning_rate": 7.803578226113438e-07, + "loss": 1.3333, + "step": 2526 + }, + { + "epoch": 2.595788392398562, + "grad_norm": 2.559192657470703, + "learning_rate": 7.784545108488771e-07, + "loss": 1.282, + "step": 2527 + }, + { + "epoch": 2.5968156137647664, + "grad_norm": 2.5788307189941406, + "learning_rate": 7.765511990864104e-07, + "loss": 1.3033, + "step": 2528 + }, + { + "epoch": 2.5978428351309706, + "grad_norm": 2.8756697177886963, + "learning_rate": 7.746478873239437e-07, + "loss": 1.3569, + "step": 2529 + }, + { + "epoch": 2.598870056497175, + "grad_norm": 2.632779598236084, + "learning_rate": 7.727445755614771e-07, + "loss": 1.319, + "step": 2530 + }, + { + "epoch": 2.598870056497175, + "eval_loss": 1.7722663879394531, + "eval_runtime": 37.9166, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 2530 + }, + { + "epoch": 2.5998972778633798, + "grad_norm": 2.76119065284729, + "learning_rate": 7.708412637990104e-07, + "loss": 1.3142, + "step": 2531 + }, + { + "epoch": 2.600924499229584, + "grad_norm": 2.593604803085327, + "learning_rate": 7.689379520365436e-07, + "loss": 1.3286, + "step": 2532 + }, + { + "epoch": 2.6019517205957885, + "grad_norm": 2.6926052570343018, + "learning_rate": 7.670346402740769e-07, + "loss": 1.2933, + "step": 2533 + }, + { + "epoch": 2.602978941961993, + "grad_norm": 2.734140157699585, + "learning_rate": 7.651313285116103e-07, + "loss": 1.3187, + "step": 2534 + }, + { + "epoch": 2.6040061633281972, + "grad_norm": 2.7742362022399902, + "learning_rate": 7.632280167491436e-07, + "loss": 1.3427, + "step": 2535 + }, + { + "epoch": 2.6050333846944014, + "grad_norm": 2.5474061965942383, + "learning_rate": 7.613247049866769e-07, + "loss": 1.2652, + "step": 2536 + }, + { + "epoch": 2.606060606060606, + "grad_norm": 2.704885959625244, + "learning_rate": 7.594213932242102e-07, + "loss": 1.2405, + "step": 2537 + }, + { + "epoch": 2.6070878274268106, + "grad_norm": 2.67887020111084, + "learning_rate": 7.575180814617435e-07, + "loss": 1.2702, + "step": 2538 + }, + { + "epoch": 2.6081150487930147, + "grad_norm": 2.6307291984558105, + "learning_rate": 7.556147696992767e-07, + "loss": 1.3046, + "step": 2539 + }, + { + "epoch": 2.6091422701592193, + "grad_norm": 2.7046597003936768, + "learning_rate": 7.537114579368102e-07, + "loss": 1.3121, + "step": 2540 + }, + { + "epoch": 2.6091422701592193, + "eval_loss": 1.7705127000808716, + "eval_runtime": 38.0131, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 2540 + }, + { + "epoch": 2.610169491525424, + "grad_norm": 2.564473867416382, + "learning_rate": 7.518081461743435e-07, + "loss": 1.28, + "step": 2541 + }, + { + "epoch": 2.611196712891628, + "grad_norm": 2.63580322265625, + "learning_rate": 7.499048344118767e-07, + "loss": 1.2945, + "step": 2542 + }, + { + "epoch": 2.6122239342578326, + "grad_norm": 2.6448638439178467, + "learning_rate": 7.4800152264941e-07, + "loss": 1.3265, + "step": 2543 + }, + { + "epoch": 2.613251155624037, + "grad_norm": 2.490447521209717, + "learning_rate": 7.460982108869432e-07, + "loss": 1.3035, + "step": 2544 + }, + { + "epoch": 2.6142783769902413, + "grad_norm": 2.5429506301879883, + "learning_rate": 7.441948991244767e-07, + "loss": 1.288, + "step": 2545 + }, + { + "epoch": 2.615305598356446, + "grad_norm": 2.6845016479492188, + "learning_rate": 7.4229158736201e-07, + "loss": 1.3119, + "step": 2546 + }, + { + "epoch": 2.61633281972265, + "grad_norm": 2.6247217655181885, + "learning_rate": 7.403882755995432e-07, + "loss": 1.3213, + "step": 2547 + }, + { + "epoch": 2.6173600410888547, + "grad_norm": 2.5123143196105957, + "learning_rate": 7.384849638370766e-07, + "loss": 1.2831, + "step": 2548 + }, + { + "epoch": 2.618387262455059, + "grad_norm": 2.5071399211883545, + "learning_rate": 7.365816520746098e-07, + "loss": 1.2393, + "step": 2549 + }, + { + "epoch": 2.6194144838212634, + "grad_norm": 2.744621515274048, + "learning_rate": 7.346783403121432e-07, + "loss": 1.3083, + "step": 2550 + }, + { + "epoch": 2.6194144838212634, + "eval_loss": 1.7846120595932007, + "eval_runtime": 37.9987, + "eval_samples_per_second": 26.317, + "eval_steps_per_second": 1.105, + "step": 2550 + }, + { + "epoch": 2.620441705187468, + "grad_norm": 2.714101552963257, + "learning_rate": 7.327750285496766e-07, + "loss": 1.287, + "step": 2551 + }, + { + "epoch": 2.621468926553672, + "grad_norm": 2.792046546936035, + "learning_rate": 7.308717167872098e-07, + "loss": 1.2873, + "step": 2552 + }, + { + "epoch": 2.6224961479198767, + "grad_norm": 2.493668556213379, + "learning_rate": 7.289684050247431e-07, + "loss": 1.255, + "step": 2553 + }, + { + "epoch": 2.6235233692860813, + "grad_norm": 2.631716012954712, + "learning_rate": 7.270650932622763e-07, + "loss": 1.3418, + "step": 2554 + }, + { + "epoch": 2.6245505906522855, + "grad_norm": 2.443617820739746, + "learning_rate": 7.251617814998098e-07, + "loss": 1.2512, + "step": 2555 + }, + { + "epoch": 2.62557781201849, + "grad_norm": 2.4819979667663574, + "learning_rate": 7.232584697373431e-07, + "loss": 1.3131, + "step": 2556 + }, + { + "epoch": 2.6266050333846946, + "grad_norm": 2.5260822772979736, + "learning_rate": 7.213551579748763e-07, + "loss": 1.2812, + "step": 2557 + }, + { + "epoch": 2.627632254750899, + "grad_norm": 2.614377498626709, + "learning_rate": 7.194518462124096e-07, + "loss": 1.3065, + "step": 2558 + }, + { + "epoch": 2.6286594761171034, + "grad_norm": 2.619166374206543, + "learning_rate": 7.175485344499429e-07, + "loss": 1.3813, + "step": 2559 + }, + { + "epoch": 2.6296866974833075, + "grad_norm": 2.6223137378692627, + "learning_rate": 7.156452226874763e-07, + "loss": 1.2856, + "step": 2560 + }, + { + "epoch": 2.6296866974833075, + "eval_loss": 1.7834736108779907, + "eval_runtime": 38.0086, + "eval_samples_per_second": 26.31, + "eval_steps_per_second": 1.105, + "step": 2560 + }, + { + "epoch": 2.630713918849512, + "grad_norm": 2.7283148765563965, + "learning_rate": 7.137419109250096e-07, + "loss": 1.375, + "step": 2561 + }, + { + "epoch": 2.6317411402157163, + "grad_norm": 2.629539728164673, + "learning_rate": 7.118385991625429e-07, + "loss": 1.2915, + "step": 2562 + }, + { + "epoch": 2.632768361581921, + "grad_norm": 2.576698064804077, + "learning_rate": 7.099352874000762e-07, + "loss": 1.2818, + "step": 2563 + }, + { + "epoch": 2.6337955829481254, + "grad_norm": 2.5807430744171143, + "learning_rate": 7.080319756376094e-07, + "loss": 1.285, + "step": 2564 + }, + { + "epoch": 2.6348228043143296, + "grad_norm": 2.461386203765869, + "learning_rate": 7.061286638751429e-07, + "loss": 1.2594, + "step": 2565 + }, + { + "epoch": 2.635850025680534, + "grad_norm": 2.6482107639312744, + "learning_rate": 7.042253521126762e-07, + "loss": 1.3038, + "step": 2566 + }, + { + "epoch": 2.6368772470467388, + "grad_norm": 2.637989044189453, + "learning_rate": 7.023220403502094e-07, + "loss": 1.2453, + "step": 2567 + }, + { + "epoch": 2.637904468412943, + "grad_norm": 2.6575927734375, + "learning_rate": 7.004187285877427e-07, + "loss": 1.3413, + "step": 2568 + }, + { + "epoch": 2.6389316897791475, + "grad_norm": 2.616055727005005, + "learning_rate": 6.98515416825276e-07, + "loss": 1.3212, + "step": 2569 + }, + { + "epoch": 2.639958911145352, + "grad_norm": 2.5758297443389893, + "learning_rate": 6.966121050628094e-07, + "loss": 1.2999, + "step": 2570 + }, + { + "epoch": 2.639958911145352, + "eval_loss": 1.7819379568099976, + "eval_runtime": 38.0358, + "eval_samples_per_second": 26.291, + "eval_steps_per_second": 1.104, + "step": 2570 + }, + { + "epoch": 2.6409861325115562, + "grad_norm": 2.7511165142059326, + "learning_rate": 6.947087933003427e-07, + "loss": 1.3519, + "step": 2571 + }, + { + "epoch": 2.642013353877761, + "grad_norm": 2.6978421211242676, + "learning_rate": 6.92805481537876e-07, + "loss": 1.2944, + "step": 2572 + }, + { + "epoch": 2.643040575243965, + "grad_norm": 2.7035439014434814, + "learning_rate": 6.909021697754093e-07, + "loss": 1.3496, + "step": 2573 + }, + { + "epoch": 2.6440677966101696, + "grad_norm": 2.6517820358276367, + "learning_rate": 6.889988580129425e-07, + "loss": 1.3233, + "step": 2574 + }, + { + "epoch": 2.6450950179763737, + "grad_norm": 2.65956974029541, + "learning_rate": 6.870955462504759e-07, + "loss": 1.2826, + "step": 2575 + }, + { + "epoch": 2.6461222393425783, + "grad_norm": 2.671339988708496, + "learning_rate": 6.851922344880093e-07, + "loss": 1.3257, + "step": 2576 + }, + { + "epoch": 2.647149460708783, + "grad_norm": 2.648261785507202, + "learning_rate": 6.832889227255425e-07, + "loss": 1.3015, + "step": 2577 + }, + { + "epoch": 2.648176682074987, + "grad_norm": 2.700082540512085, + "learning_rate": 6.813856109630758e-07, + "loss": 1.3129, + "step": 2578 + }, + { + "epoch": 2.6492039034411916, + "grad_norm": 2.7364165782928467, + "learning_rate": 6.79482299200609e-07, + "loss": 1.3185, + "step": 2579 + }, + { + "epoch": 2.650231124807396, + "grad_norm": 2.5831615924835205, + "learning_rate": 6.775789874381425e-07, + "loss": 1.2796, + "step": 2580 + }, + { + "epoch": 2.650231124807396, + "eval_loss": 1.7886762619018555, + "eval_runtime": 38.0048, + "eval_samples_per_second": 26.312, + "eval_steps_per_second": 1.105, + "step": 2580 + }, + { + "epoch": 2.6512583461736003, + "grad_norm": 2.6888351440429688, + "learning_rate": 6.756756756756758e-07, + "loss": 1.3447, + "step": 2581 + }, + { + "epoch": 2.652285567539805, + "grad_norm": 2.643916606903076, + "learning_rate": 6.73772363913209e-07, + "loss": 1.3234, + "step": 2582 + }, + { + "epoch": 2.653312788906009, + "grad_norm": 2.593222141265869, + "learning_rate": 6.718690521507424e-07, + "loss": 1.3164, + "step": 2583 + }, + { + "epoch": 2.6543400102722137, + "grad_norm": 2.535433053970337, + "learning_rate": 6.699657403882756e-07, + "loss": 1.2637, + "step": 2584 + }, + { + "epoch": 2.655367231638418, + "grad_norm": 2.5683724880218506, + "learning_rate": 6.68062428625809e-07, + "loss": 1.328, + "step": 2585 + }, + { + "epoch": 2.6563944530046224, + "grad_norm": 2.6119470596313477, + "learning_rate": 6.661591168633424e-07, + "loss": 1.3393, + "step": 2586 + }, + { + "epoch": 2.657421674370827, + "grad_norm": 2.646026134490967, + "learning_rate": 6.642558051008756e-07, + "loss": 1.3219, + "step": 2587 + }, + { + "epoch": 2.658448895737031, + "grad_norm": 2.7056374549865723, + "learning_rate": 6.623524933384089e-07, + "loss": 1.326, + "step": 2588 + }, + { + "epoch": 2.6594761171032357, + "grad_norm": 2.638749361038208, + "learning_rate": 6.604491815759421e-07, + "loss": 1.2979, + "step": 2589 + }, + { + "epoch": 2.6605033384694403, + "grad_norm": 2.6330184936523438, + "learning_rate": 6.585458698134756e-07, + "loss": 1.292, + "step": 2590 + }, + { + "epoch": 2.6605033384694403, + "eval_loss": 1.7978336811065674, + "eval_runtime": 38.019, + "eval_samples_per_second": 26.303, + "eval_steps_per_second": 1.105, + "step": 2590 + }, + { + "epoch": 2.6615305598356445, + "grad_norm": 2.577609062194824, + "learning_rate": 6.566425580510089e-07, + "loss": 1.2541, + "step": 2591 + }, + { + "epoch": 2.662557781201849, + "grad_norm": 2.666790008544922, + "learning_rate": 6.547392462885421e-07, + "loss": 1.3183, + "step": 2592 + }, + { + "epoch": 2.6635850025680536, + "grad_norm": 2.6628904342651367, + "learning_rate": 6.528359345260754e-07, + "loss": 1.3303, + "step": 2593 + }, + { + "epoch": 2.664612223934258, + "grad_norm": 2.6553924083709717, + "learning_rate": 6.509326227636087e-07, + "loss": 1.3019, + "step": 2594 + }, + { + "epoch": 2.6656394453004624, + "grad_norm": 2.7442100048065186, + "learning_rate": 6.490293110011421e-07, + "loss": 1.3111, + "step": 2595 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 2.6254689693450928, + "learning_rate": 6.471259992386753e-07, + "loss": 1.2986, + "step": 2596 + }, + { + "epoch": 2.667693888032871, + "grad_norm": 2.5442848205566406, + "learning_rate": 6.452226874762087e-07, + "loss": 1.3471, + "step": 2597 + }, + { + "epoch": 2.6687211093990753, + "grad_norm": 2.5967025756835938, + "learning_rate": 6.43319375713742e-07, + "loss": 1.2879, + "step": 2598 + }, + { + "epoch": 2.66974833076528, + "grad_norm": 2.69215989112854, + "learning_rate": 6.414160639512752e-07, + "loss": 1.3201, + "step": 2599 + }, + { + "epoch": 2.6707755521314844, + "grad_norm": 2.7470815181732178, + "learning_rate": 6.395127521888085e-07, + "loss": 1.31, + "step": 2600 + }, + { + "epoch": 2.6707755521314844, + "eval_loss": 1.8076120615005493, + "eval_runtime": 38.023, + "eval_samples_per_second": 26.3, + "eval_steps_per_second": 1.105, + "step": 2600 + }, + { + "epoch": 2.6718027734976886, + "grad_norm": 2.681523561477661, + "learning_rate": 6.37609440426342e-07, + "loss": 1.2818, + "step": 2601 + }, + { + "epoch": 2.672829994863893, + "grad_norm": 2.7073915004730225, + "learning_rate": 6.357061286638752e-07, + "loss": 1.3465, + "step": 2602 + }, + { + "epoch": 2.6738572162300978, + "grad_norm": 2.7676303386688232, + "learning_rate": 6.338028169014085e-07, + "loss": 1.3072, + "step": 2603 + }, + { + "epoch": 2.674884437596302, + "grad_norm": 2.7673637866973877, + "learning_rate": 6.318995051389418e-07, + "loss": 1.3415, + "step": 2604 + }, + { + "epoch": 2.6759116589625065, + "grad_norm": 2.518784284591675, + "learning_rate": 6.299961933764751e-07, + "loss": 1.2292, + "step": 2605 + }, + { + "epoch": 2.676938880328711, + "grad_norm": 2.569060802459717, + "learning_rate": 6.280928816140084e-07, + "loss": 1.3314, + "step": 2606 + }, + { + "epoch": 2.6779661016949152, + "grad_norm": 2.70438814163208, + "learning_rate": 6.261895698515417e-07, + "loss": 1.3806, + "step": 2607 + }, + { + "epoch": 2.67899332306112, + "grad_norm": 2.6343674659729004, + "learning_rate": 6.242862580890751e-07, + "loss": 1.326, + "step": 2608 + }, + { + "epoch": 2.680020544427324, + "grad_norm": 2.5554616451263428, + "learning_rate": 6.223829463266084e-07, + "loss": 1.2799, + "step": 2609 + }, + { + "epoch": 2.6810477657935285, + "grad_norm": 2.6620726585388184, + "learning_rate": 6.204796345641417e-07, + "loss": 1.2466, + "step": 2610 + }, + { + "epoch": 2.6810477657935285, + "eval_loss": 1.8074779510498047, + "eval_runtime": 37.6679, + "eval_samples_per_second": 26.548, + "eval_steps_per_second": 1.115, + "step": 2610 + }, + { + "epoch": 2.6820749871597327, + "grad_norm": 2.7202115058898926, + "learning_rate": 6.185763228016749e-07, + "loss": 1.3034, + "step": 2611 + }, + { + "epoch": 2.6831022085259373, + "grad_norm": 2.912074089050293, + "learning_rate": 6.166730110392083e-07, + "loss": 1.3341, + "step": 2612 + }, + { + "epoch": 2.684129429892142, + "grad_norm": 3.0352418422698975, + "learning_rate": 6.147696992767416e-07, + "loss": 1.3551, + "step": 2613 + }, + { + "epoch": 2.685156651258346, + "grad_norm": 2.603231191635132, + "learning_rate": 6.128663875142749e-07, + "loss": 1.3205, + "step": 2614 + }, + { + "epoch": 2.6861838726245506, + "grad_norm": 2.6780762672424316, + "learning_rate": 6.109630757518082e-07, + "loss": 1.2837, + "step": 2615 + }, + { + "epoch": 2.687211093990755, + "grad_norm": 2.8105902671813965, + "learning_rate": 6.090597639893414e-07, + "loss": 1.3635, + "step": 2616 + }, + { + "epoch": 2.6882383153569593, + "grad_norm": 2.609379291534424, + "learning_rate": 6.071564522268748e-07, + "loss": 1.2684, + "step": 2617 + }, + { + "epoch": 2.689265536723164, + "grad_norm": 2.6364426612854004, + "learning_rate": 6.052531404644082e-07, + "loss": 1.3132, + "step": 2618 + }, + { + "epoch": 2.6902927580893685, + "grad_norm": 2.7625892162323, + "learning_rate": 6.033498287019414e-07, + "loss": 1.3828, + "step": 2619 + }, + { + "epoch": 2.6913199794555727, + "grad_norm": 2.5761687755584717, + "learning_rate": 6.014465169394748e-07, + "loss": 1.2953, + "step": 2620 + }, + { + "epoch": 2.6913199794555727, + "eval_loss": 1.7990498542785645, + "eval_runtime": 37.8153, + "eval_samples_per_second": 26.444, + "eval_steps_per_second": 1.111, + "step": 2620 + }, + { + "epoch": 2.6923472008217773, + "grad_norm": 2.7242608070373535, + "learning_rate": 5.99543205177008e-07, + "loss": 1.2924, + "step": 2621 + }, + { + "epoch": 2.6933744221879814, + "grad_norm": 2.589207410812378, + "learning_rate": 5.976398934145414e-07, + "loss": 1.2918, + "step": 2622 + }, + { + "epoch": 2.694401643554186, + "grad_norm": 2.5671799182891846, + "learning_rate": 5.957365816520747e-07, + "loss": 1.288, + "step": 2623 + }, + { + "epoch": 2.69542886492039, + "grad_norm": 2.6533076763153076, + "learning_rate": 5.938332698896079e-07, + "loss": 1.3131, + "step": 2624 + }, + { + "epoch": 2.6964560862865947, + "grad_norm": 2.60552716255188, + "learning_rate": 5.919299581271413e-07, + "loss": 1.3049, + "step": 2625 + }, + { + "epoch": 2.6974833076527993, + "grad_norm": 2.6435978412628174, + "learning_rate": 5.900266463646745e-07, + "loss": 1.3068, + "step": 2626 + }, + { + "epoch": 2.6985105290190035, + "grad_norm": 2.551628589630127, + "learning_rate": 5.881233346022079e-07, + "loss": 1.2941, + "step": 2627 + }, + { + "epoch": 2.699537750385208, + "grad_norm": 2.829988479614258, + "learning_rate": 5.862200228397411e-07, + "loss": 1.3361, + "step": 2628 + }, + { + "epoch": 2.7005649717514126, + "grad_norm": 2.6916370391845703, + "learning_rate": 5.843167110772745e-07, + "loss": 1.3516, + "step": 2629 + }, + { + "epoch": 2.701592193117617, + "grad_norm": 2.6523451805114746, + "learning_rate": 5.824133993148079e-07, + "loss": 1.3286, + "step": 2630 + }, + { + "epoch": 2.701592193117617, + "eval_loss": 1.7828145027160645, + "eval_runtime": 37.9485, + "eval_samples_per_second": 26.351, + "eval_steps_per_second": 1.107, + "step": 2630 + }, + { + "epoch": 2.7026194144838214, + "grad_norm": 2.6095120906829834, + "learning_rate": 5.805100875523411e-07, + "loss": 1.2778, + "step": 2631 + }, + { + "epoch": 2.7036466358500255, + "grad_norm": 2.630551338195801, + "learning_rate": 5.786067757898745e-07, + "loss": 1.2549, + "step": 2632 + }, + { + "epoch": 2.70467385721623, + "grad_norm": 2.585043430328369, + "learning_rate": 5.767034640274078e-07, + "loss": 1.2639, + "step": 2633 + }, + { + "epoch": 2.7057010785824342, + "grad_norm": 2.8497092723846436, + "learning_rate": 5.74800152264941e-07, + "loss": 1.3636, + "step": 2634 + }, + { + "epoch": 2.706728299948639, + "grad_norm": 2.7136929035186768, + "learning_rate": 5.728968405024744e-07, + "loss": 1.2788, + "step": 2635 + }, + { + "epoch": 2.7077555213148434, + "grad_norm": 2.7176077365875244, + "learning_rate": 5.709935287400076e-07, + "loss": 1.3561, + "step": 2636 + }, + { + "epoch": 2.7087827426810476, + "grad_norm": 2.754312515258789, + "learning_rate": 5.69090216977541e-07, + "loss": 1.3105, + "step": 2637 + }, + { + "epoch": 2.709809964047252, + "grad_norm": 2.6190314292907715, + "learning_rate": 5.671869052150742e-07, + "loss": 1.2946, + "step": 2638 + }, + { + "epoch": 2.7108371854134568, + "grad_norm": 2.5728366374969482, + "learning_rate": 5.652835934526075e-07, + "loss": 1.2989, + "step": 2639 + }, + { + "epoch": 2.711864406779661, + "grad_norm": 2.5806870460510254, + "learning_rate": 5.633802816901409e-07, + "loss": 1.3021, + "step": 2640 + }, + { + "epoch": 2.711864406779661, + "eval_loss": 1.7884520292282104, + "eval_runtime": 38.0067, + "eval_samples_per_second": 26.311, + "eval_steps_per_second": 1.105, + "step": 2640 + }, + { + "epoch": 2.7128916281458655, + "grad_norm": 2.5025806427001953, + "learning_rate": 5.614769699276742e-07, + "loss": 1.2784, + "step": 2641 + }, + { + "epoch": 2.71391884951207, + "grad_norm": 2.7000210285186768, + "learning_rate": 5.595736581652075e-07, + "loss": 1.2955, + "step": 2642 + }, + { + "epoch": 2.714946070878274, + "grad_norm": 2.5144412517547607, + "learning_rate": 5.576703464027408e-07, + "loss": 1.3036, + "step": 2643 + }, + { + "epoch": 2.715973292244479, + "grad_norm": 2.511216878890991, + "learning_rate": 5.557670346402741e-07, + "loss": 1.3266, + "step": 2644 + }, + { + "epoch": 2.717000513610683, + "grad_norm": 2.7001969814300537, + "learning_rate": 5.538637228778075e-07, + "loss": 1.3082, + "step": 2645 + }, + { + "epoch": 2.7180277349768875, + "grad_norm": 2.7682130336761475, + "learning_rate": 5.519604111153407e-07, + "loss": 1.3354, + "step": 2646 + }, + { + "epoch": 2.7190549563430917, + "grad_norm": 2.7113163471221924, + "learning_rate": 5.500570993528741e-07, + "loss": 1.2606, + "step": 2647 + }, + { + "epoch": 2.7200821777092963, + "grad_norm": 2.772627592086792, + "learning_rate": 5.481537875904073e-07, + "loss": 1.3362, + "step": 2648 + }, + { + "epoch": 2.721109399075501, + "grad_norm": 2.7985000610351562, + "learning_rate": 5.462504758279406e-07, + "loss": 1.3503, + "step": 2649 + }, + { + "epoch": 2.722136620441705, + "grad_norm": 2.6800785064697266, + "learning_rate": 5.44347164065474e-07, + "loss": 1.2854, + "step": 2650 + }, + { + "epoch": 2.722136620441705, + "eval_loss": 1.7845864295959473, + "eval_runtime": 38.0453, + "eval_samples_per_second": 26.284, + "eval_steps_per_second": 1.104, + "step": 2650 + }, + { + "epoch": 2.7231638418079096, + "grad_norm": 2.610942840576172, + "learning_rate": 5.424438523030072e-07, + "loss": 1.3161, + "step": 2651 + }, + { + "epoch": 2.724191063174114, + "grad_norm": 2.7399675846099854, + "learning_rate": 5.405405405405406e-07, + "loss": 1.357, + "step": 2652 + }, + { + "epoch": 2.7252182845403183, + "grad_norm": 2.6037936210632324, + "learning_rate": 5.386372287780739e-07, + "loss": 1.3198, + "step": 2653 + }, + { + "epoch": 2.726245505906523, + "grad_norm": 2.5747087001800537, + "learning_rate": 5.367339170156072e-07, + "loss": 1.3382, + "step": 2654 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 2.8241353034973145, + "learning_rate": 5.348306052531406e-07, + "loss": 1.3049, + "step": 2655 + }, + { + "epoch": 2.7282999486389317, + "grad_norm": 2.624354600906372, + "learning_rate": 5.329272934906738e-07, + "loss": 1.3272, + "step": 2656 + }, + { + "epoch": 2.7293271700051362, + "grad_norm": 2.5246741771698, + "learning_rate": 5.310239817282072e-07, + "loss": 1.2704, + "step": 2657 + }, + { + "epoch": 2.7303543913713404, + "grad_norm": 2.817598342895508, + "learning_rate": 5.291206699657404e-07, + "loss": 1.3404, + "step": 2658 + }, + { + "epoch": 2.731381612737545, + "grad_norm": 2.4856812953948975, + "learning_rate": 5.272173582032737e-07, + "loss": 1.2904, + "step": 2659 + }, + { + "epoch": 2.732408834103749, + "grad_norm": 2.7321617603302, + "learning_rate": 5.253140464408071e-07, + "loss": 1.3131, + "step": 2660 + }, + { + "epoch": 2.732408834103749, + "eval_loss": 1.784244418144226, + "eval_runtime": 38.0266, + "eval_samples_per_second": 26.297, + "eval_steps_per_second": 1.104, + "step": 2660 + }, + { + "epoch": 2.7334360554699537, + "grad_norm": 2.5358948707580566, + "learning_rate": 5.234107346783403e-07, + "loss": 1.2882, + "step": 2661 + }, + { + "epoch": 2.7344632768361583, + "grad_norm": 2.5920722484588623, + "learning_rate": 5.215074229158737e-07, + "loss": 1.2577, + "step": 2662 + }, + { + "epoch": 2.7354904982023625, + "grad_norm": 2.589002847671509, + "learning_rate": 5.196041111534069e-07, + "loss": 1.2934, + "step": 2663 + }, + { + "epoch": 2.736517719568567, + "grad_norm": 2.5873234272003174, + "learning_rate": 5.177007993909403e-07, + "loss": 1.3088, + "step": 2664 + }, + { + "epoch": 2.7375449409347716, + "grad_norm": 2.7201805114746094, + "learning_rate": 5.157974876284736e-07, + "loss": 1.3226, + "step": 2665 + }, + { + "epoch": 2.7385721623009758, + "grad_norm": 2.5658745765686035, + "learning_rate": 5.138941758660069e-07, + "loss": 1.3197, + "step": 2666 + }, + { + "epoch": 2.7395993836671804, + "grad_norm": 2.6131479740142822, + "learning_rate": 5.119908641035403e-07, + "loss": 1.3035, + "step": 2667 + }, + { + "epoch": 2.740626605033385, + "grad_norm": 2.5535855293273926, + "learning_rate": 5.100875523410735e-07, + "loss": 1.2709, + "step": 2668 + }, + { + "epoch": 2.741653826399589, + "grad_norm": 2.4551708698272705, + "learning_rate": 5.081842405786068e-07, + "loss": 1.2773, + "step": 2669 + }, + { + "epoch": 2.7426810477657937, + "grad_norm": 2.6793084144592285, + "learning_rate": 5.062809288161402e-07, + "loss": 1.2932, + "step": 2670 + }, + { + "epoch": 2.7426810477657937, + "eval_loss": 1.7706480026245117, + "eval_runtime": 38.033, + "eval_samples_per_second": 26.293, + "eval_steps_per_second": 1.104, + "step": 2670 + }, + { + "epoch": 2.743708269131998, + "grad_norm": 2.6757824420928955, + "learning_rate": 5.043776170536734e-07, + "loss": 1.3053, + "step": 2671 + }, + { + "epoch": 2.7447354904982024, + "grad_norm": 2.7082114219665527, + "learning_rate": 5.024743052912067e-07, + "loss": 1.3196, + "step": 2672 + }, + { + "epoch": 2.7457627118644066, + "grad_norm": 2.6333837509155273, + "learning_rate": 5.0057099352874e-07, + "loss": 1.3112, + "step": 2673 + }, + { + "epoch": 2.746789933230611, + "grad_norm": 2.7922589778900146, + "learning_rate": 4.986676817662733e-07, + "loss": 1.3421, + "step": 2674 + }, + { + "epoch": 2.7478171545968157, + "grad_norm": 2.5689852237701416, + "learning_rate": 4.967643700038067e-07, + "loss": 1.3001, + "step": 2675 + }, + { + "epoch": 2.74884437596302, + "grad_norm": 2.6763274669647217, + "learning_rate": 4.9486105824134e-07, + "loss": 1.3132, + "step": 2676 + }, + { + "epoch": 2.7498715973292245, + "grad_norm": 2.778526544570923, + "learning_rate": 4.929577464788733e-07, + "loss": 1.3548, + "step": 2677 + }, + { + "epoch": 2.750898818695429, + "grad_norm": 2.6535632610321045, + "learning_rate": 4.910544347164066e-07, + "loss": 1.3193, + "step": 2678 + }, + { + "epoch": 2.751926040061633, + "grad_norm": 2.5167694091796875, + "learning_rate": 4.891511229539399e-07, + "loss": 1.2681, + "step": 2679 + }, + { + "epoch": 2.752953261427838, + "grad_norm": 2.6354637145996094, + "learning_rate": 4.872478111914733e-07, + "loss": 1.3046, + "step": 2680 + }, + { + "epoch": 2.752953261427838, + "eval_loss": 1.7752299308776855, + "eval_runtime": 38.0209, + "eval_samples_per_second": 26.301, + "eval_steps_per_second": 1.105, + "step": 2680 + }, + { + "epoch": 2.753980482794042, + "grad_norm": 2.5887973308563232, + "learning_rate": 4.853444994290065e-07, + "loss": 1.2766, + "step": 2681 + }, + { + "epoch": 2.7550077041602465, + "grad_norm": 2.693185567855835, + "learning_rate": 4.834411876665398e-07, + "loss": 1.3078, + "step": 2682 + }, + { + "epoch": 2.7560349255264507, + "grad_norm": 2.7474021911621094, + "learning_rate": 4.815378759040731e-07, + "loss": 1.286, + "step": 2683 + }, + { + "epoch": 2.7570621468926553, + "grad_norm": 2.7882676124572754, + "learning_rate": 4.796345641416064e-07, + "loss": 1.3296, + "step": 2684 + }, + { + "epoch": 2.75808936825886, + "grad_norm": 2.633437156677246, + "learning_rate": 4.777312523791397e-07, + "loss": 1.29, + "step": 2685 + }, + { + "epoch": 2.759116589625064, + "grad_norm": 2.5937581062316895, + "learning_rate": 4.7582794061667306e-07, + "loss": 1.3181, + "step": 2686 + }, + { + "epoch": 2.7601438109912686, + "grad_norm": 2.710019588470459, + "learning_rate": 4.739246288542063e-07, + "loss": 1.3347, + "step": 2687 + }, + { + "epoch": 2.761171032357473, + "grad_norm": 2.728793144226074, + "learning_rate": 4.720213170917397e-07, + "loss": 1.2981, + "step": 2688 + }, + { + "epoch": 2.7621982537236773, + "grad_norm": 2.6515908241271973, + "learning_rate": 4.7011800532927293e-07, + "loss": 1.3758, + "step": 2689 + }, + { + "epoch": 2.763225475089882, + "grad_norm": 2.585145950317383, + "learning_rate": 4.682146935668063e-07, + "loss": 1.2745, + "step": 2690 + }, + { + "epoch": 2.763225475089882, + "eval_loss": 1.7617937326431274, + "eval_runtime": 38.0261, + "eval_samples_per_second": 26.298, + "eval_steps_per_second": 1.105, + "step": 2690 + }, + { + "epoch": 2.7642526964560865, + "grad_norm": 2.5894737243652344, + "learning_rate": 4.663113818043396e-07, + "loss": 1.3174, + "step": 2691 + }, + { + "epoch": 2.7652799178222907, + "grad_norm": 2.718618392944336, + "learning_rate": 4.6440807004187286e-07, + "loss": 1.3024, + "step": 2692 + }, + { + "epoch": 2.7663071391884952, + "grad_norm": 2.6753177642822266, + "learning_rate": 4.625047582794062e-07, + "loss": 1.3204, + "step": 2693 + }, + { + "epoch": 2.7673343605546994, + "grad_norm": 2.5795040130615234, + "learning_rate": 4.606014465169395e-07, + "loss": 1.2935, + "step": 2694 + }, + { + "epoch": 2.768361581920904, + "grad_norm": 2.5900628566741943, + "learning_rate": 4.5869813475447284e-07, + "loss": 1.2988, + "step": 2695 + }, + { + "epoch": 2.769388803287108, + "grad_norm": 2.565168619155884, + "learning_rate": 4.5679482299200615e-07, + "loss": 1.3121, + "step": 2696 + }, + { + "epoch": 2.7704160246533127, + "grad_norm": 2.666557788848877, + "learning_rate": 4.548915112295394e-07, + "loss": 1.3062, + "step": 2697 + }, + { + "epoch": 2.7714432460195173, + "grad_norm": 2.5423285961151123, + "learning_rate": 4.529881994670728e-07, + "loss": 1.2633, + "step": 2698 + }, + { + "epoch": 2.7724704673857214, + "grad_norm": 2.5645787715911865, + "learning_rate": 4.5108488770460603e-07, + "loss": 1.3067, + "step": 2699 + }, + { + "epoch": 2.773497688751926, + "grad_norm": 2.4956552982330322, + "learning_rate": 4.491815759421394e-07, + "loss": 1.3153, + "step": 2700 + }, + { + "epoch": 2.773497688751926, + "eval_loss": 1.7727458477020264, + "eval_runtime": 38.0396, + "eval_samples_per_second": 26.288, + "eval_steps_per_second": 1.104, + "step": 2700 + }, + { + "epoch": 2.7745249101181306, + "grad_norm": 2.3624420166015625, + "learning_rate": 4.4727826417967265e-07, + "loss": 1.206, + "step": 2701 + }, + { + "epoch": 2.7755521314843348, + "grad_norm": 2.462627410888672, + "learning_rate": 4.4537495241720596e-07, + "loss": 1.2452, + "step": 2702 + }, + { + "epoch": 2.7765793528505394, + "grad_norm": 2.6404151916503906, + "learning_rate": 4.434716406547393e-07, + "loss": 1.2931, + "step": 2703 + }, + { + "epoch": 2.777606574216744, + "grad_norm": 2.5767128467559814, + "learning_rate": 4.415683288922726e-07, + "loss": 1.3024, + "step": 2704 + }, + { + "epoch": 2.778633795582948, + "grad_norm": 2.6911098957061768, + "learning_rate": 4.3966501712980594e-07, + "loss": 1.2854, + "step": 2705 + }, + { + "epoch": 2.7796610169491527, + "grad_norm": 2.5582408905029297, + "learning_rate": 4.377617053673392e-07, + "loss": 1.2672, + "step": 2706 + }, + { + "epoch": 2.780688238315357, + "grad_norm": 2.7786879539489746, + "learning_rate": 4.358583936048725e-07, + "loss": 1.3574, + "step": 2707 + }, + { + "epoch": 2.7817154596815614, + "grad_norm": 2.8259687423706055, + "learning_rate": 4.339550818424058e-07, + "loss": 1.3238, + "step": 2708 + }, + { + "epoch": 2.7827426810477656, + "grad_norm": 2.754877805709839, + "learning_rate": 4.3205177007993913e-07, + "loss": 1.3005, + "step": 2709 + }, + { + "epoch": 2.78376990241397, + "grad_norm": 2.618274688720703, + "learning_rate": 4.301484583174725e-07, + "loss": 1.3251, + "step": 2710 + }, + { + "epoch": 2.78376990241397, + "eval_loss": 1.7750599384307861, + "eval_runtime": 37.6386, + "eval_samples_per_second": 26.568, + "eval_steps_per_second": 1.116, + "step": 2710 + }, + { + "epoch": 2.7847971237801747, + "grad_norm": 2.4981017112731934, + "learning_rate": 4.2824514655500575e-07, + "loss": 1.2748, + "step": 2711 + }, + { + "epoch": 2.785824345146379, + "grad_norm": 2.6884243488311768, + "learning_rate": 4.2634183479253906e-07, + "loss": 1.3121, + "step": 2712 + }, + { + "epoch": 2.7868515665125835, + "grad_norm": 2.609513759613037, + "learning_rate": 4.2443852303007237e-07, + "loss": 1.2651, + "step": 2713 + }, + { + "epoch": 2.787878787878788, + "grad_norm": 2.650463581085205, + "learning_rate": 4.225352112676057e-07, + "loss": 1.3599, + "step": 2714 + }, + { + "epoch": 2.788906009244992, + "grad_norm": 2.6558997631073, + "learning_rate": 4.20631899505139e-07, + "loss": 1.3087, + "step": 2715 + }, + { + "epoch": 2.789933230611197, + "grad_norm": 2.54502010345459, + "learning_rate": 4.187285877426723e-07, + "loss": 1.2407, + "step": 2716 + }, + { + "epoch": 2.7909604519774014, + "grad_norm": 2.538921356201172, + "learning_rate": 4.1682527598020555e-07, + "loss": 1.2722, + "step": 2717 + }, + { + "epoch": 2.7919876733436055, + "grad_norm": 2.765803813934326, + "learning_rate": 4.149219642177389e-07, + "loss": 1.3689, + "step": 2718 + }, + { + "epoch": 2.79301489470981, + "grad_norm": 2.7181074619293213, + "learning_rate": 4.130186524552722e-07, + "loss": 1.2992, + "step": 2719 + }, + { + "epoch": 2.7940421160760143, + "grad_norm": 2.636695384979248, + "learning_rate": 4.111153406928055e-07, + "loss": 1.3008, + "step": 2720 + }, + { + "epoch": 2.7940421160760143, + "eval_loss": 1.7647284269332886, + "eval_runtime": 37.8262, + "eval_samples_per_second": 26.437, + "eval_steps_per_second": 1.11, + "step": 2720 + }, + { + "epoch": 2.795069337442219, + "grad_norm": 2.471787214279175, + "learning_rate": 4.0921202893033884e-07, + "loss": 1.2691, + "step": 2721 + }, + { + "epoch": 2.796096558808423, + "grad_norm": 2.7055912017822266, + "learning_rate": 4.073087171678721e-07, + "loss": 1.3343, + "step": 2722 + }, + { + "epoch": 2.7971237801746276, + "grad_norm": 2.736154556274414, + "learning_rate": 4.0540540540540546e-07, + "loss": 1.3289, + "step": 2723 + }, + { + "epoch": 2.798151001540832, + "grad_norm": 2.6181352138519287, + "learning_rate": 4.035020936429387e-07, + "loss": 1.2881, + "step": 2724 + }, + { + "epoch": 2.7991782229070363, + "grad_norm": 2.5176339149475098, + "learning_rate": 4.0159878188047203e-07, + "loss": 1.2523, + "step": 2725 + }, + { + "epoch": 2.800205444273241, + "grad_norm": 2.5560781955718994, + "learning_rate": 3.996954701180054e-07, + "loss": 1.2911, + "step": 2726 + }, + { + "epoch": 2.8012326656394455, + "grad_norm": 2.636842727661133, + "learning_rate": 3.9779215835553865e-07, + "loss": 1.2895, + "step": 2727 + }, + { + "epoch": 2.8022598870056497, + "grad_norm": 2.6112375259399414, + "learning_rate": 3.95888846593072e-07, + "loss": 1.3084, + "step": 2728 + }, + { + "epoch": 2.8032871083718542, + "grad_norm": 2.6065049171447754, + "learning_rate": 3.9398553483060527e-07, + "loss": 1.2969, + "step": 2729 + }, + { + "epoch": 2.8043143297380584, + "grad_norm": 2.492626905441284, + "learning_rate": 3.920822230681386e-07, + "loss": 1.3095, + "step": 2730 + }, + { + "epoch": 2.8043143297380584, + "eval_loss": 1.7692224979400635, + "eval_runtime": 37.8818, + "eval_samples_per_second": 26.398, + "eval_steps_per_second": 1.109, + "step": 2730 + }, + { + "epoch": 2.805341551104263, + "grad_norm": 2.727534055709839, + "learning_rate": 3.901789113056719e-07, + "loss": 1.3137, + "step": 2731 + }, + { + "epoch": 2.806368772470467, + "grad_norm": 2.5522494316101074, + "learning_rate": 3.882755995432052e-07, + "loss": 1.2365, + "step": 2732 + }, + { + "epoch": 2.8073959938366717, + "grad_norm": 2.674705982208252, + "learning_rate": 3.8637228778073856e-07, + "loss": 1.367, + "step": 2733 + }, + { + "epoch": 2.8084232152028763, + "grad_norm": 2.577897071838379, + "learning_rate": 3.844689760182718e-07, + "loss": 1.2882, + "step": 2734 + }, + { + "epoch": 2.8094504365690804, + "grad_norm": 2.490436553955078, + "learning_rate": 3.825656642558051e-07, + "loss": 1.2781, + "step": 2735 + }, + { + "epoch": 2.810477657935285, + "grad_norm": 2.531339406967163, + "learning_rate": 3.8066235249333844e-07, + "loss": 1.3248, + "step": 2736 + }, + { + "epoch": 2.8115048793014896, + "grad_norm": 2.6713733673095703, + "learning_rate": 3.7875904073087175e-07, + "loss": 1.3249, + "step": 2737 + }, + { + "epoch": 2.8125321006676938, + "grad_norm": 2.6347765922546387, + "learning_rate": 3.768557289684051e-07, + "loss": 1.3405, + "step": 2738 + }, + { + "epoch": 2.8135593220338984, + "grad_norm": 2.6573073863983154, + "learning_rate": 3.7495241720593836e-07, + "loss": 1.315, + "step": 2739 + }, + { + "epoch": 2.814586543400103, + "grad_norm": 2.5330655574798584, + "learning_rate": 3.730491054434716e-07, + "loss": 1.2925, + "step": 2740 + }, + { + "epoch": 2.814586543400103, + "eval_loss": 1.762029767036438, + "eval_runtime": 37.9299, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 2740 + }, + { + "epoch": 2.815613764766307, + "grad_norm": 2.5661964416503906, + "learning_rate": 3.71145793681005e-07, + "loss": 1.2857, + "step": 2741 + }, + { + "epoch": 2.8166409861325117, + "grad_norm": 2.6637752056121826, + "learning_rate": 3.692424819185383e-07, + "loss": 1.3445, + "step": 2742 + }, + { + "epoch": 2.817668207498716, + "grad_norm": 2.7280433177948, + "learning_rate": 3.673391701560716e-07, + "loss": 1.3293, + "step": 2743 + }, + { + "epoch": 2.8186954288649204, + "grad_norm": 2.5757462978363037, + "learning_rate": 3.654358583936049e-07, + "loss": 1.2899, + "step": 2744 + }, + { + "epoch": 2.8197226502311246, + "grad_norm": 2.515266180038452, + "learning_rate": 3.6353254663113817e-07, + "loss": 1.29, + "step": 2745 + }, + { + "epoch": 2.820749871597329, + "grad_norm": 2.607907295227051, + "learning_rate": 3.6162923486867153e-07, + "loss": 1.2959, + "step": 2746 + }, + { + "epoch": 2.8217770929635337, + "grad_norm": 2.6733455657958984, + "learning_rate": 3.597259231062048e-07, + "loss": 1.3026, + "step": 2747 + }, + { + "epoch": 2.822804314329738, + "grad_norm": 2.608802318572998, + "learning_rate": 3.5782261134373815e-07, + "loss": 1.2633, + "step": 2748 + }, + { + "epoch": 2.8238315356959425, + "grad_norm": 2.553936004638672, + "learning_rate": 3.5591929958127146e-07, + "loss": 1.2745, + "step": 2749 + }, + { + "epoch": 2.824858757062147, + "grad_norm": 2.7180869579315186, + "learning_rate": 3.540159878188047e-07, + "loss": 1.3157, + "step": 2750 + }, + { + "epoch": 2.824858757062147, + "eval_loss": 1.7729238271713257, + "eval_runtime": 37.9877, + "eval_samples_per_second": 26.324, + "eval_steps_per_second": 1.106, + "step": 2750 + }, + { + "epoch": 2.825885978428351, + "grad_norm": 2.7445926666259766, + "learning_rate": 3.521126760563381e-07, + "loss": 1.3174, + "step": 2751 + }, + { + "epoch": 2.826913199794556, + "grad_norm": 2.578984022140503, + "learning_rate": 3.5020936429387134e-07, + "loss": 1.3095, + "step": 2752 + }, + { + "epoch": 2.8279404211607604, + "grad_norm": 2.6088266372680664, + "learning_rate": 3.483060525314047e-07, + "loss": 1.2648, + "step": 2753 + }, + { + "epoch": 2.8289676425269645, + "grad_norm": 2.6487412452697754, + "learning_rate": 3.46402740768938e-07, + "loss": 1.3675, + "step": 2754 + }, + { + "epoch": 2.829994863893169, + "grad_norm": 2.604130506515503, + "learning_rate": 3.4449942900647127e-07, + "loss": 1.2722, + "step": 2755 + }, + { + "epoch": 2.8310220852593733, + "grad_norm": 2.6885905265808105, + "learning_rate": 3.4259611724400463e-07, + "loss": 1.3354, + "step": 2756 + }, + { + "epoch": 2.832049306625578, + "grad_norm": 2.709333896636963, + "learning_rate": 3.406928054815379e-07, + "loss": 1.403, + "step": 2757 + }, + { + "epoch": 2.833076527991782, + "grad_norm": 2.6402320861816406, + "learning_rate": 3.3878949371907125e-07, + "loss": 1.2836, + "step": 2758 + }, + { + "epoch": 2.8341037493579866, + "grad_norm": 2.6921451091766357, + "learning_rate": 3.368861819566045e-07, + "loss": 1.2743, + "step": 2759 + }, + { + "epoch": 2.835130970724191, + "grad_norm": 2.5713753700256348, + "learning_rate": 3.349828701941378e-07, + "loss": 1.2691, + "step": 2760 + }, + { + "epoch": 2.835130970724191, + "eval_loss": 1.7677931785583496, + "eval_runtime": 37.9668, + "eval_samples_per_second": 26.339, + "eval_steps_per_second": 1.106, + "step": 2760 + }, + { + "epoch": 2.8361581920903953, + "grad_norm": 2.3884944915771484, + "learning_rate": 3.330795584316712e-07, + "loss": 1.2537, + "step": 2761 + }, + { + "epoch": 2.8371854134566, + "grad_norm": 2.579089641571045, + "learning_rate": 3.3117624666920443e-07, + "loss": 1.2779, + "step": 2762 + }, + { + "epoch": 2.8382126348228045, + "grad_norm": 2.796998977661133, + "learning_rate": 3.292729349067378e-07, + "loss": 1.3303, + "step": 2763 + }, + { + "epoch": 2.8392398561890086, + "grad_norm": 2.4984240531921387, + "learning_rate": 3.2736962314427105e-07, + "loss": 1.3029, + "step": 2764 + }, + { + "epoch": 2.8402670775552132, + "grad_norm": 2.6721031665802, + "learning_rate": 3.2546631138180436e-07, + "loss": 1.352, + "step": 2765 + }, + { + "epoch": 2.841294298921418, + "grad_norm": 2.7240445613861084, + "learning_rate": 3.2356299961933767e-07, + "loss": 1.2783, + "step": 2766 + }, + { + "epoch": 2.842321520287622, + "grad_norm": 2.72751784324646, + "learning_rate": 3.21659687856871e-07, + "loss": 1.3461, + "step": 2767 + }, + { + "epoch": 2.8433487416538266, + "grad_norm": 2.6866025924682617, + "learning_rate": 3.1975637609440424e-07, + "loss": 1.2654, + "step": 2768 + }, + { + "epoch": 2.8443759630200307, + "grad_norm": 2.7173006534576416, + "learning_rate": 3.178530643319376e-07, + "loss": 1.3543, + "step": 2769 + }, + { + "epoch": 2.8454031843862353, + "grad_norm": 2.65193247795105, + "learning_rate": 3.159497525694709e-07, + "loss": 1.3036, + "step": 2770 + }, + { + "epoch": 2.8454031843862353, + "eval_loss": 1.7600867748260498, + "eval_runtime": 38.0302, + "eval_samples_per_second": 26.295, + "eval_steps_per_second": 1.104, + "step": 2770 + }, + { + "epoch": 2.8464304057524394, + "grad_norm": 2.647165298461914, + "learning_rate": 3.140464408070042e-07, + "loss": 1.3017, + "step": 2771 + }, + { + "epoch": 2.847457627118644, + "grad_norm": 2.7047841548919678, + "learning_rate": 3.1214312904453753e-07, + "loss": 1.3689, + "step": 2772 + }, + { + "epoch": 2.8484848484848486, + "grad_norm": 2.618946075439453, + "learning_rate": 3.1023981728207084e-07, + "loss": 1.3024, + "step": 2773 + }, + { + "epoch": 2.8495120698510528, + "grad_norm": 2.6727678775787354, + "learning_rate": 3.0833650551960415e-07, + "loss": 1.3304, + "step": 2774 + }, + { + "epoch": 2.8505392912172574, + "grad_norm": 2.559290647506714, + "learning_rate": 3.0643319375713746e-07, + "loss": 1.303, + "step": 2775 + }, + { + "epoch": 2.851566512583462, + "grad_norm": 2.5195682048797607, + "learning_rate": 3.045298819946707e-07, + "loss": 1.3081, + "step": 2776 + }, + { + "epoch": 2.852593733949666, + "grad_norm": 2.586885452270508, + "learning_rate": 3.026265702322041e-07, + "loss": 1.2692, + "step": 2777 + }, + { + "epoch": 2.8536209553158707, + "grad_norm": 2.5503077507019043, + "learning_rate": 3.007232584697374e-07, + "loss": 1.2742, + "step": 2778 + }, + { + "epoch": 2.854648176682075, + "grad_norm": 2.515547275543213, + "learning_rate": 2.988199467072707e-07, + "loss": 1.2725, + "step": 2779 + }, + { + "epoch": 2.8556753980482794, + "grad_norm": 2.402459144592285, + "learning_rate": 2.9691663494480396e-07, + "loss": 1.2256, + "step": 2780 + }, + { + "epoch": 2.8556753980482794, + "eval_loss": 1.76780104637146, + "eval_runtime": 38.0687, + "eval_samples_per_second": 26.268, + "eval_steps_per_second": 1.103, + "step": 2780 + }, + { + "epoch": 2.8567026194144836, + "grad_norm": 2.5463836193084717, + "learning_rate": 2.9501332318233727e-07, + "loss": 1.3251, + "step": 2781 + }, + { + "epoch": 2.857729840780688, + "grad_norm": 2.532472610473633, + "learning_rate": 2.931100114198706e-07, + "loss": 1.275, + "step": 2782 + }, + { + "epoch": 2.8587570621468927, + "grad_norm": 2.6995961666107178, + "learning_rate": 2.9120669965740394e-07, + "loss": 1.3001, + "step": 2783 + }, + { + "epoch": 2.859784283513097, + "grad_norm": 2.5292301177978516, + "learning_rate": 2.8930338789493725e-07, + "loss": 1.2636, + "step": 2784 + }, + { + "epoch": 2.8608115048793015, + "grad_norm": 2.7945351600646973, + "learning_rate": 2.874000761324705e-07, + "loss": 1.2888, + "step": 2785 + }, + { + "epoch": 2.861838726245506, + "grad_norm": 2.607816457748413, + "learning_rate": 2.854967643700038e-07, + "loss": 1.3016, + "step": 2786 + }, + { + "epoch": 2.86286594761171, + "grad_norm": 2.7001326084136963, + "learning_rate": 2.835934526075371e-07, + "loss": 1.3108, + "step": 2787 + }, + { + "epoch": 2.863893168977915, + "grad_norm": 2.8196349143981934, + "learning_rate": 2.8169014084507043e-07, + "loss": 1.3215, + "step": 2788 + }, + { + "epoch": 2.8649203903441194, + "grad_norm": 2.7477948665618896, + "learning_rate": 2.7978682908260374e-07, + "loss": 1.3218, + "step": 2789 + }, + { + "epoch": 2.8659476117103235, + "grad_norm": 2.605776309967041, + "learning_rate": 2.7788351732013705e-07, + "loss": 1.301, + "step": 2790 + }, + { + "epoch": 2.8659476117103235, + "eval_loss": 1.7605133056640625, + "eval_runtime": 38.0458, + "eval_samples_per_second": 26.284, + "eval_steps_per_second": 1.104, + "step": 2790 + }, + { + "epoch": 2.866974833076528, + "grad_norm": 2.417668104171753, + "learning_rate": 2.7598020555767036e-07, + "loss": 1.2655, + "step": 2791 + }, + { + "epoch": 2.8680020544427323, + "grad_norm": 2.551400899887085, + "learning_rate": 2.7407689379520367e-07, + "loss": 1.2828, + "step": 2792 + }, + { + "epoch": 2.869029275808937, + "grad_norm": 2.5882434844970703, + "learning_rate": 2.72173582032737e-07, + "loss": 1.3466, + "step": 2793 + }, + { + "epoch": 2.870056497175141, + "grad_norm": 2.5469772815704346, + "learning_rate": 2.702702702702703e-07, + "loss": 1.3224, + "step": 2794 + }, + { + "epoch": 2.8710837185413456, + "grad_norm": 2.5152812004089355, + "learning_rate": 2.683669585078036e-07, + "loss": 1.2682, + "step": 2795 + }, + { + "epoch": 2.87211093990755, + "grad_norm": 2.587076425552368, + "learning_rate": 2.664636467453369e-07, + "loss": 1.3311, + "step": 2796 + }, + { + "epoch": 2.8731381612737543, + "grad_norm": 2.647310256958008, + "learning_rate": 2.645603349828702e-07, + "loss": 1.2956, + "step": 2797 + }, + { + "epoch": 2.874165382639959, + "grad_norm": 2.5272367000579834, + "learning_rate": 2.6265702322040353e-07, + "loss": 1.2552, + "step": 2798 + }, + { + "epoch": 2.8751926040061635, + "grad_norm": 2.5864880084991455, + "learning_rate": 2.6075371145793684e-07, + "loss": 1.2629, + "step": 2799 + }, + { + "epoch": 2.8762198253723676, + "grad_norm": 2.6133804321289062, + "learning_rate": 2.5885039969547015e-07, + "loss": 1.2948, + "step": 2800 + }, + { + "epoch": 2.8762198253723676, + "eval_loss": 1.755886435508728, + "eval_runtime": 38.0646, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 2800 + }, + { + "epoch": 2.8772470467385722, + "grad_norm": 2.581003189086914, + "learning_rate": 2.5694708793300346e-07, + "loss": 1.251, + "step": 2801 + }, + { + "epoch": 2.878274268104777, + "grad_norm": 2.617664337158203, + "learning_rate": 2.5504377617053677e-07, + "loss": 1.2834, + "step": 2802 + }, + { + "epoch": 2.879301489470981, + "grad_norm": 2.5373620986938477, + "learning_rate": 2.531404644080701e-07, + "loss": 1.2943, + "step": 2803 + }, + { + "epoch": 2.8803287108371856, + "grad_norm": 2.5740702152252197, + "learning_rate": 2.5123715264560333e-07, + "loss": 1.3091, + "step": 2804 + }, + { + "epoch": 2.8813559322033897, + "grad_norm": 2.6244497299194336, + "learning_rate": 2.4933384088313664e-07, + "loss": 1.3096, + "step": 2805 + }, + { + "epoch": 2.8823831535695943, + "grad_norm": 2.568964719772339, + "learning_rate": 2.4743052912067e-07, + "loss": 1.32, + "step": 2806 + }, + { + "epoch": 2.8834103749357984, + "grad_norm": 2.4921417236328125, + "learning_rate": 2.455272173582033e-07, + "loss": 1.2663, + "step": 2807 + }, + { + "epoch": 2.884437596302003, + "grad_norm": 2.457496404647827, + "learning_rate": 2.436239055957366e-07, + "loss": 1.3325, + "step": 2808 + }, + { + "epoch": 2.8854648176682076, + "grad_norm": 2.797403573989868, + "learning_rate": 2.417205938332699e-07, + "loss": 1.3544, + "step": 2809 + }, + { + "epoch": 2.8864920390344118, + "grad_norm": 2.635730266571045, + "learning_rate": 2.398172820708032e-07, + "loss": 1.3008, + "step": 2810 + }, + { + "epoch": 2.8864920390344118, + "eval_loss": 1.7717928886413574, + "eval_runtime": 37.6949, + "eval_samples_per_second": 26.529, + "eval_steps_per_second": 1.114, + "step": 2810 + }, + { + "epoch": 2.8875192604006163, + "grad_norm": 2.687199592590332, + "learning_rate": 2.3791397030833653e-07, + "loss": 1.3438, + "step": 2811 + }, + { + "epoch": 2.888546481766821, + "grad_norm": 2.510181188583374, + "learning_rate": 2.3601065854586984e-07, + "loss": 1.2972, + "step": 2812 + }, + { + "epoch": 2.889573703133025, + "grad_norm": 2.5739052295684814, + "learning_rate": 2.3410734678340315e-07, + "loss": 1.3009, + "step": 2813 + }, + { + "epoch": 2.8906009244992297, + "grad_norm": 2.590186834335327, + "learning_rate": 2.3220403502093643e-07, + "loss": 1.2711, + "step": 2814 + }, + { + "epoch": 2.8916281458654343, + "grad_norm": 2.64300799369812, + "learning_rate": 2.3030072325846974e-07, + "loss": 1.2883, + "step": 2815 + }, + { + "epoch": 2.8926553672316384, + "grad_norm": 2.673630714416504, + "learning_rate": 2.2839741149600308e-07, + "loss": 1.2493, + "step": 2816 + }, + { + "epoch": 2.893682588597843, + "grad_norm": 2.8108747005462646, + "learning_rate": 2.264940997335364e-07, + "loss": 1.3046, + "step": 2817 + }, + { + "epoch": 2.894709809964047, + "grad_norm": 2.6422648429870605, + "learning_rate": 2.245907879710697e-07, + "loss": 1.3026, + "step": 2818 + }, + { + "epoch": 2.8957370313302517, + "grad_norm": 2.637664556503296, + "learning_rate": 2.2268747620860298e-07, + "loss": 1.2762, + "step": 2819 + }, + { + "epoch": 2.896764252696456, + "grad_norm": 2.700580596923828, + "learning_rate": 2.207841644461363e-07, + "loss": 1.3211, + "step": 2820 + }, + { + "epoch": 2.896764252696456, + "eval_loss": 1.7740252017974854, + "eval_runtime": 37.8573, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 2820 + }, + { + "epoch": 2.8977914740626605, + "grad_norm": 2.5184342861175537, + "learning_rate": 2.188808526836696e-07, + "loss": 1.2616, + "step": 2821 + }, + { + "epoch": 2.898818695428865, + "grad_norm": 2.716503381729126, + "learning_rate": 2.169775409212029e-07, + "loss": 1.39, + "step": 2822 + }, + { + "epoch": 2.899845916795069, + "grad_norm": 2.5148370265960693, + "learning_rate": 2.1507422915873625e-07, + "loss": 1.2786, + "step": 2823 + }, + { + "epoch": 2.900873138161274, + "grad_norm": 2.6032865047454834, + "learning_rate": 2.1317091739626953e-07, + "loss": 1.3297, + "step": 2824 + }, + { + "epoch": 2.9019003595274784, + "grad_norm": 2.642293930053711, + "learning_rate": 2.1126760563380284e-07, + "loss": 1.3277, + "step": 2825 + }, + { + "epoch": 2.9029275808936825, + "grad_norm": 2.4964346885681152, + "learning_rate": 2.0936429387133615e-07, + "loss": 1.273, + "step": 2826 + }, + { + "epoch": 2.903954802259887, + "grad_norm": 2.6109700202941895, + "learning_rate": 2.0746098210886946e-07, + "loss": 1.2879, + "step": 2827 + }, + { + "epoch": 2.9049820236260913, + "grad_norm": 2.631652593612671, + "learning_rate": 2.0555767034640274e-07, + "loss": 1.2992, + "step": 2828 + }, + { + "epoch": 2.906009244992296, + "grad_norm": 2.7252166271209717, + "learning_rate": 2.0365435858393605e-07, + "loss": 1.3625, + "step": 2829 + }, + { + "epoch": 2.9070364663585, + "grad_norm": 2.556710720062256, + "learning_rate": 2.0175104682146936e-07, + "loss": 1.2974, + "step": 2830 + }, + { + "epoch": 2.9070364663585, + "eval_loss": 1.7611210346221924, + "eval_runtime": 37.9618, + "eval_samples_per_second": 26.342, + "eval_steps_per_second": 1.106, + "step": 2830 + }, + { + "epoch": 2.9080636877247046, + "grad_norm": 2.6725974082946777, + "learning_rate": 1.998477350590027e-07, + "loss": 1.3144, + "step": 2831 + }, + { + "epoch": 2.909090909090909, + "grad_norm": 2.5859169960021973, + "learning_rate": 1.97944423296536e-07, + "loss": 1.2903, + "step": 2832 + }, + { + "epoch": 2.9101181304571133, + "grad_norm": 2.5611412525177, + "learning_rate": 1.960411115340693e-07, + "loss": 1.3295, + "step": 2833 + }, + { + "epoch": 2.911145351823318, + "grad_norm": 2.5633907318115234, + "learning_rate": 1.941377997716026e-07, + "loss": 1.3239, + "step": 2834 + }, + { + "epoch": 2.9121725731895225, + "grad_norm": 2.5077147483825684, + "learning_rate": 1.922344880091359e-07, + "loss": 1.2854, + "step": 2835 + }, + { + "epoch": 2.9131997945557266, + "grad_norm": 2.582174062728882, + "learning_rate": 1.9033117624666922e-07, + "loss": 1.2955, + "step": 2836 + }, + { + "epoch": 2.9142270159219312, + "grad_norm": 2.447506904602051, + "learning_rate": 1.8842786448420255e-07, + "loss": 1.2537, + "step": 2837 + }, + { + "epoch": 2.915254237288136, + "grad_norm": 2.6061720848083496, + "learning_rate": 1.865245527217358e-07, + "loss": 1.3014, + "step": 2838 + }, + { + "epoch": 2.91628145865434, + "grad_norm": 2.691216468811035, + "learning_rate": 1.8462124095926915e-07, + "loss": 1.3382, + "step": 2839 + }, + { + "epoch": 2.9173086800205446, + "grad_norm": 2.541931390762329, + "learning_rate": 1.8271792919680246e-07, + "loss": 1.2599, + "step": 2840 + }, + { + "epoch": 2.9173086800205446, + "eval_loss": 1.7685185670852661, + "eval_runtime": 37.9961, + "eval_samples_per_second": 26.318, + "eval_steps_per_second": 1.105, + "step": 2840 + }, + { + "epoch": 2.9183359013867487, + "grad_norm": 2.6509108543395996, + "learning_rate": 1.8081461743433577e-07, + "loss": 1.3164, + "step": 2841 + }, + { + "epoch": 2.9193631227529533, + "grad_norm": 2.5785300731658936, + "learning_rate": 1.7891130567186908e-07, + "loss": 1.2777, + "step": 2842 + }, + { + "epoch": 2.9203903441191574, + "grad_norm": 2.621399402618408, + "learning_rate": 1.7700799390940236e-07, + "loss": 1.3187, + "step": 2843 + }, + { + "epoch": 2.921417565485362, + "grad_norm": 2.5683820247650146, + "learning_rate": 1.7510468214693567e-07, + "loss": 1.2718, + "step": 2844 + }, + { + "epoch": 2.9224447868515666, + "grad_norm": 2.581298589706421, + "learning_rate": 1.73201370384469e-07, + "loss": 1.3185, + "step": 2845 + }, + { + "epoch": 2.9234720082177708, + "grad_norm": 2.4775500297546387, + "learning_rate": 1.7129805862200231e-07, + "loss": 1.3099, + "step": 2846 + }, + { + "epoch": 2.9244992295839753, + "grad_norm": 2.5610580444335938, + "learning_rate": 1.6939474685953562e-07, + "loss": 1.284, + "step": 2847 + }, + { + "epoch": 2.92552645095018, + "grad_norm": 2.5270745754241943, + "learning_rate": 1.674914350970689e-07, + "loss": 1.3279, + "step": 2848 + }, + { + "epoch": 2.926553672316384, + "grad_norm": 2.5251827239990234, + "learning_rate": 1.6558812333460222e-07, + "loss": 1.3127, + "step": 2849 + }, + { + "epoch": 2.9275808936825887, + "grad_norm": 2.5263609886169434, + "learning_rate": 1.6368481157213553e-07, + "loss": 1.3068, + "step": 2850 + }, + { + "epoch": 2.9275808936825887, + "eval_loss": 1.7712348699569702, + "eval_runtime": 38.0713, + "eval_samples_per_second": 26.267, + "eval_steps_per_second": 1.103, + "step": 2850 + }, + { + "epoch": 2.9286081150487933, + "grad_norm": 2.5240910053253174, + "learning_rate": 1.6178149980966884e-07, + "loss": 1.2879, + "step": 2851 + }, + { + "epoch": 2.9296353364149974, + "grad_norm": 2.5428097248077393, + "learning_rate": 1.5987818804720212e-07, + "loss": 1.3068, + "step": 2852 + }, + { + "epoch": 2.930662557781202, + "grad_norm": 2.562764883041382, + "learning_rate": 1.5797487628473546e-07, + "loss": 1.2917, + "step": 2853 + }, + { + "epoch": 2.931689779147406, + "grad_norm": 2.68951153755188, + "learning_rate": 1.5607156452226877e-07, + "loss": 1.3124, + "step": 2854 + }, + { + "epoch": 2.9327170005136107, + "grad_norm": 2.521895170211792, + "learning_rate": 1.5416825275980208e-07, + "loss": 1.2865, + "step": 2855 + }, + { + "epoch": 2.933744221879815, + "grad_norm": 2.6974024772644043, + "learning_rate": 1.5226494099733536e-07, + "loss": 1.3136, + "step": 2856 + }, + { + "epoch": 2.9347714432460195, + "grad_norm": 2.673924207687378, + "learning_rate": 1.503616292348687e-07, + "loss": 1.3265, + "step": 2857 + }, + { + "epoch": 2.935798664612224, + "grad_norm": 2.605788469314575, + "learning_rate": 1.4845831747240198e-07, + "loss": 1.2894, + "step": 2858 + }, + { + "epoch": 2.936825885978428, + "grad_norm": 2.7430074214935303, + "learning_rate": 1.465550057099353e-07, + "loss": 1.3085, + "step": 2859 + }, + { + "epoch": 2.937853107344633, + "grad_norm": 2.6493849754333496, + "learning_rate": 1.4465169394746862e-07, + "loss": 1.3228, + "step": 2860 + }, + { + "epoch": 2.937853107344633, + "eval_loss": 1.778856635093689, + "eval_runtime": 38.0447, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 2860 + }, + { + "epoch": 2.9388803287108374, + "grad_norm": 2.7383079528808594, + "learning_rate": 1.427483821850019e-07, + "loss": 1.296, + "step": 2861 + }, + { + "epoch": 2.9399075500770415, + "grad_norm": 2.5898325443267822, + "learning_rate": 1.4084507042253522e-07, + "loss": 1.2738, + "step": 2862 + }, + { + "epoch": 2.940934771443246, + "grad_norm": 2.617520809173584, + "learning_rate": 1.3894175866006853e-07, + "loss": 1.283, + "step": 2863 + }, + { + "epoch": 2.9419619928094507, + "grad_norm": 2.5761914253234863, + "learning_rate": 1.3703844689760184e-07, + "loss": 1.2752, + "step": 2864 + }, + { + "epoch": 2.942989214175655, + "grad_norm": 2.5721006393432617, + "learning_rate": 1.3513513513513515e-07, + "loss": 1.3019, + "step": 2865 + }, + { + "epoch": 2.9440164355418594, + "grad_norm": 2.5242764949798584, + "learning_rate": 1.3323182337266846e-07, + "loss": 1.2709, + "step": 2866 + }, + { + "epoch": 2.9450436569080636, + "grad_norm": 2.5600898265838623, + "learning_rate": 1.3132851161020176e-07, + "loss": 1.3178, + "step": 2867 + }, + { + "epoch": 2.946070878274268, + "grad_norm": 2.4490609169006348, + "learning_rate": 1.2942519984773507e-07, + "loss": 1.27, + "step": 2868 + }, + { + "epoch": 2.9470980996404723, + "grad_norm": 2.6730990409851074, + "learning_rate": 1.2752188808526838e-07, + "loss": 1.285, + "step": 2869 + }, + { + "epoch": 2.948125321006677, + "grad_norm": 2.5714759826660156, + "learning_rate": 1.2561857632280167e-07, + "loss": 1.3251, + "step": 2870 + }, + { + "epoch": 2.948125321006677, + "eval_loss": 1.7652958631515503, + "eval_runtime": 38.0497, + "eval_samples_per_second": 26.281, + "eval_steps_per_second": 1.104, + "step": 2870 + }, + { + "epoch": 2.9491525423728815, + "grad_norm": 2.668329954147339, + "learning_rate": 1.23715264560335e-07, + "loss": 1.349, + "step": 2871 + }, + { + "epoch": 2.9501797637390856, + "grad_norm": 2.5298960208892822, + "learning_rate": 1.218119527978683e-07, + "loss": 1.3064, + "step": 2872 + }, + { + "epoch": 2.9512069851052902, + "grad_norm": 2.5941309928894043, + "learning_rate": 1.199086410354016e-07, + "loss": 1.3315, + "step": 2873 + }, + { + "epoch": 2.952234206471495, + "grad_norm": 2.622615098953247, + "learning_rate": 1.1800532927293492e-07, + "loss": 1.2641, + "step": 2874 + }, + { + "epoch": 2.953261427837699, + "grad_norm": 2.590561866760254, + "learning_rate": 1.1610201751046822e-07, + "loss": 1.2865, + "step": 2875 + }, + { + "epoch": 2.9542886492039035, + "grad_norm": 2.612455129623413, + "learning_rate": 1.1419870574800154e-07, + "loss": 1.297, + "step": 2876 + }, + { + "epoch": 2.9553158705701077, + "grad_norm": 2.672116279602051, + "learning_rate": 1.1229539398553485e-07, + "loss": 1.3467, + "step": 2877 + }, + { + "epoch": 2.9563430919363123, + "grad_norm": 2.6142027378082275, + "learning_rate": 1.1039208222306814e-07, + "loss": 1.282, + "step": 2878 + }, + { + "epoch": 2.9573703133025164, + "grad_norm": 2.7042629718780518, + "learning_rate": 1.0848877046060145e-07, + "loss": 1.3018, + "step": 2879 + }, + { + "epoch": 2.958397534668721, + "grad_norm": 2.5736732482910156, + "learning_rate": 1.0658545869813476e-07, + "loss": 1.3141, + "step": 2880 + }, + { + "epoch": 2.958397534668721, + "eval_loss": 1.7524182796478271, + "eval_runtime": 38.0088, + "eval_samples_per_second": 26.31, + "eval_steps_per_second": 1.105, + "step": 2880 + }, + { + "epoch": 2.9594247560349256, + "grad_norm": 2.7072103023529053, + "learning_rate": 1.0468214693566807e-07, + "loss": 1.3132, + "step": 2881 + }, + { + "epoch": 2.9604519774011298, + "grad_norm": 2.4957451820373535, + "learning_rate": 1.0277883517320137e-07, + "loss": 1.2867, + "step": 2882 + }, + { + "epoch": 2.9614791987673343, + "grad_norm": 2.572026014328003, + "learning_rate": 1.0087552341073468e-07, + "loss": 1.2959, + "step": 2883 + }, + { + "epoch": 2.962506420133539, + "grad_norm": 2.6354286670684814, + "learning_rate": 9.8972211648268e-08, + "loss": 1.2952, + "step": 2884 + }, + { + "epoch": 2.963533641499743, + "grad_norm": 2.4952967166900635, + "learning_rate": 9.70688998858013e-08, + "loss": 1.2844, + "step": 2885 + }, + { + "epoch": 2.9645608628659477, + "grad_norm": 2.5860595703125, + "learning_rate": 9.516558812333461e-08, + "loss": 1.2462, + "step": 2886 + }, + { + "epoch": 2.9655880842321523, + "grad_norm": 2.6458609104156494, + "learning_rate": 9.32622763608679e-08, + "loss": 1.3064, + "step": 2887 + }, + { + "epoch": 2.9666153055983564, + "grad_norm": 2.5705668926239014, + "learning_rate": 9.135896459840123e-08, + "loss": 1.3074, + "step": 2888 + }, + { + "epoch": 2.967642526964561, + "grad_norm": 2.5741384029388428, + "learning_rate": 8.945565283593454e-08, + "loss": 1.2931, + "step": 2889 + }, + { + "epoch": 2.968669748330765, + "grad_norm": 2.6422197818756104, + "learning_rate": 8.755234107346783e-08, + "loss": 1.3144, + "step": 2890 + }, + { + "epoch": 2.968669748330765, + "eval_loss": 1.7483710050582886, + "eval_runtime": 38.0597, + "eval_samples_per_second": 26.275, + "eval_steps_per_second": 1.104, + "step": 2890 + }, + { + "epoch": 2.9696969696969697, + "grad_norm": 2.4819366931915283, + "learning_rate": 8.564902931100116e-08, + "loss": 1.2642, + "step": 2891 + }, + { + "epoch": 2.970724191063174, + "grad_norm": 2.5297718048095703, + "learning_rate": 8.374571754853445e-08, + "loss": 1.2536, + "step": 2892 + }, + { + "epoch": 2.9717514124293785, + "grad_norm": 2.512047052383423, + "learning_rate": 8.184240578606776e-08, + "loss": 1.3091, + "step": 2893 + }, + { + "epoch": 2.972778633795583, + "grad_norm": 2.7100260257720947, + "learning_rate": 7.993909402360106e-08, + "loss": 1.3269, + "step": 2894 + }, + { + "epoch": 2.973805855161787, + "grad_norm": 2.6287920475006104, + "learning_rate": 7.803578226113438e-08, + "loss": 1.2929, + "step": 2895 + }, + { + "epoch": 2.974833076527992, + "grad_norm": 2.650087356567383, + "learning_rate": 7.613247049866768e-08, + "loss": 1.361, + "step": 2896 + }, + { + "epoch": 2.9758602978941964, + "grad_norm": 2.646580457687378, + "learning_rate": 7.422915873620099e-08, + "loss": 1.3077, + "step": 2897 + }, + { + "epoch": 2.9768875192604005, + "grad_norm": 2.4880261421203613, + "learning_rate": 7.232584697373431e-08, + "loss": 1.276, + "step": 2898 + }, + { + "epoch": 2.977914740626605, + "grad_norm": 2.5483410358428955, + "learning_rate": 7.042253521126761e-08, + "loss": 1.304, + "step": 2899 + }, + { + "epoch": 2.9789419619928097, + "grad_norm": 2.468510389328003, + "learning_rate": 6.851922344880092e-08, + "loss": 1.2682, + "step": 2900 + }, + { + "epoch": 2.9789419619928097, + "eval_loss": 1.7464934587478638, + "eval_runtime": 38.0201, + "eval_samples_per_second": 26.302, + "eval_steps_per_second": 1.105, + "step": 2900 + }, + { + "epoch": 2.979969183359014, + "grad_norm": 2.502958059310913, + "learning_rate": 6.661591168633423e-08, + "loss": 1.2744, + "step": 2901 + }, + { + "epoch": 2.9809964047252184, + "grad_norm": 2.8128011226654053, + "learning_rate": 6.471259992386754e-08, + "loss": 1.3449, + "step": 2902 + }, + { + "epoch": 2.9820236260914226, + "grad_norm": 2.7207062244415283, + "learning_rate": 6.280928816140083e-08, + "loss": 1.3282, + "step": 2903 + }, + { + "epoch": 2.983050847457627, + "grad_norm": 2.7468044757843018, + "learning_rate": 6.090597639893416e-08, + "loss": 1.2927, + "step": 2904 + }, + { + "epoch": 2.9840780688238313, + "grad_norm": 2.5978646278381348, + "learning_rate": 5.900266463646746e-08, + "loss": 1.2789, + "step": 2905 + }, + { + "epoch": 2.985105290190036, + "grad_norm": 2.6372761726379395, + "learning_rate": 5.709935287400077e-08, + "loss": 1.2858, + "step": 2906 + }, + { + "epoch": 2.9861325115562405, + "grad_norm": 2.4788858890533447, + "learning_rate": 5.519604111153407e-08, + "loss": 1.2726, + "step": 2907 + }, + { + "epoch": 2.9871597329224446, + "grad_norm": 2.4479143619537354, + "learning_rate": 5.329272934906738e-08, + "loss": 1.2306, + "step": 2908 + }, + { + "epoch": 2.988186954288649, + "grad_norm": 2.542419910430908, + "learning_rate": 5.1389417586600685e-08, + "loss": 1.2682, + "step": 2909 + }, + { + "epoch": 2.989214175654854, + "grad_norm": 2.5090270042419434, + "learning_rate": 4.9486105824134e-08, + "loss": 1.2609, + "step": 2910 + }, + { + "epoch": 2.989214175654854, + "eval_loss": 1.751560091972351, + "eval_runtime": 37.6685, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 2910 + }, + { + "epoch": 2.990241397021058, + "grad_norm": 2.6508572101593018, + "learning_rate": 4.7582794061667304e-08, + "loss": 1.3379, + "step": 2911 + }, + { + "epoch": 2.9912686183872625, + "grad_norm": 2.511770486831665, + "learning_rate": 4.5679482299200614e-08, + "loss": 1.3058, + "step": 2912 + }, + { + "epoch": 2.992295839753467, + "grad_norm": 2.592280149459839, + "learning_rate": 4.377617053673392e-08, + "loss": 1.3348, + "step": 2913 + }, + { + "epoch": 2.9933230611196713, + "grad_norm": 2.5459935665130615, + "learning_rate": 4.187285877426723e-08, + "loss": 1.3159, + "step": 2914 + }, + { + "epoch": 2.994350282485876, + "grad_norm": 2.5060341358184814, + "learning_rate": 3.996954701180053e-08, + "loss": 1.2869, + "step": 2915 + }, + { + "epoch": 2.99537750385208, + "grad_norm": 2.4990108013153076, + "learning_rate": 3.806623524933384e-08, + "loss": 1.2845, + "step": 2916 + }, + { + "epoch": 2.9964047252182846, + "grad_norm": 2.636683940887451, + "learning_rate": 3.6162923486867156e-08, + "loss": 1.3399, + "step": 2917 + }, + { + "epoch": 2.9974319465844887, + "grad_norm": 2.4763123989105225, + "learning_rate": 3.425961172440046e-08, + "loss": 1.2782, + "step": 2918 + }, + { + "epoch": 2.9984591679506933, + "grad_norm": 2.456406354904175, + "learning_rate": 3.235629996193377e-08, + "loss": 1.2783, + "step": 2919 + } + ], + "logging_steps": 1, + "max_steps": 2919, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 9.099426165927444e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/zero_to_fp32.py b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-2919/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/generation_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/latest b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/latest new file mode 100644 index 0000000000000000000000000000000000000000..e5bdf58d4f29d34e909da25905fad376f73e7c29 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/latest @@ -0,0 +1 @@ +global_step400 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/model.safetensors.index.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..d5b3eaa75e01e0564b8641c8f89a308c359a8a6c --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf3509a33c2e615b42fa452a51f3b448cd738037d76a4830d79b89c785555ed0 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..f31c91e6fc0dcc1eb5c101a4c40e35f5b315d26f --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c32df687c2bf9ce5fb1beafdcec2d9cfb49bc02e3a06d089279c235d40a82ca8 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..c1492af50e3afb56d1aecd7362d9e876ea1c01da --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fef8236d6b1951c0a5c3e8cf66c0be4ae93551b780648507e1981a0117386378 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..9b9e69a656f49f8683f204f4ad9d7d9ee3bc9742 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5b069fde04a452bbf7f8ad607daf86d02b811d4f07a94a5d930be87ba8885d7f +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/special_tokens_map.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/tokenizer_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0b782d4602b361f7577abb52b710a8c0e088588d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/tokenizer_config.json @@ -0,0 +1,2078 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128257": { + "content": "🅁", + "lstrip": true, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/trainer_state.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4c01a3d2ccd5f77f403ec9ae4eaad0cb0b4ad616 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/trainer_state.json @@ -0,0 +1,3153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.4108885464817668, + "eval_steps": 10, + "global_step": 400, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7525, + "step": 1 + }, + { + "epoch": 0.002054442732408834, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7605, + "step": 2 + }, + { + "epoch": 0.0030816640986132513, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8008, + "step": 3 + }, + { + "epoch": 0.004108885464817668, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.809, + "step": 4 + }, + { + "epoch": 0.005136106831022085, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7841, + "step": 5 + }, + { + "epoch": 0.0061633281972265025, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7924, + "step": 6 + }, + { + "epoch": 0.007190549563430919, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7732, + "step": 7 + }, + { + "epoch": 0.008217770929635337, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7953, + "step": 8 + }, + { + "epoch": 0.009244992295839754, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7747, + "step": 9 + }, + { + "epoch": 0.01027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7909, + "step": 10 + }, + { + "epoch": 0.01027221366204417, + "eval_loss": 2.813098192214966, + "eval_runtime": 38.1187, + "eval_samples_per_second": 26.234, + "eval_steps_per_second": 1.102, + "step": 10 + }, + { + "epoch": 0.011299435028248588, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7592, + "step": 11 + }, + { + "epoch": 0.012326656394453005, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7394, + "step": 12 + }, + { + "epoch": 0.01335387776065742, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8111, + "step": 13 + }, + { + "epoch": 0.014381099126861838, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7786, + "step": 14 + }, + { + "epoch": 0.015408320493066256, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7711, + "step": 15 + }, + { + "epoch": 0.016435541859270673, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7622, + "step": 16 + }, + { + "epoch": 0.01746276322547509, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7291, + "step": 17 + }, + { + "epoch": 0.01848998459167951, + "grad_norm": 6.8249125480651855, + "learning_rate": 3.424657534246575e-08, + "loss": 2.7995, + "step": 18 + }, + { + "epoch": 0.019517205957883924, + "grad_norm": 5.581949710845947, + "learning_rate": 5.136986301369863e-08, + "loss": 2.5998, + "step": 19 + }, + { + "epoch": 0.02054442732408834, + "grad_norm": 3.0888617038726807, + "learning_rate": 6.84931506849315e-08, + "loss": 2.6031, + "step": 20 + }, + { + "epoch": 0.02054442732408834, + "eval_loss": 2.60642147064209, + "eval_runtime": 38.2226, + "eval_samples_per_second": 26.163, + "eval_steps_per_second": 1.099, + "step": 20 + }, + { + "epoch": 0.02157164869029276, + "grad_norm": 4.42941427230835, + "learning_rate": 8.561643835616439e-08, + "loss": 2.5816, + "step": 21 + }, + { + "epoch": 0.022598870056497175, + "grad_norm": 2.359813690185547, + "learning_rate": 1.0273972602739726e-07, + "loss": 2.5352, + "step": 22 + }, + { + "epoch": 0.02362609142270159, + "grad_norm": 2.9022927284240723, + "learning_rate": 1.1986301369863014e-07, + "loss": 2.4641, + "step": 23 + }, + { + "epoch": 0.02465331278890601, + "grad_norm": 1.6895997524261475, + "learning_rate": 1.36986301369863e-07, + "loss": 2.4582, + "step": 24 + }, + { + "epoch": 0.025680534155110426, + "grad_norm": 2.4555776119232178, + "learning_rate": 1.541095890410959e-07, + "loss": 2.4171, + "step": 25 + }, + { + "epoch": 0.02670775552131484, + "grad_norm": 1.9735779762268066, + "learning_rate": 1.7123287671232878e-07, + "loss": 2.4249, + "step": 26 + }, + { + "epoch": 0.02773497688751926, + "grad_norm": 2.71161150932312, + "learning_rate": 1.8835616438356165e-07, + "loss": 2.3791, + "step": 27 + }, + { + "epoch": 0.028762198253723677, + "grad_norm": 1.8482733964920044, + "learning_rate": 2.0547945205479452e-07, + "loss": 2.374, + "step": 28 + }, + { + "epoch": 0.029789419619928096, + "grad_norm": 2.191246747970581, + "learning_rate": 2.226027397260274e-07, + "loss": 2.3557, + "step": 29 + }, + { + "epoch": 0.030816640986132512, + "grad_norm": 2.0575971603393555, + "learning_rate": 2.397260273972603e-07, + "loss": 2.3315, + "step": 30 + }, + { + "epoch": 0.030816640986132512, + "eval_loss": 2.399639844894409, + "eval_runtime": 37.737, + "eval_samples_per_second": 26.499, + "eval_steps_per_second": 1.113, + "step": 30 + }, + { + "epoch": 0.03184386235233693, + "grad_norm": 2.0346758365631104, + "learning_rate": 2.568493150684932e-07, + "loss": 2.2791, + "step": 31 + }, + { + "epoch": 0.03287108371854135, + "grad_norm": 1.8710602521896362, + "learning_rate": 2.73972602739726e-07, + "loss": 2.3219, + "step": 32 + }, + { + "epoch": 0.03389830508474576, + "grad_norm": 1.9982675313949585, + "learning_rate": 2.910958904109589e-07, + "loss": 2.273, + "step": 33 + }, + { + "epoch": 0.03492552645095018, + "grad_norm": 2.092076301574707, + "learning_rate": 3.082191780821918e-07, + "loss": 2.2555, + "step": 34 + }, + { + "epoch": 0.035952747817154594, + "grad_norm": 1.870373249053955, + "learning_rate": 3.2534246575342466e-07, + "loss": 2.2314, + "step": 35 + }, + { + "epoch": 0.03697996918335902, + "grad_norm": 1.8378052711486816, + "learning_rate": 3.4246575342465755e-07, + "loss": 2.1244, + "step": 36 + }, + { + "epoch": 0.03800719054956343, + "grad_norm": 2.0156054496765137, + "learning_rate": 3.595890410958904e-07, + "loss": 2.2502, + "step": 37 + }, + { + "epoch": 0.03903441191576785, + "grad_norm": 1.6175312995910645, + "learning_rate": 3.767123287671233e-07, + "loss": 2.1566, + "step": 38 + }, + { + "epoch": 0.040061633281972264, + "grad_norm": 1.9051450490951538, + "learning_rate": 3.938356164383562e-07, + "loss": 2.167, + "step": 39 + }, + { + "epoch": 0.04108885464817668, + "grad_norm": 2.383063793182373, + "learning_rate": 4.1095890410958903e-07, + "loss": 2.2099, + "step": 40 + }, + { + "epoch": 0.04108885464817668, + "eval_loss": 2.2763354778289795, + "eval_runtime": 37.8085, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 40 + }, + { + "epoch": 0.042116076014381096, + "grad_norm": 2.1310720443725586, + "learning_rate": 4.2808219178082193e-07, + "loss": 2.1714, + "step": 41 + }, + { + "epoch": 0.04314329738058552, + "grad_norm": 1.771941900253296, + "learning_rate": 4.452054794520548e-07, + "loss": 2.0895, + "step": 42 + }, + { + "epoch": 0.044170518746789934, + "grad_norm": 2.3784172534942627, + "learning_rate": 4.6232876712328767e-07, + "loss": 2.1555, + "step": 43 + }, + { + "epoch": 0.04519774011299435, + "grad_norm": 1.6692813634872437, + "learning_rate": 4.794520547945206e-07, + "loss": 2.11, + "step": 44 + }, + { + "epoch": 0.046224961479198766, + "grad_norm": 2.1780643463134766, + "learning_rate": 4.965753424657534e-07, + "loss": 2.1256, + "step": 45 + }, + { + "epoch": 0.04725218284540318, + "grad_norm": 1.9390568733215332, + "learning_rate": 5.136986301369864e-07, + "loss": 2.0971, + "step": 46 + }, + { + "epoch": 0.048279404211607604, + "grad_norm": 2.276216506958008, + "learning_rate": 5.308219178082192e-07, + "loss": 2.1479, + "step": 47 + }, + { + "epoch": 0.04930662557781202, + "grad_norm": 2.148573160171509, + "learning_rate": 5.47945205479452e-07, + "loss": 2.0808, + "step": 48 + }, + { + "epoch": 0.050333846944016436, + "grad_norm": 2.259422540664673, + "learning_rate": 5.65068493150685e-07, + "loss": 2.0823, + "step": 49 + }, + { + "epoch": 0.05136106831022085, + "grad_norm": 1.7703359127044678, + "learning_rate": 5.821917808219178e-07, + "loss": 2.0236, + "step": 50 + }, + { + "epoch": 0.05136106831022085, + "eval_loss": 2.2037274837493896, + "eval_runtime": 37.8385, + "eval_samples_per_second": 26.428, + "eval_steps_per_second": 1.11, + "step": 50 + }, + { + "epoch": 0.05238828967642527, + "grad_norm": 2.56268310546875, + "learning_rate": 5.993150684931507e-07, + "loss": 2.0228, + "step": 51 + }, + { + "epoch": 0.05341551104262968, + "grad_norm": 2.5256848335266113, + "learning_rate": 6.164383561643836e-07, + "loss": 2.1401, + "step": 52 + }, + { + "epoch": 0.054442732408834106, + "grad_norm": 2.6329638957977295, + "learning_rate": 6.335616438356165e-07, + "loss": 2.0107, + "step": 53 + }, + { + "epoch": 0.05546995377503852, + "grad_norm": 2.477762222290039, + "learning_rate": 6.506849315068493e-07, + "loss": 2.0648, + "step": 54 + }, + { + "epoch": 0.05649717514124294, + "grad_norm": 1.8869198560714722, + "learning_rate": 6.678082191780823e-07, + "loss": 2.0214, + "step": 55 + }, + { + "epoch": 0.05752439650744735, + "grad_norm": 2.707345485687256, + "learning_rate": 6.849315068493151e-07, + "loss": 1.9893, + "step": 56 + }, + { + "epoch": 0.05855161787365177, + "grad_norm": 2.067230701446533, + "learning_rate": 7.020547945205481e-07, + "loss": 2.0308, + "step": 57 + }, + { + "epoch": 0.05957883923985619, + "grad_norm": 2.1961982250213623, + "learning_rate": 7.191780821917808e-07, + "loss": 1.948, + "step": 58 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 2.1032280921936035, + "learning_rate": 7.363013698630137e-07, + "loss": 2.0098, + "step": 59 + }, + { + "epoch": 0.061633281972265024, + "grad_norm": 2.5191144943237305, + "learning_rate": 7.534246575342466e-07, + "loss": 2.0183, + "step": 60 + }, + { + "epoch": 0.061633281972265024, + "eval_loss": 2.1474344730377197, + "eval_runtime": 37.8631, + "eval_samples_per_second": 26.411, + "eval_steps_per_second": 1.109, + "step": 60 + }, + { + "epoch": 0.06266050333846944, + "grad_norm": 2.177739143371582, + "learning_rate": 7.705479452054795e-07, + "loss": 1.9702, + "step": 61 + }, + { + "epoch": 0.06368772470467386, + "grad_norm": 2.499784231185913, + "learning_rate": 7.876712328767124e-07, + "loss": 1.9825, + "step": 62 + }, + { + "epoch": 0.06471494607087827, + "grad_norm": 2.181591749191284, + "learning_rate": 8.047945205479453e-07, + "loss": 2.0414, + "step": 63 + }, + { + "epoch": 0.0657421674370827, + "grad_norm": 1.9575999975204468, + "learning_rate": 8.219178082191781e-07, + "loss": 2.0115, + "step": 64 + }, + { + "epoch": 0.0667693888032871, + "grad_norm": 2.3431336879730225, + "learning_rate": 8.39041095890411e-07, + "loss": 1.8781, + "step": 65 + }, + { + "epoch": 0.06779661016949153, + "grad_norm": 2.042747974395752, + "learning_rate": 8.561643835616439e-07, + "loss": 1.9798, + "step": 66 + }, + { + "epoch": 0.06882383153569595, + "grad_norm": 2.784104585647583, + "learning_rate": 8.732876712328768e-07, + "loss": 1.9274, + "step": 67 + }, + { + "epoch": 0.06985105290190036, + "grad_norm": 1.9781445264816284, + "learning_rate": 8.904109589041097e-07, + "loss": 2.0036, + "step": 68 + }, + { + "epoch": 0.07087827426810478, + "grad_norm": 2.631537437438965, + "learning_rate": 9.075342465753426e-07, + "loss": 1.938, + "step": 69 + }, + { + "epoch": 0.07190549563430919, + "grad_norm": 1.902701497077942, + "learning_rate": 9.246575342465753e-07, + "loss": 2.0545, + "step": 70 + }, + { + "epoch": 0.07190549563430919, + "eval_loss": 2.115269422531128, + "eval_runtime": 37.887, + "eval_samples_per_second": 26.394, + "eval_steps_per_second": 1.109, + "step": 70 + }, + { + "epoch": 0.07293271700051361, + "grad_norm": 1.9323904514312744, + "learning_rate": 9.417808219178083e-07, + "loss": 1.9463, + "step": 71 + }, + { + "epoch": 0.07395993836671803, + "grad_norm": 2.1631956100463867, + "learning_rate": 9.589041095890411e-07, + "loss": 1.8982, + "step": 72 + }, + { + "epoch": 0.07498715973292244, + "grad_norm": 1.959518313407898, + "learning_rate": 9.76027397260274e-07, + "loss": 1.9538, + "step": 73 + }, + { + "epoch": 0.07601438109912687, + "grad_norm": 1.9834569692611694, + "learning_rate": 9.931506849315068e-07, + "loss": 1.9257, + "step": 74 + }, + { + "epoch": 0.07704160246533127, + "grad_norm": 2.2818732261657715, + "learning_rate": 1.0102739726027399e-06, + "loss": 1.9318, + "step": 75 + }, + { + "epoch": 0.0780688238315357, + "grad_norm": 1.7696406841278076, + "learning_rate": 1.0273972602739727e-06, + "loss": 1.912, + "step": 76 + }, + { + "epoch": 0.07909604519774012, + "grad_norm": 2.396303176879883, + "learning_rate": 1.0445205479452056e-06, + "loss": 1.9482, + "step": 77 + }, + { + "epoch": 0.08012326656394453, + "grad_norm": 2.2371881008148193, + "learning_rate": 1.0616438356164384e-06, + "loss": 1.9448, + "step": 78 + }, + { + "epoch": 0.08115048793014895, + "grad_norm": 2.40077543258667, + "learning_rate": 1.0787671232876712e-06, + "loss": 1.8764, + "step": 79 + }, + { + "epoch": 0.08217770929635336, + "grad_norm": 2.267916679382324, + "learning_rate": 1.095890410958904e-06, + "loss": 1.9478, + "step": 80 + }, + { + "epoch": 0.08217770929635336, + "eval_loss": 2.0851893424987793, + "eval_runtime": 37.9176, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 80 + }, + { + "epoch": 0.08320493066255778, + "grad_norm": 2.286259651184082, + "learning_rate": 1.1130136986301371e-06, + "loss": 1.8856, + "step": 81 + }, + { + "epoch": 0.08423215202876219, + "grad_norm": 2.160848379135132, + "learning_rate": 1.13013698630137e-06, + "loss": 1.9196, + "step": 82 + }, + { + "epoch": 0.08525937339496661, + "grad_norm": 2.3618931770324707, + "learning_rate": 1.1472602739726028e-06, + "loss": 1.9229, + "step": 83 + }, + { + "epoch": 0.08628659476117104, + "grad_norm": 2.1963961124420166, + "learning_rate": 1.1643835616438357e-06, + "loss": 1.8998, + "step": 84 + }, + { + "epoch": 0.08731381612737545, + "grad_norm": 2.197981357574463, + "learning_rate": 1.1815068493150685e-06, + "loss": 1.8716, + "step": 85 + }, + { + "epoch": 0.08834103749357987, + "grad_norm": 2.3562612533569336, + "learning_rate": 1.1986301369863014e-06, + "loss": 1.9336, + "step": 86 + }, + { + "epoch": 0.08936825885978428, + "grad_norm": 2.49712872505188, + "learning_rate": 1.2157534246575344e-06, + "loss": 1.8827, + "step": 87 + }, + { + "epoch": 0.0903954802259887, + "grad_norm": 2.315244436264038, + "learning_rate": 1.2328767123287673e-06, + "loss": 1.8801, + "step": 88 + }, + { + "epoch": 0.09142270159219312, + "grad_norm": 2.774587869644165, + "learning_rate": 1.25e-06, + "loss": 1.8972, + "step": 89 + }, + { + "epoch": 0.09244992295839753, + "grad_norm": 2.262364387512207, + "learning_rate": 1.267123287671233e-06, + "loss": 1.8426, + "step": 90 + }, + { + "epoch": 0.09244992295839753, + "eval_loss": 2.069462299346924, + "eval_runtime": 37.9319, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 90 + }, + { + "epoch": 0.09347714432460195, + "grad_norm": 2.6195597648620605, + "learning_rate": 1.284246575342466e-06, + "loss": 1.871, + "step": 91 + }, + { + "epoch": 0.09450436569080636, + "grad_norm": 2.051847219467163, + "learning_rate": 1.3013698630136986e-06, + "loss": 1.8922, + "step": 92 + }, + { + "epoch": 0.09553158705701079, + "grad_norm": 2.5930886268615723, + "learning_rate": 1.3184931506849317e-06, + "loss": 1.8818, + "step": 93 + }, + { + "epoch": 0.09655880842321521, + "grad_norm": 2.304164171218872, + "learning_rate": 1.3356164383561645e-06, + "loss": 1.8779, + "step": 94 + }, + { + "epoch": 0.09758602978941962, + "grad_norm": 2.187354564666748, + "learning_rate": 1.3527397260273976e-06, + "loss": 1.9272, + "step": 95 + }, + { + "epoch": 0.09861325115562404, + "grad_norm": 2.2492458820343018, + "learning_rate": 1.3698630136986302e-06, + "loss": 1.805, + "step": 96 + }, + { + "epoch": 0.09964047252182845, + "grad_norm": 2.3325352668762207, + "learning_rate": 1.386986301369863e-06, + "loss": 1.8817, + "step": 97 + }, + { + "epoch": 0.10066769388803287, + "grad_norm": 2.1708617210388184, + "learning_rate": 1.4041095890410961e-06, + "loss": 1.7956, + "step": 98 + }, + { + "epoch": 0.1016949152542373, + "grad_norm": 2.4267706871032715, + "learning_rate": 1.421232876712329e-06, + "loss": 1.8116, + "step": 99 + }, + { + "epoch": 0.1027221366204417, + "grad_norm": 2.3286805152893066, + "learning_rate": 1.4383561643835616e-06, + "loss": 1.8299, + "step": 100 + }, + { + "epoch": 0.1027221366204417, + "eval_loss": 2.0549135208129883, + "eval_runtime": 37.935, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 100 + }, + { + "epoch": 0.10374935798664613, + "grad_norm": 2.5391643047332764, + "learning_rate": 1.4554794520547946e-06, + "loss": 1.8521, + "step": 101 + }, + { + "epoch": 0.10477657935285054, + "grad_norm": 2.018380880355835, + "learning_rate": 1.4726027397260275e-06, + "loss": 1.8209, + "step": 102 + }, + { + "epoch": 0.10580380071905496, + "grad_norm": 2.3706746101379395, + "learning_rate": 1.4897260273972605e-06, + "loss": 1.79, + "step": 103 + }, + { + "epoch": 0.10683102208525937, + "grad_norm": 2.1290764808654785, + "learning_rate": 1.5068493150684932e-06, + "loss": 1.8141, + "step": 104 + }, + { + "epoch": 0.10785824345146379, + "grad_norm": 2.2768120765686035, + "learning_rate": 1.5239726027397262e-06, + "loss": 1.8826, + "step": 105 + }, + { + "epoch": 0.10888546481766821, + "grad_norm": 2.316584348678589, + "learning_rate": 1.541095890410959e-06, + "loss": 1.8604, + "step": 106 + }, + { + "epoch": 0.10991268618387262, + "grad_norm": 2.14679217338562, + "learning_rate": 1.5582191780821921e-06, + "loss": 1.865, + "step": 107 + }, + { + "epoch": 0.11093990755007704, + "grad_norm": 2.3288021087646484, + "learning_rate": 1.5753424657534248e-06, + "loss": 1.8512, + "step": 108 + }, + { + "epoch": 0.11196712891628145, + "grad_norm": 2.300482988357544, + "learning_rate": 1.5924657534246576e-06, + "loss": 1.7455, + "step": 109 + }, + { + "epoch": 0.11299435028248588, + "grad_norm": 2.4226229190826416, + "learning_rate": 1.6095890410958907e-06, + "loss": 1.8058, + "step": 110 + }, + { + "epoch": 0.11299435028248588, + "eval_loss": 2.025261640548706, + "eval_runtime": 37.6779, + "eval_samples_per_second": 26.541, + "eval_steps_per_second": 1.115, + "step": 110 + }, + { + "epoch": 0.1140215716486903, + "grad_norm": 2.0930111408233643, + "learning_rate": 1.6267123287671235e-06, + "loss": 1.7375, + "step": 111 + }, + { + "epoch": 0.1150487930148947, + "grad_norm": 2.4973645210266113, + "learning_rate": 1.6438356164383561e-06, + "loss": 1.8268, + "step": 112 + }, + { + "epoch": 0.11607601438109913, + "grad_norm": 2.3295114040374756, + "learning_rate": 1.6609589041095892e-06, + "loss": 1.7882, + "step": 113 + }, + { + "epoch": 0.11710323574730354, + "grad_norm": 2.3693318367004395, + "learning_rate": 1.678082191780822e-06, + "loss": 1.8192, + "step": 114 + }, + { + "epoch": 0.11813045711350796, + "grad_norm": 2.553637742996216, + "learning_rate": 1.695205479452055e-06, + "loss": 1.7491, + "step": 115 + }, + { + "epoch": 0.11915767847971238, + "grad_norm": 2.303945541381836, + "learning_rate": 1.7123287671232877e-06, + "loss": 1.8737, + "step": 116 + }, + { + "epoch": 0.12018489984591679, + "grad_norm": 2.250356435775757, + "learning_rate": 1.7294520547945206e-06, + "loss": 1.8039, + "step": 117 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 2.3854403495788574, + "learning_rate": 1.7465753424657536e-06, + "loss": 1.8357, + "step": 118 + }, + { + "epoch": 0.12223934257832562, + "grad_norm": 2.4049572944641113, + "learning_rate": 1.7636986301369865e-06, + "loss": 1.8209, + "step": 119 + }, + { + "epoch": 0.12326656394453005, + "grad_norm": 2.313551425933838, + "learning_rate": 1.7808219178082193e-06, + "loss": 1.9026, + "step": 120 + }, + { + "epoch": 0.12326656394453005, + "eval_loss": 2.012171745300293, + "eval_runtime": 37.7908, + "eval_samples_per_second": 26.461, + "eval_steps_per_second": 1.111, + "step": 120 + }, + { + "epoch": 0.12429378531073447, + "grad_norm": 2.518320322036743, + "learning_rate": 1.7979452054794521e-06, + "loss": 1.8241, + "step": 121 + }, + { + "epoch": 0.12532100667693888, + "grad_norm": 2.6504592895507812, + "learning_rate": 1.8150684931506852e-06, + "loss": 1.8091, + "step": 122 + }, + { + "epoch": 0.1263482280431433, + "grad_norm": 2.556610107421875, + "learning_rate": 1.832191780821918e-06, + "loss": 1.7547, + "step": 123 + }, + { + "epoch": 0.12737544940934772, + "grad_norm": 2.247633695602417, + "learning_rate": 1.8493150684931507e-06, + "loss": 1.7923, + "step": 124 + }, + { + "epoch": 0.12840267077555212, + "grad_norm": 2.195596933364868, + "learning_rate": 1.8664383561643837e-06, + "loss": 1.747, + "step": 125 + }, + { + "epoch": 0.12942989214175654, + "grad_norm": 2.121396780014038, + "learning_rate": 1.8835616438356166e-06, + "loss": 1.8088, + "step": 126 + }, + { + "epoch": 0.13045711350796096, + "grad_norm": 2.4572501182556152, + "learning_rate": 1.9006849315068496e-06, + "loss": 1.7285, + "step": 127 + }, + { + "epoch": 0.1314843348741654, + "grad_norm": 2.2590177059173584, + "learning_rate": 1.9178082191780823e-06, + "loss": 1.7929, + "step": 128 + }, + { + "epoch": 0.1325115562403698, + "grad_norm": 2.5099234580993652, + "learning_rate": 1.9349315068493153e-06, + "loss": 1.8365, + "step": 129 + }, + { + "epoch": 0.1335387776065742, + "grad_norm": 2.3064465522766113, + "learning_rate": 1.952054794520548e-06, + "loss": 1.8351, + "step": 130 + }, + { + "epoch": 0.1335387776065742, + "eval_loss": 2.0043866634368896, + "eval_runtime": 37.8661, + "eval_samples_per_second": 26.409, + "eval_steps_per_second": 1.109, + "step": 130 + }, + { + "epoch": 0.13456599897277863, + "grad_norm": 2.5977718830108643, + "learning_rate": 1.969178082191781e-06, + "loss": 1.7603, + "step": 131 + }, + { + "epoch": 0.13559322033898305, + "grad_norm": 2.23221755027771, + "learning_rate": 1.9863013698630136e-06, + "loss": 1.8237, + "step": 132 + }, + { + "epoch": 0.13662044170518747, + "grad_norm": 2.247044324874878, + "learning_rate": 2.0034246575342467e-06, + "loss": 1.8006, + "step": 133 + }, + { + "epoch": 0.1376476630713919, + "grad_norm": 2.1796107292175293, + "learning_rate": 2.0205479452054797e-06, + "loss": 1.7354, + "step": 134 + }, + { + "epoch": 0.1386748844375963, + "grad_norm": 2.3269693851470947, + "learning_rate": 2.037671232876713e-06, + "loss": 1.783, + "step": 135 + }, + { + "epoch": 0.1397021058038007, + "grad_norm": 2.2903053760528564, + "learning_rate": 2.0547945205479454e-06, + "loss": 1.8072, + "step": 136 + }, + { + "epoch": 0.14072932717000514, + "grad_norm": 2.634150981903076, + "learning_rate": 2.071917808219178e-06, + "loss": 1.7251, + "step": 137 + }, + { + "epoch": 0.14175654853620956, + "grad_norm": 2.42579984664917, + "learning_rate": 2.089041095890411e-06, + "loss": 1.7997, + "step": 138 + }, + { + "epoch": 0.14278376990241398, + "grad_norm": 2.615828514099121, + "learning_rate": 2.106164383561644e-06, + "loss": 1.766, + "step": 139 + }, + { + "epoch": 0.14381099126861838, + "grad_norm": 2.312714099884033, + "learning_rate": 2.123287671232877e-06, + "loss": 1.7538, + "step": 140 + }, + { + "epoch": 0.14381099126861838, + "eval_loss": 1.9918214082717896, + "eval_runtime": 37.8698, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 140 + }, + { + "epoch": 0.1448382126348228, + "grad_norm": 2.7540030479431152, + "learning_rate": 2.14041095890411e-06, + "loss": 1.7856, + "step": 141 + }, + { + "epoch": 0.14586543400102722, + "grad_norm": 2.289191246032715, + "learning_rate": 2.1575342465753425e-06, + "loss": 1.7288, + "step": 142 + }, + { + "epoch": 0.14689265536723164, + "grad_norm": 2.788001775741577, + "learning_rate": 2.1746575342465755e-06, + "loss": 1.8703, + "step": 143 + }, + { + "epoch": 0.14791987673343607, + "grad_norm": 2.4128854274749756, + "learning_rate": 2.191780821917808e-06, + "loss": 1.7461, + "step": 144 + }, + { + "epoch": 0.14894709809964046, + "grad_norm": 2.4108099937438965, + "learning_rate": 2.2089041095890412e-06, + "loss": 1.73, + "step": 145 + }, + { + "epoch": 0.14997431946584489, + "grad_norm": 2.4315948486328125, + "learning_rate": 2.2260273972602743e-06, + "loss": 1.7303, + "step": 146 + }, + { + "epoch": 0.1510015408320493, + "grad_norm": 2.494056224822998, + "learning_rate": 2.243150684931507e-06, + "loss": 1.7616, + "step": 147 + }, + { + "epoch": 0.15202876219825373, + "grad_norm": 2.4557394981384277, + "learning_rate": 2.26027397260274e-06, + "loss": 1.7655, + "step": 148 + }, + { + "epoch": 0.15305598356445815, + "grad_norm": 2.4114091396331787, + "learning_rate": 2.2773972602739726e-06, + "loss": 1.8538, + "step": 149 + }, + { + "epoch": 0.15408320493066255, + "grad_norm": 2.609219789505005, + "learning_rate": 2.2945205479452057e-06, + "loss": 1.7767, + "step": 150 + }, + { + "epoch": 0.15408320493066255, + "eval_loss": 1.981367588043213, + "eval_runtime": 37.9562, + "eval_samples_per_second": 26.346, + "eval_steps_per_second": 1.107, + "step": 150 + }, + { + "epoch": 0.15511042629686697, + "grad_norm": 2.213622808456421, + "learning_rate": 2.3116438356164387e-06, + "loss": 1.687, + "step": 151 + }, + { + "epoch": 0.1561376476630714, + "grad_norm": 2.6517975330352783, + "learning_rate": 2.3287671232876713e-06, + "loss": 1.7311, + "step": 152 + }, + { + "epoch": 0.15716486902927582, + "grad_norm": 2.5174458026885986, + "learning_rate": 2.3458904109589044e-06, + "loss": 1.7835, + "step": 153 + }, + { + "epoch": 0.15819209039548024, + "grad_norm": 2.485480308532715, + "learning_rate": 2.363013698630137e-06, + "loss": 1.7749, + "step": 154 + }, + { + "epoch": 0.15921931176168463, + "grad_norm": 2.5386061668395996, + "learning_rate": 2.38013698630137e-06, + "loss": 1.7223, + "step": 155 + }, + { + "epoch": 0.16024653312788906, + "grad_norm": 2.2895941734313965, + "learning_rate": 2.3972602739726027e-06, + "loss": 1.6845, + "step": 156 + }, + { + "epoch": 0.16127375449409348, + "grad_norm": 2.5453789234161377, + "learning_rate": 2.4143835616438358e-06, + "loss": 1.7627, + "step": 157 + }, + { + "epoch": 0.1623009758602979, + "grad_norm": 2.5322303771972656, + "learning_rate": 2.431506849315069e-06, + "loss": 1.764, + "step": 158 + }, + { + "epoch": 0.1633281972265023, + "grad_norm": 2.2669899463653564, + "learning_rate": 2.4486301369863015e-06, + "loss": 1.7487, + "step": 159 + }, + { + "epoch": 0.16435541859270672, + "grad_norm": 2.4577853679656982, + "learning_rate": 2.4657534246575345e-06, + "loss": 1.7635, + "step": 160 + }, + { + "epoch": 0.16435541859270672, + "eval_loss": 1.9793355464935303, + "eval_runtime": 37.8576, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 160 + }, + { + "epoch": 0.16538263995891114, + "grad_norm": 2.354893922805786, + "learning_rate": 2.482876712328767e-06, + "loss": 1.7538, + "step": 161 + }, + { + "epoch": 0.16640986132511557, + "grad_norm": 2.3255856037139893, + "learning_rate": 2.5e-06, + "loss": 1.7277, + "step": 162 + }, + { + "epoch": 0.16743708269132, + "grad_norm": 2.3501994609832764, + "learning_rate": 2.5171232876712333e-06, + "loss": 1.6985, + "step": 163 + }, + { + "epoch": 0.16846430405752438, + "grad_norm": 2.386627674102783, + "learning_rate": 2.534246575342466e-06, + "loss": 1.7552, + "step": 164 + }, + { + "epoch": 0.1694915254237288, + "grad_norm": 2.2238457202911377, + "learning_rate": 2.551369863013699e-06, + "loss": 1.7805, + "step": 165 + }, + { + "epoch": 0.17051874678993323, + "grad_norm": 2.2653403282165527, + "learning_rate": 2.568493150684932e-06, + "loss": 1.7656, + "step": 166 + }, + { + "epoch": 0.17154596815613765, + "grad_norm": 2.233132839202881, + "learning_rate": 2.585616438356164e-06, + "loss": 1.6997, + "step": 167 + }, + { + "epoch": 0.17257318952234207, + "grad_norm": 2.1696043014526367, + "learning_rate": 2.6027397260273973e-06, + "loss": 1.7487, + "step": 168 + }, + { + "epoch": 0.17360041088854647, + "grad_norm": 2.4479899406433105, + "learning_rate": 2.6198630136986303e-06, + "loss": 1.7957, + "step": 169 + }, + { + "epoch": 0.1746276322547509, + "grad_norm": 2.3054702281951904, + "learning_rate": 2.6369863013698634e-06, + "loss": 1.766, + "step": 170 + }, + { + "epoch": 0.1746276322547509, + "eval_loss": 1.9543790817260742, + "eval_runtime": 37.8837, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 170 + }, + { + "epoch": 0.17565485362095531, + "grad_norm": 2.3591110706329346, + "learning_rate": 2.654109589041096e-06, + "loss": 1.7149, + "step": 171 + }, + { + "epoch": 0.17668207498715974, + "grad_norm": 2.450526237487793, + "learning_rate": 2.671232876712329e-06, + "loss": 1.7074, + "step": 172 + }, + { + "epoch": 0.17770929635336416, + "grad_norm": 2.1511754989624023, + "learning_rate": 2.688356164383562e-06, + "loss": 1.7036, + "step": 173 + }, + { + "epoch": 0.17873651771956856, + "grad_norm": 2.5944275856018066, + "learning_rate": 2.705479452054795e-06, + "loss": 1.6566, + "step": 174 + }, + { + "epoch": 0.17976373908577298, + "grad_norm": 2.4301369190216064, + "learning_rate": 2.7226027397260274e-06, + "loss": 1.7977, + "step": 175 + }, + { + "epoch": 0.1807909604519774, + "grad_norm": 2.287665367126465, + "learning_rate": 2.7397260273972604e-06, + "loss": 1.7526, + "step": 176 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.380032777786255, + "learning_rate": 2.7568493150684935e-06, + "loss": 1.6753, + "step": 177 + }, + { + "epoch": 0.18284540318438625, + "grad_norm": 2.4040510654449463, + "learning_rate": 2.773972602739726e-06, + "loss": 1.758, + "step": 178 + }, + { + "epoch": 0.18387262455059064, + "grad_norm": 2.5074045658111572, + "learning_rate": 2.791095890410959e-06, + "loss": 1.6498, + "step": 179 + }, + { + "epoch": 0.18489984591679506, + "grad_norm": 2.3926949501037598, + "learning_rate": 2.8082191780821922e-06, + "loss": 1.8006, + "step": 180 + }, + { + "epoch": 0.18489984591679506, + "eval_loss": 1.9446762800216675, + "eval_runtime": 37.8572, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 180 + }, + { + "epoch": 0.1859270672829995, + "grad_norm": 2.4991633892059326, + "learning_rate": 2.8253424657534253e-06, + "loss": 1.7649, + "step": 181 + }, + { + "epoch": 0.1869542886492039, + "grad_norm": 2.2558324337005615, + "learning_rate": 2.842465753424658e-06, + "loss": 1.7743, + "step": 182 + }, + { + "epoch": 0.18798151001540833, + "grad_norm": 2.219717025756836, + "learning_rate": 2.8595890410958905e-06, + "loss": 1.7807, + "step": 183 + }, + { + "epoch": 0.18900873138161273, + "grad_norm": 2.406562566757202, + "learning_rate": 2.876712328767123e-06, + "loss": 1.6289, + "step": 184 + }, + { + "epoch": 0.19003595274781715, + "grad_norm": 2.356579303741455, + "learning_rate": 2.8938356164383562e-06, + "loss": 1.6747, + "step": 185 + }, + { + "epoch": 0.19106317411402157, + "grad_norm": 2.4844322204589844, + "learning_rate": 2.9109589041095893e-06, + "loss": 1.6731, + "step": 186 + }, + { + "epoch": 0.192090395480226, + "grad_norm": 2.8179149627685547, + "learning_rate": 2.9280821917808223e-06, + "loss": 1.6786, + "step": 187 + }, + { + "epoch": 0.19311761684643042, + "grad_norm": 2.395416259765625, + "learning_rate": 2.945205479452055e-06, + "loss": 1.7313, + "step": 188 + }, + { + "epoch": 0.1941448382126348, + "grad_norm": 2.848740816116333, + "learning_rate": 2.962328767123288e-06, + "loss": 1.7258, + "step": 189 + }, + { + "epoch": 0.19517205957883924, + "grad_norm": 2.427395820617676, + "learning_rate": 2.979452054794521e-06, + "loss": 1.6633, + "step": 190 + }, + { + "epoch": 0.19517205957883924, + "eval_loss": 1.9443999528884888, + "eval_runtime": 37.8579, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 190 + }, + { + "epoch": 0.19619928094504366, + "grad_norm": 2.5347256660461426, + "learning_rate": 2.9965753424657533e-06, + "loss": 1.7342, + "step": 191 + }, + { + "epoch": 0.19722650231124808, + "grad_norm": 2.8696866035461426, + "learning_rate": 3.0136986301369864e-06, + "loss": 1.7037, + "step": 192 + }, + { + "epoch": 0.1982537236774525, + "grad_norm": 2.427468776702881, + "learning_rate": 3.0308219178082194e-06, + "loss": 1.672, + "step": 193 + }, + { + "epoch": 0.1992809450436569, + "grad_norm": 2.4253737926483154, + "learning_rate": 3.0479452054794525e-06, + "loss": 1.7252, + "step": 194 + }, + { + "epoch": 0.20030816640986132, + "grad_norm": 2.801668643951416, + "learning_rate": 3.065068493150685e-06, + "loss": 1.6694, + "step": 195 + }, + { + "epoch": 0.20133538777606574, + "grad_norm": 2.442263603210449, + "learning_rate": 3.082191780821918e-06, + "loss": 1.6866, + "step": 196 + }, + { + "epoch": 0.20236260914227017, + "grad_norm": 2.777463674545288, + "learning_rate": 3.099315068493151e-06, + "loss": 1.686, + "step": 197 + }, + { + "epoch": 0.2033898305084746, + "grad_norm": 2.491227388381958, + "learning_rate": 3.1164383561643843e-06, + "loss": 1.6414, + "step": 198 + }, + { + "epoch": 0.20441705187467898, + "grad_norm": 2.33585262298584, + "learning_rate": 3.1335616438356165e-06, + "loss": 1.7808, + "step": 199 + }, + { + "epoch": 0.2054442732408834, + "grad_norm": 2.518160581588745, + "learning_rate": 3.1506849315068495e-06, + "loss": 1.6592, + "step": 200 + }, + { + "epoch": 0.2054442732408834, + "eval_loss": 1.9371097087860107, + "eval_runtime": 37.8602, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 200 + }, + { + "epoch": 0.20647149460708783, + "grad_norm": 2.460204839706421, + "learning_rate": 3.167808219178082e-06, + "loss": 1.6668, + "step": 201 + }, + { + "epoch": 0.20749871597329225, + "grad_norm": 2.377824068069458, + "learning_rate": 3.184931506849315e-06, + "loss": 1.6462, + "step": 202 + }, + { + "epoch": 0.20852593733949668, + "grad_norm": 2.426177740097046, + "learning_rate": 3.2020547945205483e-06, + "loss": 1.6272, + "step": 203 + }, + { + "epoch": 0.20955315870570107, + "grad_norm": 2.4532034397125244, + "learning_rate": 3.2191780821917813e-06, + "loss": 1.6169, + "step": 204 + }, + { + "epoch": 0.2105803800719055, + "grad_norm": 2.517059564590454, + "learning_rate": 3.236301369863014e-06, + "loss": 1.5991, + "step": 205 + }, + { + "epoch": 0.21160760143810992, + "grad_norm": 2.22658109664917, + "learning_rate": 3.253424657534247e-06, + "loss": 1.6754, + "step": 206 + }, + { + "epoch": 0.21263482280431434, + "grad_norm": 2.344799518585205, + "learning_rate": 3.2705479452054796e-06, + "loss": 1.6648, + "step": 207 + }, + { + "epoch": 0.21366204417051873, + "grad_norm": 2.331875801086426, + "learning_rate": 3.2876712328767123e-06, + "loss": 1.6112, + "step": 208 + }, + { + "epoch": 0.21468926553672316, + "grad_norm": 2.262343645095825, + "learning_rate": 3.3047945205479453e-06, + "loss": 1.6479, + "step": 209 + }, + { + "epoch": 0.21571648690292758, + "grad_norm": 2.282278537750244, + "learning_rate": 3.3219178082191784e-06, + "loss": 1.6753, + "step": 210 + }, + { + "epoch": 0.21571648690292758, + "eval_loss": 1.9338418245315552, + "eval_runtime": 37.6083, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 210 + }, + { + "epoch": 0.216743708269132, + "grad_norm": 2.578476905822754, + "learning_rate": 3.3390410958904114e-06, + "loss": 1.7043, + "step": 211 + }, + { + "epoch": 0.21777092963533642, + "grad_norm": 2.3118813037872314, + "learning_rate": 3.356164383561644e-06, + "loss": 1.7541, + "step": 212 + }, + { + "epoch": 0.21879815100154082, + "grad_norm": 2.2418692111968994, + "learning_rate": 3.373287671232877e-06, + "loss": 1.6886, + "step": 213 + }, + { + "epoch": 0.21982537236774524, + "grad_norm": 2.3583438396453857, + "learning_rate": 3.39041095890411e-06, + "loss": 1.7241, + "step": 214 + }, + { + "epoch": 0.22085259373394966, + "grad_norm": 2.345090866088867, + "learning_rate": 3.4075342465753424e-06, + "loss": 1.6194, + "step": 215 + }, + { + "epoch": 0.2218798151001541, + "grad_norm": 2.3996200561523438, + "learning_rate": 3.4246575342465754e-06, + "loss": 1.6889, + "step": 216 + }, + { + "epoch": 0.2229070364663585, + "grad_norm": 2.3552374839782715, + "learning_rate": 3.4417808219178085e-06, + "loss": 1.6713, + "step": 217 + }, + { + "epoch": 0.2239342578325629, + "grad_norm": 2.2653744220733643, + "learning_rate": 3.458904109589041e-06, + "loss": 1.673, + "step": 218 + }, + { + "epoch": 0.22496147919876733, + "grad_norm": 2.263288974761963, + "learning_rate": 3.476027397260274e-06, + "loss": 1.673, + "step": 219 + }, + { + "epoch": 0.22598870056497175, + "grad_norm": 2.3173880577087402, + "learning_rate": 3.4931506849315072e-06, + "loss": 1.6803, + "step": 220 + }, + { + "epoch": 0.22598870056497175, + "eval_loss": 1.9166476726531982, + "eval_runtime": 37.7659, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 220 + }, + { + "epoch": 0.22701592193117617, + "grad_norm": 2.4005839824676514, + "learning_rate": 3.5102739726027403e-06, + "loss": 1.6809, + "step": 221 + }, + { + "epoch": 0.2280431432973806, + "grad_norm": 2.5032525062561035, + "learning_rate": 3.527397260273973e-06, + "loss": 1.6411, + "step": 222 + }, + { + "epoch": 0.229070364663585, + "grad_norm": 2.533583879470825, + "learning_rate": 3.5445205479452056e-06, + "loss": 1.6608, + "step": 223 + }, + { + "epoch": 0.2300975860297894, + "grad_norm": 2.3575572967529297, + "learning_rate": 3.5616438356164386e-06, + "loss": 1.7236, + "step": 224 + }, + { + "epoch": 0.23112480739599384, + "grad_norm": 2.3408942222595215, + "learning_rate": 3.5787671232876712e-06, + "loss": 1.7101, + "step": 225 + }, + { + "epoch": 0.23215202876219826, + "grad_norm": 2.3819596767425537, + "learning_rate": 3.5958904109589043e-06, + "loss": 1.6414, + "step": 226 + }, + { + "epoch": 0.23317925012840268, + "grad_norm": 2.2874338626861572, + "learning_rate": 3.6130136986301373e-06, + "loss": 1.6767, + "step": 227 + }, + { + "epoch": 0.23420647149460708, + "grad_norm": 2.3523948192596436, + "learning_rate": 3.6301369863013704e-06, + "loss": 1.6932, + "step": 228 + }, + { + "epoch": 0.2352336928608115, + "grad_norm": 2.476128578186035, + "learning_rate": 3.647260273972603e-06, + "loss": 1.7125, + "step": 229 + }, + { + "epoch": 0.23626091422701592, + "grad_norm": 2.3841593265533447, + "learning_rate": 3.664383561643836e-06, + "loss": 1.6469, + "step": 230 + }, + { + "epoch": 0.23626091422701592, + "eval_loss": 1.925522804260254, + "eval_runtime": 37.8581, + "eval_samples_per_second": 26.414, + "eval_steps_per_second": 1.109, + "step": 230 + }, + { + "epoch": 0.23728813559322035, + "grad_norm": 2.4401519298553467, + "learning_rate": 3.6815068493150687e-06, + "loss": 1.6726, + "step": 231 + }, + { + "epoch": 0.23831535695942477, + "grad_norm": 2.329674005508423, + "learning_rate": 3.6986301369863014e-06, + "loss": 1.6349, + "step": 232 + }, + { + "epoch": 0.23934257832562916, + "grad_norm": 2.654290199279785, + "learning_rate": 3.7157534246575344e-06, + "loss": 1.7309, + "step": 233 + }, + { + "epoch": 0.24036979969183359, + "grad_norm": 2.1599373817443848, + "learning_rate": 3.7328767123287675e-06, + "loss": 1.6437, + "step": 234 + }, + { + "epoch": 0.241397021058038, + "grad_norm": 2.6170263290405273, + "learning_rate": 3.7500000000000005e-06, + "loss": 1.6862, + "step": 235 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 2.2836415767669678, + "learning_rate": 3.767123287671233e-06, + "loss": 1.7073, + "step": 236 + }, + { + "epoch": 0.24345146379044685, + "grad_norm": 2.3001456260681152, + "learning_rate": 3.784246575342466e-06, + "loss": 1.6455, + "step": 237 + }, + { + "epoch": 0.24447868515665125, + "grad_norm": 2.4549131393432617, + "learning_rate": 3.8013698630136993e-06, + "loss": 1.6566, + "step": 238 + }, + { + "epoch": 0.24550590652285567, + "grad_norm": 2.4387831687927246, + "learning_rate": 3.818493150684932e-06, + "loss": 1.627, + "step": 239 + }, + { + "epoch": 0.2465331278890601, + "grad_norm": 2.136474847793579, + "learning_rate": 3.8356164383561645e-06, + "loss": 1.7286, + "step": 240 + }, + { + "epoch": 0.2465331278890601, + "eval_loss": 1.9188542366027832, + "eval_runtime": 37.9079, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 240 + }, + { + "epoch": 0.24756034925526452, + "grad_norm": 2.430297374725342, + "learning_rate": 3.852739726027397e-06, + "loss": 1.7014, + "step": 241 + }, + { + "epoch": 0.24858757062146894, + "grad_norm": 2.302480697631836, + "learning_rate": 3.869863013698631e-06, + "loss": 1.6412, + "step": 242 + }, + { + "epoch": 0.24961479198767333, + "grad_norm": 2.2905426025390625, + "learning_rate": 3.886986301369863e-06, + "loss": 1.6445, + "step": 243 + }, + { + "epoch": 0.25064201335387776, + "grad_norm": 2.55385684967041, + "learning_rate": 3.904109589041096e-06, + "loss": 1.6265, + "step": 244 + }, + { + "epoch": 0.2516692347200822, + "grad_norm": 2.3053576946258545, + "learning_rate": 3.921232876712329e-06, + "loss": 1.6338, + "step": 245 + }, + { + "epoch": 0.2526964560862866, + "grad_norm": 2.4894707202911377, + "learning_rate": 3.938356164383562e-06, + "loss": 1.5966, + "step": 246 + }, + { + "epoch": 0.253723677452491, + "grad_norm": 2.6198930740356445, + "learning_rate": 3.9554794520547955e-06, + "loss": 1.6935, + "step": 247 + }, + { + "epoch": 0.25475089881869545, + "grad_norm": 2.3548998832702637, + "learning_rate": 3.972602739726027e-06, + "loss": 1.6227, + "step": 248 + }, + { + "epoch": 0.25577812018489987, + "grad_norm": 2.535460948944092, + "learning_rate": 3.989726027397261e-06, + "loss": 1.6699, + "step": 249 + }, + { + "epoch": 0.25680534155110424, + "grad_norm": 2.3680567741394043, + "learning_rate": 4.006849315068493e-06, + "loss": 1.6666, + "step": 250 + }, + { + "epoch": 0.25680534155110424, + "eval_loss": 1.9149967432022095, + "eval_runtime": 37.9423, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 250 + }, + { + "epoch": 0.25783256291730866, + "grad_norm": 2.3914358615875244, + "learning_rate": 4.023972602739726e-06, + "loss": 1.6633, + "step": 251 + }, + { + "epoch": 0.2588597842835131, + "grad_norm": 2.4293038845062256, + "learning_rate": 4.0410958904109595e-06, + "loss": 1.6573, + "step": 252 + }, + { + "epoch": 0.2598870056497175, + "grad_norm": 2.3409900665283203, + "learning_rate": 4.058219178082192e-06, + "loss": 1.6815, + "step": 253 + }, + { + "epoch": 0.26091422701592193, + "grad_norm": 2.5807456970214844, + "learning_rate": 4.075342465753426e-06, + "loss": 1.6412, + "step": 254 + }, + { + "epoch": 0.26194144838212635, + "grad_norm": 2.317861795425415, + "learning_rate": 4.092465753424658e-06, + "loss": 1.6311, + "step": 255 + }, + { + "epoch": 0.2629686697483308, + "grad_norm": 2.231682777404785, + "learning_rate": 4.109589041095891e-06, + "loss": 1.7021, + "step": 256 + }, + { + "epoch": 0.2639958911145352, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.1267123287671235e-06, + "loss": 1.6446, + "step": 257 + }, + { + "epoch": 0.2650231124807396, + "grad_norm": 2.2619071006774902, + "learning_rate": 4.143835616438356e-06, + "loss": 1.7101, + "step": 258 + }, + { + "epoch": 0.26605033384694404, + "grad_norm": 2.461945056915283, + "learning_rate": 4.16095890410959e-06, + "loss": 1.6901, + "step": 259 + }, + { + "epoch": 0.2670775552131484, + "grad_norm": 2.516348123550415, + "learning_rate": 4.178082191780822e-06, + "loss": 1.644, + "step": 260 + }, + { + "epoch": 0.2670775552131484, + "eval_loss": 1.904897928237915, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 260 + }, + { + "epoch": 0.26810477657935283, + "grad_norm": 2.2311813831329346, + "learning_rate": 4.195205479452055e-06, + "loss": 1.6301, + "step": 261 + }, + { + "epoch": 0.26913199794555726, + "grad_norm": 2.605109214782715, + "learning_rate": 4.212328767123288e-06, + "loss": 1.6393, + "step": 262 + }, + { + "epoch": 0.2701592193117617, + "grad_norm": 2.6103134155273438, + "learning_rate": 4.229452054794521e-06, + "loss": 1.6468, + "step": 263 + }, + { + "epoch": 0.2711864406779661, + "grad_norm": 2.5538523197174072, + "learning_rate": 4.246575342465754e-06, + "loss": 1.6294, + "step": 264 + }, + { + "epoch": 0.2722136620441705, + "grad_norm": 2.5841965675354004, + "learning_rate": 4.263698630136986e-06, + "loss": 1.5837, + "step": 265 + }, + { + "epoch": 0.27324088341037495, + "grad_norm": 2.4604523181915283, + "learning_rate": 4.28082191780822e-06, + "loss": 1.7473, + "step": 266 + }, + { + "epoch": 0.27426810477657937, + "grad_norm": 2.6701765060424805, + "learning_rate": 4.297945205479452e-06, + "loss": 1.6539, + "step": 267 + }, + { + "epoch": 0.2752953261427838, + "grad_norm": 2.360069990158081, + "learning_rate": 4.315068493150685e-06, + "loss": 1.6385, + "step": 268 + }, + { + "epoch": 0.2763225475089882, + "grad_norm": 2.4922797679901123, + "learning_rate": 4.3321917808219185e-06, + "loss": 1.6586, + "step": 269 + }, + { + "epoch": 0.2773497688751926, + "grad_norm": 2.5443379878997803, + "learning_rate": 4.349315068493151e-06, + "loss": 1.6318, + "step": 270 + }, + { + "epoch": 0.2773497688751926, + "eval_loss": 1.9062609672546387, + "eval_runtime": 37.92, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 270 + }, + { + "epoch": 0.278376990241397, + "grad_norm": 2.4064481258392334, + "learning_rate": 4.3664383561643846e-06, + "loss": 1.6178, + "step": 271 + }, + { + "epoch": 0.2794042116076014, + "grad_norm": 2.681637763977051, + "learning_rate": 4.383561643835616e-06, + "loss": 1.7071, + "step": 272 + }, + { + "epoch": 0.28043143297380585, + "grad_norm": 2.5805184841156006, + "learning_rate": 4.40068493150685e-06, + "loss": 1.6553, + "step": 273 + }, + { + "epoch": 0.2814586543400103, + "grad_norm": 2.4169883728027344, + "learning_rate": 4.4178082191780825e-06, + "loss": 1.6249, + "step": 274 + }, + { + "epoch": 0.2824858757062147, + "grad_norm": 2.557570457458496, + "learning_rate": 4.434931506849315e-06, + "loss": 1.5932, + "step": 275 + }, + { + "epoch": 0.2835130970724191, + "grad_norm": 2.3808107376098633, + "learning_rate": 4.4520547945205486e-06, + "loss": 1.6377, + "step": 276 + }, + { + "epoch": 0.28454031843862354, + "grad_norm": 2.402404546737671, + "learning_rate": 4.469178082191781e-06, + "loss": 1.6285, + "step": 277 + }, + { + "epoch": 0.28556753980482796, + "grad_norm": 2.4676456451416016, + "learning_rate": 4.486301369863014e-06, + "loss": 1.6838, + "step": 278 + }, + { + "epoch": 0.28659476117103233, + "grad_norm": 2.659083366394043, + "learning_rate": 4.503424657534247e-06, + "loss": 1.5872, + "step": 279 + }, + { + "epoch": 0.28762198253723675, + "grad_norm": 2.326733350753784, + "learning_rate": 4.52054794520548e-06, + "loss": 1.6142, + "step": 280 + }, + { + "epoch": 0.28762198253723675, + "eval_loss": 1.904759407043457, + "eval_runtime": 37.9651, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 280 + }, + { + "epoch": 0.2886492039034412, + "grad_norm": 2.299504518508911, + "learning_rate": 4.537671232876713e-06, + "loss": 1.6673, + "step": 281 + }, + { + "epoch": 0.2896764252696456, + "grad_norm": 2.4932825565338135, + "learning_rate": 4.554794520547945e-06, + "loss": 1.6586, + "step": 282 + }, + { + "epoch": 0.29070364663585, + "grad_norm": 2.2599050998687744, + "learning_rate": 4.571917808219179e-06, + "loss": 1.6568, + "step": 283 + }, + { + "epoch": 0.29173086800205444, + "grad_norm": 2.435098886489868, + "learning_rate": 4.589041095890411e-06, + "loss": 1.6113, + "step": 284 + }, + { + "epoch": 0.29275808936825887, + "grad_norm": 2.539246082305908, + "learning_rate": 4.606164383561644e-06, + "loss": 1.6306, + "step": 285 + }, + { + "epoch": 0.2937853107344633, + "grad_norm": 2.596198320388794, + "learning_rate": 4.6232876712328774e-06, + "loss": 1.6473, + "step": 286 + }, + { + "epoch": 0.2948125321006677, + "grad_norm": 2.349133014678955, + "learning_rate": 4.64041095890411e-06, + "loss": 1.6042, + "step": 287 + }, + { + "epoch": 0.29583975346687214, + "grad_norm": 2.3572661876678467, + "learning_rate": 4.657534246575343e-06, + "loss": 1.6635, + "step": 288 + }, + { + "epoch": 0.2968669748330765, + "grad_norm": 2.4210307598114014, + "learning_rate": 4.674657534246575e-06, + "loss": 1.624, + "step": 289 + }, + { + "epoch": 0.2978941961992809, + "grad_norm": 2.427403688430786, + "learning_rate": 4.691780821917809e-06, + "loss": 1.6824, + "step": 290 + }, + { + "epoch": 0.2978941961992809, + "eval_loss": 1.8888994455337524, + "eval_runtime": 37.9777, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 290 + }, + { + "epoch": 0.29892141756548535, + "grad_norm": 2.534029722213745, + "learning_rate": 4.7089041095890414e-06, + "loss": 1.6456, + "step": 291 + }, + { + "epoch": 0.29994863893168977, + "grad_norm": 2.329305648803711, + "learning_rate": 4.726027397260274e-06, + "loss": 1.646, + "step": 292 + }, + { + "epoch": 0.3009758602978942, + "grad_norm": 2.561548948287964, + "learning_rate": 4.7431506849315075e-06, + "loss": 1.6511, + "step": 293 + }, + { + "epoch": 0.3020030816640986, + "grad_norm": 2.707712411880493, + "learning_rate": 4.76027397260274e-06, + "loss": 1.5984, + "step": 294 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 2.403089761734009, + "learning_rate": 4.777397260273973e-06, + "loss": 1.6164, + "step": 295 + }, + { + "epoch": 0.30405752439650746, + "grad_norm": 2.424910306930542, + "learning_rate": 4.7945205479452054e-06, + "loss": 1.6291, + "step": 296 + }, + { + "epoch": 0.3050847457627119, + "grad_norm": 2.3474302291870117, + "learning_rate": 4.811643835616439e-06, + "loss": 1.553, + "step": 297 + }, + { + "epoch": 0.3061119671289163, + "grad_norm": 2.2211227416992188, + "learning_rate": 4.8287671232876716e-06, + "loss": 1.6108, + "step": 298 + }, + { + "epoch": 0.3071391884951207, + "grad_norm": 2.495983600616455, + "learning_rate": 4.845890410958904e-06, + "loss": 1.6135, + "step": 299 + }, + { + "epoch": 0.3081664098613251, + "grad_norm": 2.525495767593384, + "learning_rate": 4.863013698630138e-06, + "loss": 1.6061, + "step": 300 + }, + { + "epoch": 0.3081664098613251, + "eval_loss": 1.8982168436050415, + "eval_runtime": 37.9848, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 300 + }, + { + "epoch": 0.3091936312275295, + "grad_norm": 2.449064254760742, + "learning_rate": 4.88013698630137e-06, + "loss": 1.5827, + "step": 301 + }, + { + "epoch": 0.31022085259373394, + "grad_norm": 2.288095712661743, + "learning_rate": 4.897260273972603e-06, + "loss": 1.704, + "step": 302 + }, + { + "epoch": 0.31124807395993837, + "grad_norm": 2.5301458835601807, + "learning_rate": 4.914383561643836e-06, + "loss": 1.6342, + "step": 303 + }, + { + "epoch": 0.3122752953261428, + "grad_norm": 2.2793986797332764, + "learning_rate": 4.931506849315069e-06, + "loss": 1.6688, + "step": 304 + }, + { + "epoch": 0.3133025166923472, + "grad_norm": 2.38783597946167, + "learning_rate": 4.948630136986302e-06, + "loss": 1.7014, + "step": 305 + }, + { + "epoch": 0.31432973805855163, + "grad_norm": 2.5224838256835938, + "learning_rate": 4.965753424657534e-06, + "loss": 1.5812, + "step": 306 + }, + { + "epoch": 0.31535695942475606, + "grad_norm": 2.434432029724121, + "learning_rate": 4.982876712328768e-06, + "loss": 1.6186, + "step": 307 + }, + { + "epoch": 0.3163841807909605, + "grad_norm": 2.3708460330963135, + "learning_rate": 5e-06, + "loss": 1.5669, + "step": 308 + }, + { + "epoch": 0.31741140215716485, + "grad_norm": 2.5524990558624268, + "learning_rate": 4.998096688237534e-06, + "loss": 1.6688, + "step": 309 + }, + { + "epoch": 0.31843862352336927, + "grad_norm": 2.322441577911377, + "learning_rate": 4.996193376475067e-06, + "loss": 1.5571, + "step": 310 + }, + { + "epoch": 0.31843862352336927, + "eval_loss": 1.8914265632629395, + "eval_runtime": 37.6505, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 310 + }, + { + "epoch": 0.3194658448895737, + "grad_norm": 2.5248687267303467, + "learning_rate": 4.9942900647126005e-06, + "loss": 1.5585, + "step": 311 + }, + { + "epoch": 0.3204930662557781, + "grad_norm": 2.2464914321899414, + "learning_rate": 4.992386752950134e-06, + "loss": 1.5921, + "step": 312 + }, + { + "epoch": 0.32152028762198254, + "grad_norm": 2.5355331897735596, + "learning_rate": 4.990483441187667e-06, + "loss": 1.6263, + "step": 313 + }, + { + "epoch": 0.32254750898818696, + "grad_norm": 2.5611143112182617, + "learning_rate": 4.9885801294252e-06, + "loss": 1.6267, + "step": 314 + }, + { + "epoch": 0.3235747303543914, + "grad_norm": 2.364356517791748, + "learning_rate": 4.986676817662733e-06, + "loss": 1.5847, + "step": 315 + }, + { + "epoch": 0.3246019517205958, + "grad_norm": 2.455165147781372, + "learning_rate": 4.9847735059002665e-06, + "loss": 1.661, + "step": 316 + }, + { + "epoch": 0.3256291730868002, + "grad_norm": 2.4312195777893066, + "learning_rate": 4.9828701941378e-06, + "loss": 1.5995, + "step": 317 + }, + { + "epoch": 0.3266563944530046, + "grad_norm": 2.4081499576568604, + "learning_rate": 4.980966882375333e-06, + "loss": 1.5756, + "step": 318 + }, + { + "epoch": 0.327683615819209, + "grad_norm": 2.2608511447906494, + "learning_rate": 4.979063570612867e-06, + "loss": 1.543, + "step": 319 + }, + { + "epoch": 0.32871083718541344, + "grad_norm": 2.6551144123077393, + "learning_rate": 4.9771602588504e-06, + "loss": 1.6467, + "step": 320 + }, + { + "epoch": 0.32871083718541344, + "eval_loss": 1.89058518409729, + "eval_runtime": 37.8192, + "eval_samples_per_second": 26.442, + "eval_steps_per_second": 1.111, + "step": 320 + }, + { + "epoch": 0.32973805855161786, + "grad_norm": 2.3897273540496826, + "learning_rate": 4.975256947087933e-06, + "loss": 1.5607, + "step": 321 + }, + { + "epoch": 0.3307652799178223, + "grad_norm": 2.3869469165802, + "learning_rate": 4.973353635325467e-06, + "loss": 1.6079, + "step": 322 + }, + { + "epoch": 0.3317925012840267, + "grad_norm": 2.159193515777588, + "learning_rate": 4.971450323563e-06, + "loss": 1.5475, + "step": 323 + }, + { + "epoch": 0.33281972265023113, + "grad_norm": 2.3069653511047363, + "learning_rate": 4.969547011800533e-06, + "loss": 1.616, + "step": 324 + }, + { + "epoch": 0.33384694401643555, + "grad_norm": 2.3131422996520996, + "learning_rate": 4.967643700038067e-06, + "loss": 1.6106, + "step": 325 + }, + { + "epoch": 0.33487416538264, + "grad_norm": 2.4398200511932373, + "learning_rate": 4.9657403882756e-06, + "loss": 1.5779, + "step": 326 + }, + { + "epoch": 0.3359013867488444, + "grad_norm": 2.278024435043335, + "learning_rate": 4.9638370765131335e-06, + "loss": 1.5724, + "step": 327 + }, + { + "epoch": 0.33692860811504877, + "grad_norm": 2.4989173412323, + "learning_rate": 4.961933764750667e-06, + "loss": 1.6337, + "step": 328 + }, + { + "epoch": 0.3379558294812532, + "grad_norm": 2.46319580078125, + "learning_rate": 4.9600304529882e-06, + "loss": 1.6519, + "step": 329 + }, + { + "epoch": 0.3389830508474576, + "grad_norm": 2.2703049182891846, + "learning_rate": 4.9581271412257335e-06, + "loss": 1.6081, + "step": 330 + }, + { + "epoch": 0.3389830508474576, + "eval_loss": 1.8755545616149902, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 330 + }, + { + "epoch": 0.34001027221366203, + "grad_norm": 2.502196788787842, + "learning_rate": 4.956223829463266e-06, + "loss": 1.5221, + "step": 331 + }, + { + "epoch": 0.34103749357986646, + "grad_norm": 2.284280776977539, + "learning_rate": 4.954320517700799e-06, + "loss": 1.6268, + "step": 332 + }, + { + "epoch": 0.3420647149460709, + "grad_norm": 2.3551881313323975, + "learning_rate": 4.952417205938333e-06, + "loss": 1.6078, + "step": 333 + }, + { + "epoch": 0.3430919363122753, + "grad_norm": 2.343780755996704, + "learning_rate": 4.950513894175866e-06, + "loss": 1.591, + "step": 334 + }, + { + "epoch": 0.3441191576784797, + "grad_norm": 2.3524727821350098, + "learning_rate": 4.9486105824133995e-06, + "loss": 1.605, + "step": 335 + }, + { + "epoch": 0.34514637904468415, + "grad_norm": 2.349896192550659, + "learning_rate": 4.946707270650933e-06, + "loss": 1.6106, + "step": 336 + }, + { + "epoch": 0.34617360041088857, + "grad_norm": 2.5154407024383545, + "learning_rate": 4.944803958888466e-06, + "loss": 1.6429, + "step": 337 + }, + { + "epoch": 0.34720082177709294, + "grad_norm": 2.2692947387695312, + "learning_rate": 4.942900647126e-06, + "loss": 1.5865, + "step": 338 + }, + { + "epoch": 0.34822804314329736, + "grad_norm": 2.2672805786132812, + "learning_rate": 4.940997335363533e-06, + "loss": 1.5464, + "step": 339 + }, + { + "epoch": 0.3492552645095018, + "grad_norm": 2.5147039890289307, + "learning_rate": 4.939094023601066e-06, + "loss": 1.5837, + "step": 340 + }, + { + "epoch": 0.3492552645095018, + "eval_loss": 1.8672455549240112, + "eval_runtime": 37.9644, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 340 + }, + { + "epoch": 0.3502824858757062, + "grad_norm": 2.3416433334350586, + "learning_rate": 4.9371907118386e-06, + "loss": 1.5937, + "step": 341 + }, + { + "epoch": 0.35130970724191063, + "grad_norm": 2.3044960498809814, + "learning_rate": 4.935287400076133e-06, + "loss": 1.6132, + "step": 342 + }, + { + "epoch": 0.35233692860811505, + "grad_norm": 2.4360430240631104, + "learning_rate": 4.933384088313666e-06, + "loss": 1.5731, + "step": 343 + }, + { + "epoch": 0.3533641499743195, + "grad_norm": 2.392050266265869, + "learning_rate": 4.9314807765512e-06, + "loss": 1.5098, + "step": 344 + }, + { + "epoch": 0.3543913713405239, + "grad_norm": 2.3815808296203613, + "learning_rate": 4.929577464788733e-06, + "loss": 1.5827, + "step": 345 + }, + { + "epoch": 0.3554185927067283, + "grad_norm": 2.423077344894409, + "learning_rate": 4.9276741530262665e-06, + "loss": 1.5857, + "step": 346 + }, + { + "epoch": 0.35644581407293274, + "grad_norm": 2.5535898208618164, + "learning_rate": 4.9257708412638e-06, + "loss": 1.5884, + "step": 347 + }, + { + "epoch": 0.3574730354391371, + "grad_norm": 2.2665445804595947, + "learning_rate": 4.923867529501332e-06, + "loss": 1.5561, + "step": 348 + }, + { + "epoch": 0.35850025680534153, + "grad_norm": 2.3357481956481934, + "learning_rate": 4.921964217738866e-06, + "loss": 1.6444, + "step": 349 + }, + { + "epoch": 0.35952747817154596, + "grad_norm": 2.4537885189056396, + "learning_rate": 4.920060905976399e-06, + "loss": 1.6262, + "step": 350 + }, + { + "epoch": 0.35952747817154596, + "eval_loss": 1.8621584177017212, + "eval_runtime": 38.0027, + "eval_samples_per_second": 26.314, + "eval_steps_per_second": 1.105, + "step": 350 + }, + { + "epoch": 0.3605546995377504, + "grad_norm": 2.379754066467285, + "learning_rate": 4.9181575942139325e-06, + "loss": 1.6227, + "step": 351 + }, + { + "epoch": 0.3615819209039548, + "grad_norm": 2.432905435562134, + "learning_rate": 4.916254282451466e-06, + "loss": 1.6554, + "step": 352 + }, + { + "epoch": 0.3626091422701592, + "grad_norm": 2.34236216545105, + "learning_rate": 4.914350970688999e-06, + "loss": 1.5613, + "step": 353 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.3019444942474365, + "learning_rate": 4.9124476589265325e-06, + "loss": 1.5535, + "step": 354 + }, + { + "epoch": 0.36466358500256807, + "grad_norm": 2.3354644775390625, + "learning_rate": 4.910544347164066e-06, + "loss": 1.6384, + "step": 355 + }, + { + "epoch": 0.3656908063687725, + "grad_norm": 2.5154788494110107, + "learning_rate": 4.908641035401599e-06, + "loss": 1.6986, + "step": 356 + }, + { + "epoch": 0.3667180277349769, + "grad_norm": 2.2707178592681885, + "learning_rate": 4.906737723639133e-06, + "loss": 1.5705, + "step": 357 + }, + { + "epoch": 0.3677452491011813, + "grad_norm": 2.3058297634124756, + "learning_rate": 4.904834411876665e-06, + "loss": 1.6033, + "step": 358 + }, + { + "epoch": 0.3687724704673857, + "grad_norm": 2.3923370838165283, + "learning_rate": 4.902931100114199e-06, + "loss": 1.5333, + "step": 359 + }, + { + "epoch": 0.3697996918335901, + "grad_norm": 2.4023196697235107, + "learning_rate": 4.901027788351733e-06, + "loss": 1.5775, + "step": 360 + }, + { + "epoch": 0.3697996918335901, + "eval_loss": 1.867856740951538, + "eval_runtime": 38.0125, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 360 + }, + { + "epoch": 0.37082691319979455, + "grad_norm": 2.4903581142425537, + "learning_rate": 4.899124476589266e-06, + "loss": 1.5491, + "step": 361 + }, + { + "epoch": 0.371854134565999, + "grad_norm": 2.556683301925659, + "learning_rate": 4.8972211648267994e-06, + "loss": 1.5967, + "step": 362 + }, + { + "epoch": 0.3728813559322034, + "grad_norm": 2.3017377853393555, + "learning_rate": 4.895317853064333e-06, + "loss": 1.5661, + "step": 363 + }, + { + "epoch": 0.3739085772984078, + "grad_norm": 2.286191940307617, + "learning_rate": 4.893414541301866e-06, + "loss": 1.5679, + "step": 364 + }, + { + "epoch": 0.37493579866461224, + "grad_norm": 2.260812520980835, + "learning_rate": 4.891511229539399e-06, + "loss": 1.5965, + "step": 365 + }, + { + "epoch": 0.37596302003081666, + "grad_norm": 2.4569408893585205, + "learning_rate": 4.889607917776932e-06, + "loss": 1.5923, + "step": 366 + }, + { + "epoch": 0.37699024139702103, + "grad_norm": 2.571552038192749, + "learning_rate": 4.887704606014465e-06, + "loss": 1.6021, + "step": 367 + }, + { + "epoch": 0.37801746276322545, + "grad_norm": 2.287754774093628, + "learning_rate": 4.885801294251999e-06, + "loss": 1.5304, + "step": 368 + }, + { + "epoch": 0.3790446841294299, + "grad_norm": 2.317784070968628, + "learning_rate": 4.883897982489532e-06, + "loss": 1.6233, + "step": 369 + }, + { + "epoch": 0.3800719054956343, + "grad_norm": 2.290601968765259, + "learning_rate": 4.8819946707270655e-06, + "loss": 1.624, + "step": 370 + }, + { + "epoch": 0.3800719054956343, + "eval_loss": 1.8604528903961182, + "eval_runtime": 38.0176, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 370 + }, + { + "epoch": 0.3810991268618387, + "grad_norm": 2.28519868850708, + "learning_rate": 4.880091358964599e-06, + "loss": 1.5286, + "step": 371 + }, + { + "epoch": 0.38212634822804314, + "grad_norm": 2.271810293197632, + "learning_rate": 4.878188047202132e-06, + "loss": 1.6879, + "step": 372 + }, + { + "epoch": 0.38315356959424757, + "grad_norm": 2.391848564147949, + "learning_rate": 4.876284735439666e-06, + "loss": 1.6458, + "step": 373 + }, + { + "epoch": 0.384180790960452, + "grad_norm": 2.3787031173706055, + "learning_rate": 4.874381423677199e-06, + "loss": 1.5738, + "step": 374 + }, + { + "epoch": 0.3852080123266564, + "grad_norm": 2.404219150543213, + "learning_rate": 4.8724781119147315e-06, + "loss": 1.6184, + "step": 375 + }, + { + "epoch": 0.38623523369286084, + "grad_norm": 2.372661590576172, + "learning_rate": 4.870574800152265e-06, + "loss": 1.5316, + "step": 376 + }, + { + "epoch": 0.3872624550590652, + "grad_norm": 2.4397103786468506, + "learning_rate": 4.868671488389798e-06, + "loss": 1.5605, + "step": 377 + }, + { + "epoch": 0.3882896764252696, + "grad_norm": 2.371006727218628, + "learning_rate": 4.8667681766273316e-06, + "loss": 1.5184, + "step": 378 + }, + { + "epoch": 0.38931689779147405, + "grad_norm": 2.2621910572052, + "learning_rate": 4.864864864864866e-06, + "loss": 1.5788, + "step": 379 + }, + { + "epoch": 0.39034411915767847, + "grad_norm": 2.2888615131378174, + "learning_rate": 4.862961553102399e-06, + "loss": 1.6244, + "step": 380 + }, + { + "epoch": 0.39034411915767847, + "eval_loss": 1.8586912155151367, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 380 + }, + { + "epoch": 0.3913713405238829, + "grad_norm": 2.4439220428466797, + "learning_rate": 4.8610582413399325e-06, + "loss": 1.5749, + "step": 381 + }, + { + "epoch": 0.3923985618900873, + "grad_norm": 2.2423763275146484, + "learning_rate": 4.859154929577465e-06, + "loss": 1.5738, + "step": 382 + }, + { + "epoch": 0.39342578325629174, + "grad_norm": 2.267467737197876, + "learning_rate": 4.857251617814998e-06, + "loss": 1.6212, + "step": 383 + }, + { + "epoch": 0.39445300462249616, + "grad_norm": 2.3680620193481445, + "learning_rate": 4.855348306052532e-06, + "loss": 1.6129, + "step": 384 + }, + { + "epoch": 0.3954802259887006, + "grad_norm": 2.3782172203063965, + "learning_rate": 4.853444994290065e-06, + "loss": 1.5605, + "step": 385 + }, + { + "epoch": 0.396507447354905, + "grad_norm": 2.5745720863342285, + "learning_rate": 4.8515416825275985e-06, + "loss": 1.634, + "step": 386 + }, + { + "epoch": 0.3975346687211094, + "grad_norm": 2.38212513923645, + "learning_rate": 4.849638370765132e-06, + "loss": 1.5671, + "step": 387 + }, + { + "epoch": 0.3985618900873138, + "grad_norm": 2.3395450115203857, + "learning_rate": 4.847735059002665e-06, + "loss": 1.5088, + "step": 388 + }, + { + "epoch": 0.3995891114535182, + "grad_norm": 2.474440574645996, + "learning_rate": 4.8458317472401985e-06, + "loss": 1.5997, + "step": 389 + }, + { + "epoch": 0.40061633281972264, + "grad_norm": 2.429133176803589, + "learning_rate": 4.843928435477732e-06, + "loss": 1.5611, + "step": 390 + }, + { + "epoch": 0.40061633281972264, + "eval_loss": 1.8613603115081787, + "eval_runtime": 37.9862, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 390 + }, + { + "epoch": 0.40164355418592707, + "grad_norm": 2.403207778930664, + "learning_rate": 4.842025123715265e-06, + "loss": 1.5976, + "step": 391 + }, + { + "epoch": 0.4026707755521315, + "grad_norm": 2.383018732070923, + "learning_rate": 4.840121811952798e-06, + "loss": 1.5791, + "step": 392 + }, + { + "epoch": 0.4036979969183359, + "grad_norm": 2.5303306579589844, + "learning_rate": 4.838218500190331e-06, + "loss": 1.6035, + "step": 393 + }, + { + "epoch": 0.40472521828454033, + "grad_norm": 2.391075372695923, + "learning_rate": 4.8363151884278645e-06, + "loss": 1.5795, + "step": 394 + }, + { + "epoch": 0.40575243965074476, + "grad_norm": 2.396934747695923, + "learning_rate": 4.834411876665398e-06, + "loss": 1.5965, + "step": 395 + }, + { + "epoch": 0.4067796610169492, + "grad_norm": 2.4631423950195312, + "learning_rate": 4.832508564902931e-06, + "loss": 1.5372, + "step": 396 + }, + { + "epoch": 0.40780688238315355, + "grad_norm": 2.3457326889038086, + "learning_rate": 4.830605253140465e-06, + "loss": 1.5729, + "step": 397 + }, + { + "epoch": 0.40883410374935797, + "grad_norm": 2.3763277530670166, + "learning_rate": 4.828701941377998e-06, + "loss": 1.5309, + "step": 398 + }, + { + "epoch": 0.4098613251155624, + "grad_norm": 2.442439556121826, + "learning_rate": 4.826798629615531e-06, + "loss": 1.5465, + "step": 399 + }, + { + "epoch": 0.4108885464817668, + "grad_norm": 2.3464581966400146, + "learning_rate": 4.824895317853065e-06, + "loss": 1.5665, + "step": 400 + }, + { + "epoch": 0.4108885464817668, + "eval_loss": 1.8671027421951294, + "eval_runtime": 38.0166, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 400 + } + ], + "logging_steps": 1, + "max_steps": 2919, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.246923763744768e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/zero_to_fp32.py b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-400/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..d5f37008b65b2b6b774ba8ca3e08ea7f2886beeb --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:898e9b2b9cf458e184c0b3debd303a4fd7c10dad3b82242d58682e89cace8b1e +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..43698f33ac8bdb91266f12645eda8b36787da828 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:050132a910188198ac7405dcc27a0ec731bdadfd7fd9c07e7468ac838c90fae0 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..90d0b3f139f4cb1368a680fad6f12df4c66f078f --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:21ab24991874894fe74f6d6882102cbe2b931a423ab076862a8f74e33fa92f52 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..b05caf0f4e74443a7d480d4070615caf539adc44 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:021e4082dc4c1b24f078b79e27d8c4147d8c515f01a9535cc62d7153d4cc609f +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-600/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/config.json @@ -0,0 +1,36 @@ +{ + "_name_or_path": "meta-llama/Llama-3.2-3B", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "eos_token_id": 128001, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 3072, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 24, + "num_hidden_layers": 28, + "num_key_value_heads": 8, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_type": "llama3" + }, + "rope_theta": 500000.0, + "tie_word_embeddings": true, + "torch_dtype": "float16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 128256 +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/generation_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": 128001, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "4.45.1" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/latest b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/latest new file mode 100644 index 0000000000000000000000000000000000000000..57729c0be88118cbd582c8c68b4149cee821f0b4 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/latest @@ -0,0 +1 @@ +global_step800 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/model.safetensors.index.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/model.safetensors.index.json @@ -0,0 +1,262 @@ +{ + "metadata": { + "total_size": 7213504512 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_0.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..10b2f5e4f6c9f32023941f5b8cf39645a86289fd --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ccffad8722b9db65bab96085aca5b7293a1ef8d534c7b8817b948837c87f6b8 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_1.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..1f46eec58ec6affe22de910a8d659fe43a6018f4 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:633d50e430585944b70d0f16b52abb22f10fe2d185c48b1ba52f7c6a7004014f +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_2.pth b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..89cef672a6b7edc3804190d221d1a77cd6a2d146 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8c200394d27cbd1e2e417e51ec9d687fba67a2bda4d806d7a86762d3cf2c3f77 +size 16567 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/scheduler.pt b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..45845e1ff1252d20c17334b9502ee01c37ca66fe --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:02b3b35f168f82d92ace614b05cfe0c776f90e2d4e7f5defe9bf74f88fa89802 +size 563 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/special_tokens_map.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/special_tokens_map.json @@ -0,0 +1,23 @@ +{ + "bos_token": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/tokenizer.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..c76f0701115b50c463b66cc3645d7a3654ab5205 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fda2ecf7b5daf3cb77324ea6c5b9cb8bf47119249a6d756ef2dd4d62540ac6f +size 17210554 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/tokenizer_config.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0b782d4602b361f7577abb52b710a8c0e088588d --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/tokenizer_config.json @@ -0,0 +1,2078 @@ +{ + "added_tokens_decoder": { + "128000": { + "content": "<|begin_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128001": { + "content": "<|end_of_text|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128002": { + "content": "<|reserved_special_token_0|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128003": { + "content": "<|reserved_special_token_1|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128004": { + "content": "<|finetune_right_pad_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128005": { + "content": "<|reserved_special_token_2|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128006": { + "content": "<|start_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128007": { + "content": "<|end_header_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128008": { + "content": "<|eom_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128009": { + "content": "<|eot_id|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128010": { + "content": "<|python_tag|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128011": { + "content": "<|reserved_special_token_3|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128012": { + "content": "<|reserved_special_token_4|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128013": { + "content": "<|reserved_special_token_5|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128014": { + "content": "<|reserved_special_token_6|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128015": { + "content": "<|reserved_special_token_7|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128016": { + "content": "<|reserved_special_token_8|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128017": { + "content": "<|reserved_special_token_9|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128018": { + "content": "<|reserved_special_token_10|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128019": { + "content": "<|reserved_special_token_11|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128020": { + "content": "<|reserved_special_token_12|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128021": { + "content": "<|reserved_special_token_13|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128022": { + "content": "<|reserved_special_token_14|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128023": { + "content": "<|reserved_special_token_15|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128024": { + "content": "<|reserved_special_token_16|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128025": { + "content": "<|reserved_special_token_17|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128026": { + "content": "<|reserved_special_token_18|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128027": { + "content": "<|reserved_special_token_19|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128028": { + "content": "<|reserved_special_token_20|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128029": { + "content": "<|reserved_special_token_21|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128030": { + "content": "<|reserved_special_token_22|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128031": { + "content": "<|reserved_special_token_23|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128032": { + "content": "<|reserved_special_token_24|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128033": { + "content": "<|reserved_special_token_25|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128034": { + "content": "<|reserved_special_token_26|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128035": { + "content": "<|reserved_special_token_27|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128036": { + "content": "<|reserved_special_token_28|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128037": { + "content": "<|reserved_special_token_29|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128038": { + "content": "<|reserved_special_token_30|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128039": { + "content": "<|reserved_special_token_31|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128040": { + "content": "<|reserved_special_token_32|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128041": { + "content": "<|reserved_special_token_33|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128042": { + "content": "<|reserved_special_token_34|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128043": { + "content": "<|reserved_special_token_35|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128044": { + "content": "<|reserved_special_token_36|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128045": { + "content": "<|reserved_special_token_37|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128046": { + "content": "<|reserved_special_token_38|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128047": { + "content": "<|reserved_special_token_39|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128048": { + "content": "<|reserved_special_token_40|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128049": { + "content": "<|reserved_special_token_41|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128050": { + "content": "<|reserved_special_token_42|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128051": { + "content": "<|reserved_special_token_43|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128052": { + "content": "<|reserved_special_token_44|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128053": { + "content": "<|reserved_special_token_45|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128054": { + "content": "<|reserved_special_token_46|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128055": { + "content": "<|reserved_special_token_47|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128056": { + "content": "<|reserved_special_token_48|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128057": { + "content": "<|reserved_special_token_49|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128058": { + "content": "<|reserved_special_token_50|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128059": { + "content": "<|reserved_special_token_51|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128060": { + "content": "<|reserved_special_token_52|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128061": { + "content": "<|reserved_special_token_53|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128062": { + "content": "<|reserved_special_token_54|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128063": { + "content": "<|reserved_special_token_55|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128064": { + "content": "<|reserved_special_token_56|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128065": { + "content": "<|reserved_special_token_57|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128066": { + "content": "<|reserved_special_token_58|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128067": { + "content": "<|reserved_special_token_59|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128068": { + "content": "<|reserved_special_token_60|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128069": { + "content": "<|reserved_special_token_61|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128070": { + "content": "<|reserved_special_token_62|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128071": { + "content": "<|reserved_special_token_63|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128072": { + "content": "<|reserved_special_token_64|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128073": { + "content": "<|reserved_special_token_65|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128074": { + "content": "<|reserved_special_token_66|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128075": { + "content": "<|reserved_special_token_67|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128076": { + "content": "<|reserved_special_token_68|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128077": { + "content": "<|reserved_special_token_69|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128078": { + "content": "<|reserved_special_token_70|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128079": { + "content": "<|reserved_special_token_71|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128080": { + "content": "<|reserved_special_token_72|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128081": { + "content": "<|reserved_special_token_73|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128082": { + "content": "<|reserved_special_token_74|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128083": { + "content": "<|reserved_special_token_75|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128084": { + "content": "<|reserved_special_token_76|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128085": { + "content": "<|reserved_special_token_77|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128086": { + "content": "<|reserved_special_token_78|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128087": { + "content": "<|reserved_special_token_79|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128088": { + "content": "<|reserved_special_token_80|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128089": { + "content": "<|reserved_special_token_81|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128090": { + "content": "<|reserved_special_token_82|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128091": { + "content": "<|reserved_special_token_83|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128092": { + "content": "<|reserved_special_token_84|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128093": { + "content": "<|reserved_special_token_85|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128094": { + "content": "<|reserved_special_token_86|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128095": { + "content": "<|reserved_special_token_87|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128096": { + "content": "<|reserved_special_token_88|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128097": { + "content": "<|reserved_special_token_89|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128098": { + "content": "<|reserved_special_token_90|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128099": { + "content": "<|reserved_special_token_91|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128100": { + "content": "<|reserved_special_token_92|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128101": { + "content": "<|reserved_special_token_93|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128102": { + "content": "<|reserved_special_token_94|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128103": { + "content": "<|reserved_special_token_95|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128104": { + "content": "<|reserved_special_token_96|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128105": { + "content": "<|reserved_special_token_97|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128106": { + "content": "<|reserved_special_token_98|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128107": { + "content": "<|reserved_special_token_99|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128108": { + "content": "<|reserved_special_token_100|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128109": { + "content": "<|reserved_special_token_101|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128110": { + "content": "<|reserved_special_token_102|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128111": { + "content": "<|reserved_special_token_103|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128112": { + "content": "<|reserved_special_token_104|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128113": { + "content": "<|reserved_special_token_105|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128114": { + "content": "<|reserved_special_token_106|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128115": { + "content": "<|reserved_special_token_107|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128116": { + "content": "<|reserved_special_token_108|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128117": { + "content": "<|reserved_special_token_109|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128118": { + "content": "<|reserved_special_token_110|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128119": { + "content": "<|reserved_special_token_111|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128120": { + "content": "<|reserved_special_token_112|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128121": { + "content": "<|reserved_special_token_113|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128122": { + "content": "<|reserved_special_token_114|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128123": { + "content": "<|reserved_special_token_115|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128124": { + "content": "<|reserved_special_token_116|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128125": { + "content": "<|reserved_special_token_117|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128126": { + "content": "<|reserved_special_token_118|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128127": { + "content": "<|reserved_special_token_119|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128128": { + "content": "<|reserved_special_token_120|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128129": { + "content": "<|reserved_special_token_121|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128130": { + "content": "<|reserved_special_token_122|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128131": { + "content": "<|reserved_special_token_123|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128132": { + "content": "<|reserved_special_token_124|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128133": { + "content": "<|reserved_special_token_125|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128134": { + "content": "<|reserved_special_token_126|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128135": { + "content": "<|reserved_special_token_127|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128136": { + "content": "<|reserved_special_token_128|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128137": { + "content": "<|reserved_special_token_129|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128138": { + "content": "<|reserved_special_token_130|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128139": { + "content": "<|reserved_special_token_131|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128140": { + "content": "<|reserved_special_token_132|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128141": { + "content": "<|reserved_special_token_133|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128142": { + "content": "<|reserved_special_token_134|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128143": { + "content": "<|reserved_special_token_135|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128144": { + "content": "<|reserved_special_token_136|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128145": { + "content": "<|reserved_special_token_137|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128146": { + "content": "<|reserved_special_token_138|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128147": { + "content": "<|reserved_special_token_139|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128148": { + "content": "<|reserved_special_token_140|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128149": { + "content": "<|reserved_special_token_141|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128150": { + "content": "<|reserved_special_token_142|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128151": { + "content": "<|reserved_special_token_143|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128152": { + "content": "<|reserved_special_token_144|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128153": { + "content": "<|reserved_special_token_145|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128154": { + "content": "<|reserved_special_token_146|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128155": { + "content": "<|reserved_special_token_147|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128156": { + "content": "<|reserved_special_token_148|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128157": { + "content": "<|reserved_special_token_149|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128158": { + "content": "<|reserved_special_token_150|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128159": { + "content": "<|reserved_special_token_151|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128160": { + "content": "<|reserved_special_token_152|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128161": { + "content": "<|reserved_special_token_153|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128162": { + "content": "<|reserved_special_token_154|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128163": { + "content": "<|reserved_special_token_155|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128164": { + "content": "<|reserved_special_token_156|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128165": { + "content": "<|reserved_special_token_157|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128166": { + "content": "<|reserved_special_token_158|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128167": { + "content": "<|reserved_special_token_159|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128168": { + "content": "<|reserved_special_token_160|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128169": { + "content": "<|reserved_special_token_161|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128170": { + "content": "<|reserved_special_token_162|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128171": { + "content": "<|reserved_special_token_163|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128172": { + "content": "<|reserved_special_token_164|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128173": { + "content": "<|reserved_special_token_165|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128174": { + "content": "<|reserved_special_token_166|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128175": { + "content": "<|reserved_special_token_167|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128176": { + "content": "<|reserved_special_token_168|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128177": { + "content": "<|reserved_special_token_169|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128178": { + "content": "<|reserved_special_token_170|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128179": { + "content": "<|reserved_special_token_171|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128180": { + "content": "<|reserved_special_token_172|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128181": { + "content": "<|reserved_special_token_173|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128182": { + "content": "<|reserved_special_token_174|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128183": { + "content": "<|reserved_special_token_175|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128184": { + "content": "<|reserved_special_token_176|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128185": { + "content": "<|reserved_special_token_177|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128186": { + "content": "<|reserved_special_token_178|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128187": { + "content": "<|reserved_special_token_179|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128188": { + "content": "<|reserved_special_token_180|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128189": { + "content": "<|reserved_special_token_181|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128190": { + "content": "<|reserved_special_token_182|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128191": { + "content": "<|reserved_special_token_183|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128192": { + "content": "<|reserved_special_token_184|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128193": { + "content": "<|reserved_special_token_185|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128194": { + "content": "<|reserved_special_token_186|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128195": { + "content": "<|reserved_special_token_187|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128196": { + "content": "<|reserved_special_token_188|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128197": { + "content": "<|reserved_special_token_189|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128198": { + "content": "<|reserved_special_token_190|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128199": { + "content": "<|reserved_special_token_191|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128200": { + "content": "<|reserved_special_token_192|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128201": { + "content": "<|reserved_special_token_193|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128202": { + "content": "<|reserved_special_token_194|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128203": { + "content": "<|reserved_special_token_195|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128204": { + "content": "<|reserved_special_token_196|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128205": { + "content": "<|reserved_special_token_197|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128206": { + "content": "<|reserved_special_token_198|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128207": { + "content": "<|reserved_special_token_199|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128208": { + "content": "<|reserved_special_token_200|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128209": { + "content": "<|reserved_special_token_201|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128210": { + "content": "<|reserved_special_token_202|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128211": { + "content": "<|reserved_special_token_203|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128212": { + "content": "<|reserved_special_token_204|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128213": { + "content": "<|reserved_special_token_205|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128214": { + "content": "<|reserved_special_token_206|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128215": { + "content": "<|reserved_special_token_207|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128216": { + "content": "<|reserved_special_token_208|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128217": { + "content": "<|reserved_special_token_209|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128218": { + "content": "<|reserved_special_token_210|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128219": { + "content": "<|reserved_special_token_211|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128220": { + "content": "<|reserved_special_token_212|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128221": { + "content": "<|reserved_special_token_213|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128222": { + "content": "<|reserved_special_token_214|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128223": { + "content": "<|reserved_special_token_215|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128224": { + "content": "<|reserved_special_token_216|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128225": { + "content": "<|reserved_special_token_217|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128226": { + "content": "<|reserved_special_token_218|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128227": { + "content": "<|reserved_special_token_219|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128228": { + "content": "<|reserved_special_token_220|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128229": { + "content": "<|reserved_special_token_221|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128230": { + "content": "<|reserved_special_token_222|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128231": { + "content": "<|reserved_special_token_223|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128232": { + "content": "<|reserved_special_token_224|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128233": { + "content": "<|reserved_special_token_225|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128234": { + "content": "<|reserved_special_token_226|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128235": { + "content": "<|reserved_special_token_227|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128236": { + "content": "<|reserved_special_token_228|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128237": { + "content": "<|reserved_special_token_229|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128238": { + "content": "<|reserved_special_token_230|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128239": { + "content": "<|reserved_special_token_231|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128240": { + "content": "<|reserved_special_token_232|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128241": { + "content": "<|reserved_special_token_233|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128242": { + "content": "<|reserved_special_token_234|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128243": { + "content": "<|reserved_special_token_235|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128244": { + "content": "<|reserved_special_token_236|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128245": { + "content": "<|reserved_special_token_237|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128246": { + "content": "<|reserved_special_token_238|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128247": { + "content": "<|reserved_special_token_239|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128248": { + "content": "<|reserved_special_token_240|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128249": { + "content": "<|reserved_special_token_241|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128250": { + "content": "<|reserved_special_token_242|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128251": { + "content": "<|reserved_special_token_243|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128252": { + "content": "<|reserved_special_token_244|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128253": { + "content": "<|reserved_special_token_245|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128254": { + "content": "<|reserved_special_token_246|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128255": { + "content": "<|reserved_special_token_247|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128256": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "128257": { + "content": "🅁", + "lstrip": true, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|end_of_text|>", + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "[PAD]", + "tokenizer_class": "PreTrainedTokenizerFast" +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/trainer_state.json b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4a6f5d7076642a3995d476066712e3b983f2e40b --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/trainer_state.json @@ -0,0 +1,6273 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.8217770929635336, + "eval_steps": 10, + "global_step": 800, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7525, + "step": 1 + }, + { + "epoch": 0.002054442732408834, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7605, + "step": 2 + }, + { + "epoch": 0.0030816640986132513, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8008, + "step": 3 + }, + { + "epoch": 0.004108885464817668, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.809, + "step": 4 + }, + { + "epoch": 0.005136106831022085, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7841, + "step": 5 + }, + { + "epoch": 0.0061633281972265025, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7924, + "step": 6 + }, + { + "epoch": 0.007190549563430919, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7732, + "step": 7 + }, + { + "epoch": 0.008217770929635337, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7953, + "step": 8 + }, + { + "epoch": 0.009244992295839754, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7747, + "step": 9 + }, + { + "epoch": 0.01027221366204417, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7909, + "step": 10 + }, + { + "epoch": 0.01027221366204417, + "eval_loss": 2.813098192214966, + "eval_runtime": 38.1187, + "eval_samples_per_second": 26.234, + "eval_steps_per_second": 1.102, + "step": 10 + }, + { + "epoch": 0.011299435028248588, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7592, + "step": 11 + }, + { + "epoch": 0.012326656394453005, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7394, + "step": 12 + }, + { + "epoch": 0.01335387776065742, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8111, + "step": 13 + }, + { + "epoch": 0.014381099126861838, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.7786, + "step": 14 + }, + { + "epoch": 0.015408320493066256, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7711, + "step": 15 + }, + { + "epoch": 0.016435541859270673, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7622, + "step": 16 + }, + { + "epoch": 0.01746276322547509, + "grad_norm": 5.814355850219727, + "learning_rate": 1.7123287671232876e-08, + "loss": 2.7291, + "step": 17 + }, + { + "epoch": 0.01848998459167951, + "grad_norm": 6.8249125480651855, + "learning_rate": 3.424657534246575e-08, + "loss": 2.7995, + "step": 18 + }, + { + "epoch": 0.019517205957883924, + "grad_norm": 5.581949710845947, + "learning_rate": 5.136986301369863e-08, + "loss": 2.5998, + "step": 19 + }, + { + "epoch": 0.02054442732408834, + "grad_norm": 3.0888617038726807, + "learning_rate": 6.84931506849315e-08, + "loss": 2.6031, + "step": 20 + }, + { + "epoch": 0.02054442732408834, + "eval_loss": 2.60642147064209, + "eval_runtime": 38.2226, + "eval_samples_per_second": 26.163, + "eval_steps_per_second": 1.099, + "step": 20 + }, + { + "epoch": 0.02157164869029276, + "grad_norm": 4.42941427230835, + "learning_rate": 8.561643835616439e-08, + "loss": 2.5816, + "step": 21 + }, + { + "epoch": 0.022598870056497175, + "grad_norm": 2.359813690185547, + "learning_rate": 1.0273972602739726e-07, + "loss": 2.5352, + "step": 22 + }, + { + "epoch": 0.02362609142270159, + "grad_norm": 2.9022927284240723, + "learning_rate": 1.1986301369863014e-07, + "loss": 2.4641, + "step": 23 + }, + { + "epoch": 0.02465331278890601, + "grad_norm": 1.6895997524261475, + "learning_rate": 1.36986301369863e-07, + "loss": 2.4582, + "step": 24 + }, + { + "epoch": 0.025680534155110426, + "grad_norm": 2.4555776119232178, + "learning_rate": 1.541095890410959e-07, + "loss": 2.4171, + "step": 25 + }, + { + "epoch": 0.02670775552131484, + "grad_norm": 1.9735779762268066, + "learning_rate": 1.7123287671232878e-07, + "loss": 2.4249, + "step": 26 + }, + { + "epoch": 0.02773497688751926, + "grad_norm": 2.71161150932312, + "learning_rate": 1.8835616438356165e-07, + "loss": 2.3791, + "step": 27 + }, + { + "epoch": 0.028762198253723677, + "grad_norm": 1.8482733964920044, + "learning_rate": 2.0547945205479452e-07, + "loss": 2.374, + "step": 28 + }, + { + "epoch": 0.029789419619928096, + "grad_norm": 2.191246747970581, + "learning_rate": 2.226027397260274e-07, + "loss": 2.3557, + "step": 29 + }, + { + "epoch": 0.030816640986132512, + "grad_norm": 2.0575971603393555, + "learning_rate": 2.397260273972603e-07, + "loss": 2.3315, + "step": 30 + }, + { + "epoch": 0.030816640986132512, + "eval_loss": 2.399639844894409, + "eval_runtime": 37.737, + "eval_samples_per_second": 26.499, + "eval_steps_per_second": 1.113, + "step": 30 + }, + { + "epoch": 0.03184386235233693, + "grad_norm": 2.0346758365631104, + "learning_rate": 2.568493150684932e-07, + "loss": 2.2791, + "step": 31 + }, + { + "epoch": 0.03287108371854135, + "grad_norm": 1.8710602521896362, + "learning_rate": 2.73972602739726e-07, + "loss": 2.3219, + "step": 32 + }, + { + "epoch": 0.03389830508474576, + "grad_norm": 1.9982675313949585, + "learning_rate": 2.910958904109589e-07, + "loss": 2.273, + "step": 33 + }, + { + "epoch": 0.03492552645095018, + "grad_norm": 2.092076301574707, + "learning_rate": 3.082191780821918e-07, + "loss": 2.2555, + "step": 34 + }, + { + "epoch": 0.035952747817154594, + "grad_norm": 1.870373249053955, + "learning_rate": 3.2534246575342466e-07, + "loss": 2.2314, + "step": 35 + }, + { + "epoch": 0.03697996918335902, + "grad_norm": 1.8378052711486816, + "learning_rate": 3.4246575342465755e-07, + "loss": 2.1244, + "step": 36 + }, + { + "epoch": 0.03800719054956343, + "grad_norm": 2.0156054496765137, + "learning_rate": 3.595890410958904e-07, + "loss": 2.2502, + "step": 37 + }, + { + "epoch": 0.03903441191576785, + "grad_norm": 1.6175312995910645, + "learning_rate": 3.767123287671233e-07, + "loss": 2.1566, + "step": 38 + }, + { + "epoch": 0.040061633281972264, + "grad_norm": 1.9051450490951538, + "learning_rate": 3.938356164383562e-07, + "loss": 2.167, + "step": 39 + }, + { + "epoch": 0.04108885464817668, + "grad_norm": 2.383063793182373, + "learning_rate": 4.1095890410958903e-07, + "loss": 2.2099, + "step": 40 + }, + { + "epoch": 0.04108885464817668, + "eval_loss": 2.2763354778289795, + "eval_runtime": 37.8085, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 40 + }, + { + "epoch": 0.042116076014381096, + "grad_norm": 2.1310720443725586, + "learning_rate": 4.2808219178082193e-07, + "loss": 2.1714, + "step": 41 + }, + { + "epoch": 0.04314329738058552, + "grad_norm": 1.771941900253296, + "learning_rate": 4.452054794520548e-07, + "loss": 2.0895, + "step": 42 + }, + { + "epoch": 0.044170518746789934, + "grad_norm": 2.3784172534942627, + "learning_rate": 4.6232876712328767e-07, + "loss": 2.1555, + "step": 43 + }, + { + "epoch": 0.04519774011299435, + "grad_norm": 1.6692813634872437, + "learning_rate": 4.794520547945206e-07, + "loss": 2.11, + "step": 44 + }, + { + "epoch": 0.046224961479198766, + "grad_norm": 2.1780643463134766, + "learning_rate": 4.965753424657534e-07, + "loss": 2.1256, + "step": 45 + }, + { + "epoch": 0.04725218284540318, + "grad_norm": 1.9390568733215332, + "learning_rate": 5.136986301369864e-07, + "loss": 2.0971, + "step": 46 + }, + { + "epoch": 0.048279404211607604, + "grad_norm": 2.276216506958008, + "learning_rate": 5.308219178082192e-07, + "loss": 2.1479, + "step": 47 + }, + { + "epoch": 0.04930662557781202, + "grad_norm": 2.148573160171509, + "learning_rate": 5.47945205479452e-07, + "loss": 2.0808, + "step": 48 + }, + { + "epoch": 0.050333846944016436, + "grad_norm": 2.259422540664673, + "learning_rate": 5.65068493150685e-07, + "loss": 2.0823, + "step": 49 + }, + { + "epoch": 0.05136106831022085, + "grad_norm": 1.7703359127044678, + "learning_rate": 5.821917808219178e-07, + "loss": 2.0236, + "step": 50 + }, + { + "epoch": 0.05136106831022085, + "eval_loss": 2.2037274837493896, + "eval_runtime": 37.8385, + "eval_samples_per_second": 26.428, + "eval_steps_per_second": 1.11, + "step": 50 + }, + { + "epoch": 0.05238828967642527, + "grad_norm": 2.56268310546875, + "learning_rate": 5.993150684931507e-07, + "loss": 2.0228, + "step": 51 + }, + { + "epoch": 0.05341551104262968, + "grad_norm": 2.5256848335266113, + "learning_rate": 6.164383561643836e-07, + "loss": 2.1401, + "step": 52 + }, + { + "epoch": 0.054442732408834106, + "grad_norm": 2.6329638957977295, + "learning_rate": 6.335616438356165e-07, + "loss": 2.0107, + "step": 53 + }, + { + "epoch": 0.05546995377503852, + "grad_norm": 2.477762222290039, + "learning_rate": 6.506849315068493e-07, + "loss": 2.0648, + "step": 54 + }, + { + "epoch": 0.05649717514124294, + "grad_norm": 1.8869198560714722, + "learning_rate": 6.678082191780823e-07, + "loss": 2.0214, + "step": 55 + }, + { + "epoch": 0.05752439650744735, + "grad_norm": 2.707345485687256, + "learning_rate": 6.849315068493151e-07, + "loss": 1.9893, + "step": 56 + }, + { + "epoch": 0.05855161787365177, + "grad_norm": 2.067230701446533, + "learning_rate": 7.020547945205481e-07, + "loss": 2.0308, + "step": 57 + }, + { + "epoch": 0.05957883923985619, + "grad_norm": 2.1961982250213623, + "learning_rate": 7.191780821917808e-07, + "loss": 1.948, + "step": 58 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 2.1032280921936035, + "learning_rate": 7.363013698630137e-07, + "loss": 2.0098, + "step": 59 + }, + { + "epoch": 0.061633281972265024, + "grad_norm": 2.5191144943237305, + "learning_rate": 7.534246575342466e-07, + "loss": 2.0183, + "step": 60 + }, + { + "epoch": 0.061633281972265024, + "eval_loss": 2.1474344730377197, + "eval_runtime": 37.8631, + "eval_samples_per_second": 26.411, + "eval_steps_per_second": 1.109, + "step": 60 + }, + { + "epoch": 0.06266050333846944, + "grad_norm": 2.177739143371582, + "learning_rate": 7.705479452054795e-07, + "loss": 1.9702, + "step": 61 + }, + { + "epoch": 0.06368772470467386, + "grad_norm": 2.499784231185913, + "learning_rate": 7.876712328767124e-07, + "loss": 1.9825, + "step": 62 + }, + { + "epoch": 0.06471494607087827, + "grad_norm": 2.181591749191284, + "learning_rate": 8.047945205479453e-07, + "loss": 2.0414, + "step": 63 + }, + { + "epoch": 0.0657421674370827, + "grad_norm": 1.9575999975204468, + "learning_rate": 8.219178082191781e-07, + "loss": 2.0115, + "step": 64 + }, + { + "epoch": 0.0667693888032871, + "grad_norm": 2.3431336879730225, + "learning_rate": 8.39041095890411e-07, + "loss": 1.8781, + "step": 65 + }, + { + "epoch": 0.06779661016949153, + "grad_norm": 2.042747974395752, + "learning_rate": 8.561643835616439e-07, + "loss": 1.9798, + "step": 66 + }, + { + "epoch": 0.06882383153569595, + "grad_norm": 2.784104585647583, + "learning_rate": 8.732876712328768e-07, + "loss": 1.9274, + "step": 67 + }, + { + "epoch": 0.06985105290190036, + "grad_norm": 1.9781445264816284, + "learning_rate": 8.904109589041097e-07, + "loss": 2.0036, + "step": 68 + }, + { + "epoch": 0.07087827426810478, + "grad_norm": 2.631537437438965, + "learning_rate": 9.075342465753426e-07, + "loss": 1.938, + "step": 69 + }, + { + "epoch": 0.07190549563430919, + "grad_norm": 1.902701497077942, + "learning_rate": 9.246575342465753e-07, + "loss": 2.0545, + "step": 70 + }, + { + "epoch": 0.07190549563430919, + "eval_loss": 2.115269422531128, + "eval_runtime": 37.887, + "eval_samples_per_second": 26.394, + "eval_steps_per_second": 1.109, + "step": 70 + }, + { + "epoch": 0.07293271700051361, + "grad_norm": 1.9323904514312744, + "learning_rate": 9.417808219178083e-07, + "loss": 1.9463, + "step": 71 + }, + { + "epoch": 0.07395993836671803, + "grad_norm": 2.1631956100463867, + "learning_rate": 9.589041095890411e-07, + "loss": 1.8982, + "step": 72 + }, + { + "epoch": 0.07498715973292244, + "grad_norm": 1.959518313407898, + "learning_rate": 9.76027397260274e-07, + "loss": 1.9538, + "step": 73 + }, + { + "epoch": 0.07601438109912687, + "grad_norm": 1.9834569692611694, + "learning_rate": 9.931506849315068e-07, + "loss": 1.9257, + "step": 74 + }, + { + "epoch": 0.07704160246533127, + "grad_norm": 2.2818732261657715, + "learning_rate": 1.0102739726027399e-06, + "loss": 1.9318, + "step": 75 + }, + { + "epoch": 0.0780688238315357, + "grad_norm": 1.7696406841278076, + "learning_rate": 1.0273972602739727e-06, + "loss": 1.912, + "step": 76 + }, + { + "epoch": 0.07909604519774012, + "grad_norm": 2.396303176879883, + "learning_rate": 1.0445205479452056e-06, + "loss": 1.9482, + "step": 77 + }, + { + "epoch": 0.08012326656394453, + "grad_norm": 2.2371881008148193, + "learning_rate": 1.0616438356164384e-06, + "loss": 1.9448, + "step": 78 + }, + { + "epoch": 0.08115048793014895, + "grad_norm": 2.40077543258667, + "learning_rate": 1.0787671232876712e-06, + "loss": 1.8764, + "step": 79 + }, + { + "epoch": 0.08217770929635336, + "grad_norm": 2.267916679382324, + "learning_rate": 1.095890410958904e-06, + "loss": 1.9478, + "step": 80 + }, + { + "epoch": 0.08217770929635336, + "eval_loss": 2.0851893424987793, + "eval_runtime": 37.9176, + "eval_samples_per_second": 26.373, + "eval_steps_per_second": 1.108, + "step": 80 + }, + { + "epoch": 0.08320493066255778, + "grad_norm": 2.286259651184082, + "learning_rate": 1.1130136986301371e-06, + "loss": 1.8856, + "step": 81 + }, + { + "epoch": 0.08423215202876219, + "grad_norm": 2.160848379135132, + "learning_rate": 1.13013698630137e-06, + "loss": 1.9196, + "step": 82 + }, + { + "epoch": 0.08525937339496661, + "grad_norm": 2.3618931770324707, + "learning_rate": 1.1472602739726028e-06, + "loss": 1.9229, + "step": 83 + }, + { + "epoch": 0.08628659476117104, + "grad_norm": 2.1963961124420166, + "learning_rate": 1.1643835616438357e-06, + "loss": 1.8998, + "step": 84 + }, + { + "epoch": 0.08731381612737545, + "grad_norm": 2.197981357574463, + "learning_rate": 1.1815068493150685e-06, + "loss": 1.8716, + "step": 85 + }, + { + "epoch": 0.08834103749357987, + "grad_norm": 2.3562612533569336, + "learning_rate": 1.1986301369863014e-06, + "loss": 1.9336, + "step": 86 + }, + { + "epoch": 0.08936825885978428, + "grad_norm": 2.49712872505188, + "learning_rate": 1.2157534246575344e-06, + "loss": 1.8827, + "step": 87 + }, + { + "epoch": 0.0903954802259887, + "grad_norm": 2.315244436264038, + "learning_rate": 1.2328767123287673e-06, + "loss": 1.8801, + "step": 88 + }, + { + "epoch": 0.09142270159219312, + "grad_norm": 2.774587869644165, + "learning_rate": 1.25e-06, + "loss": 1.8972, + "step": 89 + }, + { + "epoch": 0.09244992295839753, + "grad_norm": 2.262364387512207, + "learning_rate": 1.267123287671233e-06, + "loss": 1.8426, + "step": 90 + }, + { + "epoch": 0.09244992295839753, + "eval_loss": 2.069462299346924, + "eval_runtime": 37.9319, + "eval_samples_per_second": 26.363, + "eval_steps_per_second": 1.107, + "step": 90 + }, + { + "epoch": 0.09347714432460195, + "grad_norm": 2.6195597648620605, + "learning_rate": 1.284246575342466e-06, + "loss": 1.871, + "step": 91 + }, + { + "epoch": 0.09450436569080636, + "grad_norm": 2.051847219467163, + "learning_rate": 1.3013698630136986e-06, + "loss": 1.8922, + "step": 92 + }, + { + "epoch": 0.09553158705701079, + "grad_norm": 2.5930886268615723, + "learning_rate": 1.3184931506849317e-06, + "loss": 1.8818, + "step": 93 + }, + { + "epoch": 0.09655880842321521, + "grad_norm": 2.304164171218872, + "learning_rate": 1.3356164383561645e-06, + "loss": 1.8779, + "step": 94 + }, + { + "epoch": 0.09758602978941962, + "grad_norm": 2.187354564666748, + "learning_rate": 1.3527397260273976e-06, + "loss": 1.9272, + "step": 95 + }, + { + "epoch": 0.09861325115562404, + "grad_norm": 2.2492458820343018, + "learning_rate": 1.3698630136986302e-06, + "loss": 1.805, + "step": 96 + }, + { + "epoch": 0.09964047252182845, + "grad_norm": 2.3325352668762207, + "learning_rate": 1.386986301369863e-06, + "loss": 1.8817, + "step": 97 + }, + { + "epoch": 0.10066769388803287, + "grad_norm": 2.1708617210388184, + "learning_rate": 1.4041095890410961e-06, + "loss": 1.7956, + "step": 98 + }, + { + "epoch": 0.1016949152542373, + "grad_norm": 2.4267706871032715, + "learning_rate": 1.421232876712329e-06, + "loss": 1.8116, + "step": 99 + }, + { + "epoch": 0.1027221366204417, + "grad_norm": 2.3286805152893066, + "learning_rate": 1.4383561643835616e-06, + "loss": 1.8299, + "step": 100 + }, + { + "epoch": 0.1027221366204417, + "eval_loss": 2.0549135208129883, + "eval_runtime": 37.935, + "eval_samples_per_second": 26.361, + "eval_steps_per_second": 1.107, + "step": 100 + }, + { + "epoch": 0.10374935798664613, + "grad_norm": 2.5391643047332764, + "learning_rate": 1.4554794520547946e-06, + "loss": 1.8521, + "step": 101 + }, + { + "epoch": 0.10477657935285054, + "grad_norm": 2.018380880355835, + "learning_rate": 1.4726027397260275e-06, + "loss": 1.8209, + "step": 102 + }, + { + "epoch": 0.10580380071905496, + "grad_norm": 2.3706746101379395, + "learning_rate": 1.4897260273972605e-06, + "loss": 1.79, + "step": 103 + }, + { + "epoch": 0.10683102208525937, + "grad_norm": 2.1290764808654785, + "learning_rate": 1.5068493150684932e-06, + "loss": 1.8141, + "step": 104 + }, + { + "epoch": 0.10785824345146379, + "grad_norm": 2.2768120765686035, + "learning_rate": 1.5239726027397262e-06, + "loss": 1.8826, + "step": 105 + }, + { + "epoch": 0.10888546481766821, + "grad_norm": 2.316584348678589, + "learning_rate": 1.541095890410959e-06, + "loss": 1.8604, + "step": 106 + }, + { + "epoch": 0.10991268618387262, + "grad_norm": 2.14679217338562, + "learning_rate": 1.5582191780821921e-06, + "loss": 1.865, + "step": 107 + }, + { + "epoch": 0.11093990755007704, + "grad_norm": 2.3288021087646484, + "learning_rate": 1.5753424657534248e-06, + "loss": 1.8512, + "step": 108 + }, + { + "epoch": 0.11196712891628145, + "grad_norm": 2.300482988357544, + "learning_rate": 1.5924657534246576e-06, + "loss": 1.7455, + "step": 109 + }, + { + "epoch": 0.11299435028248588, + "grad_norm": 2.4226229190826416, + "learning_rate": 1.6095890410958907e-06, + "loss": 1.8058, + "step": 110 + }, + { + "epoch": 0.11299435028248588, + "eval_loss": 2.025261640548706, + "eval_runtime": 37.6779, + "eval_samples_per_second": 26.541, + "eval_steps_per_second": 1.115, + "step": 110 + }, + { + "epoch": 0.1140215716486903, + "grad_norm": 2.0930111408233643, + "learning_rate": 1.6267123287671235e-06, + "loss": 1.7375, + "step": 111 + }, + { + "epoch": 0.1150487930148947, + "grad_norm": 2.4973645210266113, + "learning_rate": 1.6438356164383561e-06, + "loss": 1.8268, + "step": 112 + }, + { + "epoch": 0.11607601438109913, + "grad_norm": 2.3295114040374756, + "learning_rate": 1.6609589041095892e-06, + "loss": 1.7882, + "step": 113 + }, + { + "epoch": 0.11710323574730354, + "grad_norm": 2.3693318367004395, + "learning_rate": 1.678082191780822e-06, + "loss": 1.8192, + "step": 114 + }, + { + "epoch": 0.11813045711350796, + "grad_norm": 2.553637742996216, + "learning_rate": 1.695205479452055e-06, + "loss": 1.7491, + "step": 115 + }, + { + "epoch": 0.11915767847971238, + "grad_norm": 2.303945541381836, + "learning_rate": 1.7123287671232877e-06, + "loss": 1.8737, + "step": 116 + }, + { + "epoch": 0.12018489984591679, + "grad_norm": 2.250356435775757, + "learning_rate": 1.7294520547945206e-06, + "loss": 1.8039, + "step": 117 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 2.3854403495788574, + "learning_rate": 1.7465753424657536e-06, + "loss": 1.8357, + "step": 118 + }, + { + "epoch": 0.12223934257832562, + "grad_norm": 2.4049572944641113, + "learning_rate": 1.7636986301369865e-06, + "loss": 1.8209, + "step": 119 + }, + { + "epoch": 0.12326656394453005, + "grad_norm": 2.313551425933838, + "learning_rate": 1.7808219178082193e-06, + "loss": 1.9026, + "step": 120 + }, + { + "epoch": 0.12326656394453005, + "eval_loss": 2.012171745300293, + "eval_runtime": 37.7908, + "eval_samples_per_second": 26.461, + "eval_steps_per_second": 1.111, + "step": 120 + }, + { + "epoch": 0.12429378531073447, + "grad_norm": 2.518320322036743, + "learning_rate": 1.7979452054794521e-06, + "loss": 1.8241, + "step": 121 + }, + { + "epoch": 0.12532100667693888, + "grad_norm": 2.6504592895507812, + "learning_rate": 1.8150684931506852e-06, + "loss": 1.8091, + "step": 122 + }, + { + "epoch": 0.1263482280431433, + "grad_norm": 2.556610107421875, + "learning_rate": 1.832191780821918e-06, + "loss": 1.7547, + "step": 123 + }, + { + "epoch": 0.12737544940934772, + "grad_norm": 2.247633695602417, + "learning_rate": 1.8493150684931507e-06, + "loss": 1.7923, + "step": 124 + }, + { + "epoch": 0.12840267077555212, + "grad_norm": 2.195596933364868, + "learning_rate": 1.8664383561643837e-06, + "loss": 1.747, + "step": 125 + }, + { + "epoch": 0.12942989214175654, + "grad_norm": 2.121396780014038, + "learning_rate": 1.8835616438356166e-06, + "loss": 1.8088, + "step": 126 + }, + { + "epoch": 0.13045711350796096, + "grad_norm": 2.4572501182556152, + "learning_rate": 1.9006849315068496e-06, + "loss": 1.7285, + "step": 127 + }, + { + "epoch": 0.1314843348741654, + "grad_norm": 2.2590177059173584, + "learning_rate": 1.9178082191780823e-06, + "loss": 1.7929, + "step": 128 + }, + { + "epoch": 0.1325115562403698, + "grad_norm": 2.5099234580993652, + "learning_rate": 1.9349315068493153e-06, + "loss": 1.8365, + "step": 129 + }, + { + "epoch": 0.1335387776065742, + "grad_norm": 2.3064465522766113, + "learning_rate": 1.952054794520548e-06, + "loss": 1.8351, + "step": 130 + }, + { + "epoch": 0.1335387776065742, + "eval_loss": 2.0043866634368896, + "eval_runtime": 37.8661, + "eval_samples_per_second": 26.409, + "eval_steps_per_second": 1.109, + "step": 130 + }, + { + "epoch": 0.13456599897277863, + "grad_norm": 2.5977718830108643, + "learning_rate": 1.969178082191781e-06, + "loss": 1.7603, + "step": 131 + }, + { + "epoch": 0.13559322033898305, + "grad_norm": 2.23221755027771, + "learning_rate": 1.9863013698630136e-06, + "loss": 1.8237, + "step": 132 + }, + { + "epoch": 0.13662044170518747, + "grad_norm": 2.247044324874878, + "learning_rate": 2.0034246575342467e-06, + "loss": 1.8006, + "step": 133 + }, + { + "epoch": 0.1376476630713919, + "grad_norm": 2.1796107292175293, + "learning_rate": 2.0205479452054797e-06, + "loss": 1.7354, + "step": 134 + }, + { + "epoch": 0.1386748844375963, + "grad_norm": 2.3269693851470947, + "learning_rate": 2.037671232876713e-06, + "loss": 1.783, + "step": 135 + }, + { + "epoch": 0.1397021058038007, + "grad_norm": 2.2903053760528564, + "learning_rate": 2.0547945205479454e-06, + "loss": 1.8072, + "step": 136 + }, + { + "epoch": 0.14072932717000514, + "grad_norm": 2.634150981903076, + "learning_rate": 2.071917808219178e-06, + "loss": 1.7251, + "step": 137 + }, + { + "epoch": 0.14175654853620956, + "grad_norm": 2.42579984664917, + "learning_rate": 2.089041095890411e-06, + "loss": 1.7997, + "step": 138 + }, + { + "epoch": 0.14278376990241398, + "grad_norm": 2.615828514099121, + "learning_rate": 2.106164383561644e-06, + "loss": 1.766, + "step": 139 + }, + { + "epoch": 0.14381099126861838, + "grad_norm": 2.312714099884033, + "learning_rate": 2.123287671232877e-06, + "loss": 1.7538, + "step": 140 + }, + { + "epoch": 0.14381099126861838, + "eval_loss": 1.9918214082717896, + "eval_runtime": 37.8698, + "eval_samples_per_second": 26.406, + "eval_steps_per_second": 1.109, + "step": 140 + }, + { + "epoch": 0.1448382126348228, + "grad_norm": 2.7540030479431152, + "learning_rate": 2.14041095890411e-06, + "loss": 1.7856, + "step": 141 + }, + { + "epoch": 0.14586543400102722, + "grad_norm": 2.289191246032715, + "learning_rate": 2.1575342465753425e-06, + "loss": 1.7288, + "step": 142 + }, + { + "epoch": 0.14689265536723164, + "grad_norm": 2.788001775741577, + "learning_rate": 2.1746575342465755e-06, + "loss": 1.8703, + "step": 143 + }, + { + "epoch": 0.14791987673343607, + "grad_norm": 2.4128854274749756, + "learning_rate": 2.191780821917808e-06, + "loss": 1.7461, + "step": 144 + }, + { + "epoch": 0.14894709809964046, + "grad_norm": 2.4108099937438965, + "learning_rate": 2.2089041095890412e-06, + "loss": 1.73, + "step": 145 + }, + { + "epoch": 0.14997431946584489, + "grad_norm": 2.4315948486328125, + "learning_rate": 2.2260273972602743e-06, + "loss": 1.7303, + "step": 146 + }, + { + "epoch": 0.1510015408320493, + "grad_norm": 2.494056224822998, + "learning_rate": 2.243150684931507e-06, + "loss": 1.7616, + "step": 147 + }, + { + "epoch": 0.15202876219825373, + "grad_norm": 2.4557394981384277, + "learning_rate": 2.26027397260274e-06, + "loss": 1.7655, + "step": 148 + }, + { + "epoch": 0.15305598356445815, + "grad_norm": 2.4114091396331787, + "learning_rate": 2.2773972602739726e-06, + "loss": 1.8538, + "step": 149 + }, + { + "epoch": 0.15408320493066255, + "grad_norm": 2.609219789505005, + "learning_rate": 2.2945205479452057e-06, + "loss": 1.7767, + "step": 150 + }, + { + "epoch": 0.15408320493066255, + "eval_loss": 1.981367588043213, + "eval_runtime": 37.9562, + "eval_samples_per_second": 26.346, + "eval_steps_per_second": 1.107, + "step": 150 + }, + { + "epoch": 0.15511042629686697, + "grad_norm": 2.213622808456421, + "learning_rate": 2.3116438356164387e-06, + "loss": 1.687, + "step": 151 + }, + { + "epoch": 0.1561376476630714, + "grad_norm": 2.6517975330352783, + "learning_rate": 2.3287671232876713e-06, + "loss": 1.7311, + "step": 152 + }, + { + "epoch": 0.15716486902927582, + "grad_norm": 2.5174458026885986, + "learning_rate": 2.3458904109589044e-06, + "loss": 1.7835, + "step": 153 + }, + { + "epoch": 0.15819209039548024, + "grad_norm": 2.485480308532715, + "learning_rate": 2.363013698630137e-06, + "loss": 1.7749, + "step": 154 + }, + { + "epoch": 0.15921931176168463, + "grad_norm": 2.5386061668395996, + "learning_rate": 2.38013698630137e-06, + "loss": 1.7223, + "step": 155 + }, + { + "epoch": 0.16024653312788906, + "grad_norm": 2.2895941734313965, + "learning_rate": 2.3972602739726027e-06, + "loss": 1.6845, + "step": 156 + }, + { + "epoch": 0.16127375449409348, + "grad_norm": 2.5453789234161377, + "learning_rate": 2.4143835616438358e-06, + "loss": 1.7627, + "step": 157 + }, + { + "epoch": 0.1623009758602979, + "grad_norm": 2.5322303771972656, + "learning_rate": 2.431506849315069e-06, + "loss": 1.764, + "step": 158 + }, + { + "epoch": 0.1633281972265023, + "grad_norm": 2.2669899463653564, + "learning_rate": 2.4486301369863015e-06, + "loss": 1.7487, + "step": 159 + }, + { + "epoch": 0.16435541859270672, + "grad_norm": 2.4577853679656982, + "learning_rate": 2.4657534246575345e-06, + "loss": 1.7635, + "step": 160 + }, + { + "epoch": 0.16435541859270672, + "eval_loss": 1.9793355464935303, + "eval_runtime": 37.8576, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 160 + }, + { + "epoch": 0.16538263995891114, + "grad_norm": 2.354893922805786, + "learning_rate": 2.482876712328767e-06, + "loss": 1.7538, + "step": 161 + }, + { + "epoch": 0.16640986132511557, + "grad_norm": 2.3255856037139893, + "learning_rate": 2.5e-06, + "loss": 1.7277, + "step": 162 + }, + { + "epoch": 0.16743708269132, + "grad_norm": 2.3501994609832764, + "learning_rate": 2.5171232876712333e-06, + "loss": 1.6985, + "step": 163 + }, + { + "epoch": 0.16846430405752438, + "grad_norm": 2.386627674102783, + "learning_rate": 2.534246575342466e-06, + "loss": 1.7552, + "step": 164 + }, + { + "epoch": 0.1694915254237288, + "grad_norm": 2.2238457202911377, + "learning_rate": 2.551369863013699e-06, + "loss": 1.7805, + "step": 165 + }, + { + "epoch": 0.17051874678993323, + "grad_norm": 2.2653403282165527, + "learning_rate": 2.568493150684932e-06, + "loss": 1.7656, + "step": 166 + }, + { + "epoch": 0.17154596815613765, + "grad_norm": 2.233132839202881, + "learning_rate": 2.585616438356164e-06, + "loss": 1.6997, + "step": 167 + }, + { + "epoch": 0.17257318952234207, + "grad_norm": 2.1696043014526367, + "learning_rate": 2.6027397260273973e-06, + "loss": 1.7487, + "step": 168 + }, + { + "epoch": 0.17360041088854647, + "grad_norm": 2.4479899406433105, + "learning_rate": 2.6198630136986303e-06, + "loss": 1.7957, + "step": 169 + }, + { + "epoch": 0.1746276322547509, + "grad_norm": 2.3054702281951904, + "learning_rate": 2.6369863013698634e-06, + "loss": 1.766, + "step": 170 + }, + { + "epoch": 0.1746276322547509, + "eval_loss": 1.9543790817260742, + "eval_runtime": 37.8837, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 170 + }, + { + "epoch": 0.17565485362095531, + "grad_norm": 2.3591110706329346, + "learning_rate": 2.654109589041096e-06, + "loss": 1.7149, + "step": 171 + }, + { + "epoch": 0.17668207498715974, + "grad_norm": 2.450526237487793, + "learning_rate": 2.671232876712329e-06, + "loss": 1.7074, + "step": 172 + }, + { + "epoch": 0.17770929635336416, + "grad_norm": 2.1511754989624023, + "learning_rate": 2.688356164383562e-06, + "loss": 1.7036, + "step": 173 + }, + { + "epoch": 0.17873651771956856, + "grad_norm": 2.5944275856018066, + "learning_rate": 2.705479452054795e-06, + "loss": 1.6566, + "step": 174 + }, + { + "epoch": 0.17976373908577298, + "grad_norm": 2.4301369190216064, + "learning_rate": 2.7226027397260274e-06, + "loss": 1.7977, + "step": 175 + }, + { + "epoch": 0.1807909604519774, + "grad_norm": 2.287665367126465, + "learning_rate": 2.7397260273972604e-06, + "loss": 1.7526, + "step": 176 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.380032777786255, + "learning_rate": 2.7568493150684935e-06, + "loss": 1.6753, + "step": 177 + }, + { + "epoch": 0.18284540318438625, + "grad_norm": 2.4040510654449463, + "learning_rate": 2.773972602739726e-06, + "loss": 1.758, + "step": 178 + }, + { + "epoch": 0.18387262455059064, + "grad_norm": 2.5074045658111572, + "learning_rate": 2.791095890410959e-06, + "loss": 1.6498, + "step": 179 + }, + { + "epoch": 0.18489984591679506, + "grad_norm": 2.3926949501037598, + "learning_rate": 2.8082191780821922e-06, + "loss": 1.8006, + "step": 180 + }, + { + "epoch": 0.18489984591679506, + "eval_loss": 1.9446762800216675, + "eval_runtime": 37.8572, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 180 + }, + { + "epoch": 0.1859270672829995, + "grad_norm": 2.4991633892059326, + "learning_rate": 2.8253424657534253e-06, + "loss": 1.7649, + "step": 181 + }, + { + "epoch": 0.1869542886492039, + "grad_norm": 2.2558324337005615, + "learning_rate": 2.842465753424658e-06, + "loss": 1.7743, + "step": 182 + }, + { + "epoch": 0.18798151001540833, + "grad_norm": 2.219717025756836, + "learning_rate": 2.8595890410958905e-06, + "loss": 1.7807, + "step": 183 + }, + { + "epoch": 0.18900873138161273, + "grad_norm": 2.406562566757202, + "learning_rate": 2.876712328767123e-06, + "loss": 1.6289, + "step": 184 + }, + { + "epoch": 0.19003595274781715, + "grad_norm": 2.356579303741455, + "learning_rate": 2.8938356164383562e-06, + "loss": 1.6747, + "step": 185 + }, + { + "epoch": 0.19106317411402157, + "grad_norm": 2.4844322204589844, + "learning_rate": 2.9109589041095893e-06, + "loss": 1.6731, + "step": 186 + }, + { + "epoch": 0.192090395480226, + "grad_norm": 2.8179149627685547, + "learning_rate": 2.9280821917808223e-06, + "loss": 1.6786, + "step": 187 + }, + { + "epoch": 0.19311761684643042, + "grad_norm": 2.395416259765625, + "learning_rate": 2.945205479452055e-06, + "loss": 1.7313, + "step": 188 + }, + { + "epoch": 0.1941448382126348, + "grad_norm": 2.848740816116333, + "learning_rate": 2.962328767123288e-06, + "loss": 1.7258, + "step": 189 + }, + { + "epoch": 0.19517205957883924, + "grad_norm": 2.427395820617676, + "learning_rate": 2.979452054794521e-06, + "loss": 1.6633, + "step": 190 + }, + { + "epoch": 0.19517205957883924, + "eval_loss": 1.9443999528884888, + "eval_runtime": 37.8579, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 1.109, + "step": 190 + }, + { + "epoch": 0.19619928094504366, + "grad_norm": 2.5347256660461426, + "learning_rate": 2.9965753424657533e-06, + "loss": 1.7342, + "step": 191 + }, + { + "epoch": 0.19722650231124808, + "grad_norm": 2.8696866035461426, + "learning_rate": 3.0136986301369864e-06, + "loss": 1.7037, + "step": 192 + }, + { + "epoch": 0.1982537236774525, + "grad_norm": 2.427468776702881, + "learning_rate": 3.0308219178082194e-06, + "loss": 1.672, + "step": 193 + }, + { + "epoch": 0.1992809450436569, + "grad_norm": 2.4253737926483154, + "learning_rate": 3.0479452054794525e-06, + "loss": 1.7252, + "step": 194 + }, + { + "epoch": 0.20030816640986132, + "grad_norm": 2.801668643951416, + "learning_rate": 3.065068493150685e-06, + "loss": 1.6694, + "step": 195 + }, + { + "epoch": 0.20133538777606574, + "grad_norm": 2.442263603210449, + "learning_rate": 3.082191780821918e-06, + "loss": 1.6866, + "step": 196 + }, + { + "epoch": 0.20236260914227017, + "grad_norm": 2.777463674545288, + "learning_rate": 3.099315068493151e-06, + "loss": 1.686, + "step": 197 + }, + { + "epoch": 0.2033898305084746, + "grad_norm": 2.491227388381958, + "learning_rate": 3.1164383561643843e-06, + "loss": 1.6414, + "step": 198 + }, + { + "epoch": 0.20441705187467898, + "grad_norm": 2.33585262298584, + "learning_rate": 3.1335616438356165e-06, + "loss": 1.7808, + "step": 199 + }, + { + "epoch": 0.2054442732408834, + "grad_norm": 2.518160581588745, + "learning_rate": 3.1506849315068495e-06, + "loss": 1.6592, + "step": 200 + }, + { + "epoch": 0.2054442732408834, + "eval_loss": 1.9371097087860107, + "eval_runtime": 37.8602, + "eval_samples_per_second": 26.413, + "eval_steps_per_second": 1.109, + "step": 200 + }, + { + "epoch": 0.20647149460708783, + "grad_norm": 2.460204839706421, + "learning_rate": 3.167808219178082e-06, + "loss": 1.6668, + "step": 201 + }, + { + "epoch": 0.20749871597329225, + "grad_norm": 2.377824068069458, + "learning_rate": 3.184931506849315e-06, + "loss": 1.6462, + "step": 202 + }, + { + "epoch": 0.20852593733949668, + "grad_norm": 2.426177740097046, + "learning_rate": 3.2020547945205483e-06, + "loss": 1.6272, + "step": 203 + }, + { + "epoch": 0.20955315870570107, + "grad_norm": 2.4532034397125244, + "learning_rate": 3.2191780821917813e-06, + "loss": 1.6169, + "step": 204 + }, + { + "epoch": 0.2105803800719055, + "grad_norm": 2.517059564590454, + "learning_rate": 3.236301369863014e-06, + "loss": 1.5991, + "step": 205 + }, + { + "epoch": 0.21160760143810992, + "grad_norm": 2.22658109664917, + "learning_rate": 3.253424657534247e-06, + "loss": 1.6754, + "step": 206 + }, + { + "epoch": 0.21263482280431434, + "grad_norm": 2.344799518585205, + "learning_rate": 3.2705479452054796e-06, + "loss": 1.6648, + "step": 207 + }, + { + "epoch": 0.21366204417051873, + "grad_norm": 2.331875801086426, + "learning_rate": 3.2876712328767123e-06, + "loss": 1.6112, + "step": 208 + }, + { + "epoch": 0.21468926553672316, + "grad_norm": 2.262343645095825, + "learning_rate": 3.3047945205479453e-06, + "loss": 1.6479, + "step": 209 + }, + { + "epoch": 0.21571648690292758, + "grad_norm": 2.282278537750244, + "learning_rate": 3.3219178082191784e-06, + "loss": 1.6753, + "step": 210 + }, + { + "epoch": 0.21571648690292758, + "eval_loss": 1.9338418245315552, + "eval_runtime": 37.6083, + "eval_samples_per_second": 26.59, + "eval_steps_per_second": 1.117, + "step": 210 + }, + { + "epoch": 0.216743708269132, + "grad_norm": 2.578476905822754, + "learning_rate": 3.3390410958904114e-06, + "loss": 1.7043, + "step": 211 + }, + { + "epoch": 0.21777092963533642, + "grad_norm": 2.3118813037872314, + "learning_rate": 3.356164383561644e-06, + "loss": 1.7541, + "step": 212 + }, + { + "epoch": 0.21879815100154082, + "grad_norm": 2.2418692111968994, + "learning_rate": 3.373287671232877e-06, + "loss": 1.6886, + "step": 213 + }, + { + "epoch": 0.21982537236774524, + "grad_norm": 2.3583438396453857, + "learning_rate": 3.39041095890411e-06, + "loss": 1.7241, + "step": 214 + }, + { + "epoch": 0.22085259373394966, + "grad_norm": 2.345090866088867, + "learning_rate": 3.4075342465753424e-06, + "loss": 1.6194, + "step": 215 + }, + { + "epoch": 0.2218798151001541, + "grad_norm": 2.3996200561523438, + "learning_rate": 3.4246575342465754e-06, + "loss": 1.6889, + "step": 216 + }, + { + "epoch": 0.2229070364663585, + "grad_norm": 2.3552374839782715, + "learning_rate": 3.4417808219178085e-06, + "loss": 1.6713, + "step": 217 + }, + { + "epoch": 0.2239342578325629, + "grad_norm": 2.2653744220733643, + "learning_rate": 3.458904109589041e-06, + "loss": 1.673, + "step": 218 + }, + { + "epoch": 0.22496147919876733, + "grad_norm": 2.263288974761963, + "learning_rate": 3.476027397260274e-06, + "loss": 1.673, + "step": 219 + }, + { + "epoch": 0.22598870056497175, + "grad_norm": 2.3173880577087402, + "learning_rate": 3.4931506849315072e-06, + "loss": 1.6803, + "step": 220 + }, + { + "epoch": 0.22598870056497175, + "eval_loss": 1.9166476726531982, + "eval_runtime": 37.7659, + "eval_samples_per_second": 26.479, + "eval_steps_per_second": 1.112, + "step": 220 + }, + { + "epoch": 0.22701592193117617, + "grad_norm": 2.4005839824676514, + "learning_rate": 3.5102739726027403e-06, + "loss": 1.6809, + "step": 221 + }, + { + "epoch": 0.2280431432973806, + "grad_norm": 2.5032525062561035, + "learning_rate": 3.527397260273973e-06, + "loss": 1.6411, + "step": 222 + }, + { + "epoch": 0.229070364663585, + "grad_norm": 2.533583879470825, + "learning_rate": 3.5445205479452056e-06, + "loss": 1.6608, + "step": 223 + }, + { + "epoch": 0.2300975860297894, + "grad_norm": 2.3575572967529297, + "learning_rate": 3.5616438356164386e-06, + "loss": 1.7236, + "step": 224 + }, + { + "epoch": 0.23112480739599384, + "grad_norm": 2.3408942222595215, + "learning_rate": 3.5787671232876712e-06, + "loss": 1.7101, + "step": 225 + }, + { + "epoch": 0.23215202876219826, + "grad_norm": 2.3819596767425537, + "learning_rate": 3.5958904109589043e-06, + "loss": 1.6414, + "step": 226 + }, + { + "epoch": 0.23317925012840268, + "grad_norm": 2.2874338626861572, + "learning_rate": 3.6130136986301373e-06, + "loss": 1.6767, + "step": 227 + }, + { + "epoch": 0.23420647149460708, + "grad_norm": 2.3523948192596436, + "learning_rate": 3.6301369863013704e-06, + "loss": 1.6932, + "step": 228 + }, + { + "epoch": 0.2352336928608115, + "grad_norm": 2.476128578186035, + "learning_rate": 3.647260273972603e-06, + "loss": 1.7125, + "step": 229 + }, + { + "epoch": 0.23626091422701592, + "grad_norm": 2.3841593265533447, + "learning_rate": 3.664383561643836e-06, + "loss": 1.6469, + "step": 230 + }, + { + "epoch": 0.23626091422701592, + "eval_loss": 1.925522804260254, + "eval_runtime": 37.8581, + "eval_samples_per_second": 26.414, + "eval_steps_per_second": 1.109, + "step": 230 + }, + { + "epoch": 0.23728813559322035, + "grad_norm": 2.4401519298553467, + "learning_rate": 3.6815068493150687e-06, + "loss": 1.6726, + "step": 231 + }, + { + "epoch": 0.23831535695942477, + "grad_norm": 2.329674005508423, + "learning_rate": 3.6986301369863014e-06, + "loss": 1.6349, + "step": 232 + }, + { + "epoch": 0.23934257832562916, + "grad_norm": 2.654290199279785, + "learning_rate": 3.7157534246575344e-06, + "loss": 1.7309, + "step": 233 + }, + { + "epoch": 0.24036979969183359, + "grad_norm": 2.1599373817443848, + "learning_rate": 3.7328767123287675e-06, + "loss": 1.6437, + "step": 234 + }, + { + "epoch": 0.241397021058038, + "grad_norm": 2.6170263290405273, + "learning_rate": 3.7500000000000005e-06, + "loss": 1.6862, + "step": 235 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 2.2836415767669678, + "learning_rate": 3.767123287671233e-06, + "loss": 1.7073, + "step": 236 + }, + { + "epoch": 0.24345146379044685, + "grad_norm": 2.3001456260681152, + "learning_rate": 3.784246575342466e-06, + "loss": 1.6455, + "step": 237 + }, + { + "epoch": 0.24447868515665125, + "grad_norm": 2.4549131393432617, + "learning_rate": 3.8013698630136993e-06, + "loss": 1.6566, + "step": 238 + }, + { + "epoch": 0.24550590652285567, + "grad_norm": 2.4387831687927246, + "learning_rate": 3.818493150684932e-06, + "loss": 1.627, + "step": 239 + }, + { + "epoch": 0.2465331278890601, + "grad_norm": 2.136474847793579, + "learning_rate": 3.8356164383561645e-06, + "loss": 1.7286, + "step": 240 + }, + { + "epoch": 0.2465331278890601, + "eval_loss": 1.9188542366027832, + "eval_runtime": 37.9079, + "eval_samples_per_second": 26.38, + "eval_steps_per_second": 1.108, + "step": 240 + }, + { + "epoch": 0.24756034925526452, + "grad_norm": 2.430297374725342, + "learning_rate": 3.852739726027397e-06, + "loss": 1.7014, + "step": 241 + }, + { + "epoch": 0.24858757062146894, + "grad_norm": 2.302480697631836, + "learning_rate": 3.869863013698631e-06, + "loss": 1.6412, + "step": 242 + }, + { + "epoch": 0.24961479198767333, + "grad_norm": 2.2905426025390625, + "learning_rate": 3.886986301369863e-06, + "loss": 1.6445, + "step": 243 + }, + { + "epoch": 0.25064201335387776, + "grad_norm": 2.55385684967041, + "learning_rate": 3.904109589041096e-06, + "loss": 1.6265, + "step": 244 + }, + { + "epoch": 0.2516692347200822, + "grad_norm": 2.3053576946258545, + "learning_rate": 3.921232876712329e-06, + "loss": 1.6338, + "step": 245 + }, + { + "epoch": 0.2526964560862866, + "grad_norm": 2.4894707202911377, + "learning_rate": 3.938356164383562e-06, + "loss": 1.5966, + "step": 246 + }, + { + "epoch": 0.253723677452491, + "grad_norm": 2.6198930740356445, + "learning_rate": 3.9554794520547955e-06, + "loss": 1.6935, + "step": 247 + }, + { + "epoch": 0.25475089881869545, + "grad_norm": 2.3548998832702637, + "learning_rate": 3.972602739726027e-06, + "loss": 1.6227, + "step": 248 + }, + { + "epoch": 0.25577812018489987, + "grad_norm": 2.535460948944092, + "learning_rate": 3.989726027397261e-06, + "loss": 1.6699, + "step": 249 + }, + { + "epoch": 0.25680534155110424, + "grad_norm": 2.3680567741394043, + "learning_rate": 4.006849315068493e-06, + "loss": 1.6666, + "step": 250 + }, + { + "epoch": 0.25680534155110424, + "eval_loss": 1.9149967432022095, + "eval_runtime": 37.9423, + "eval_samples_per_second": 26.356, + "eval_steps_per_second": 1.107, + "step": 250 + }, + { + "epoch": 0.25783256291730866, + "grad_norm": 2.3914358615875244, + "learning_rate": 4.023972602739726e-06, + "loss": 1.6633, + "step": 251 + }, + { + "epoch": 0.2588597842835131, + "grad_norm": 2.4293038845062256, + "learning_rate": 4.0410958904109595e-06, + "loss": 1.6573, + "step": 252 + }, + { + "epoch": 0.2598870056497175, + "grad_norm": 2.3409900665283203, + "learning_rate": 4.058219178082192e-06, + "loss": 1.6815, + "step": 253 + }, + { + "epoch": 0.26091422701592193, + "grad_norm": 2.5807456970214844, + "learning_rate": 4.075342465753426e-06, + "loss": 1.6412, + "step": 254 + }, + { + "epoch": 0.26194144838212635, + "grad_norm": 2.317861795425415, + "learning_rate": 4.092465753424658e-06, + "loss": 1.6311, + "step": 255 + }, + { + "epoch": 0.2629686697483308, + "grad_norm": 2.231682777404785, + "learning_rate": 4.109589041095891e-06, + "loss": 1.7021, + "step": 256 + }, + { + "epoch": 0.2639958911145352, + "grad_norm": 2.7119040489196777, + "learning_rate": 4.1267123287671235e-06, + "loss": 1.6446, + "step": 257 + }, + { + "epoch": 0.2650231124807396, + "grad_norm": 2.2619071006774902, + "learning_rate": 4.143835616438356e-06, + "loss": 1.7101, + "step": 258 + }, + { + "epoch": 0.26605033384694404, + "grad_norm": 2.461945056915283, + "learning_rate": 4.16095890410959e-06, + "loss": 1.6901, + "step": 259 + }, + { + "epoch": 0.2670775552131484, + "grad_norm": 2.516348123550415, + "learning_rate": 4.178082191780822e-06, + "loss": 1.644, + "step": 260 + }, + { + "epoch": 0.2670775552131484, + "eval_loss": 1.904897928237915, + "eval_runtime": 37.93, + "eval_samples_per_second": 26.364, + "eval_steps_per_second": 1.107, + "step": 260 + }, + { + "epoch": 0.26810477657935283, + "grad_norm": 2.2311813831329346, + "learning_rate": 4.195205479452055e-06, + "loss": 1.6301, + "step": 261 + }, + { + "epoch": 0.26913199794555726, + "grad_norm": 2.605109214782715, + "learning_rate": 4.212328767123288e-06, + "loss": 1.6393, + "step": 262 + }, + { + "epoch": 0.2701592193117617, + "grad_norm": 2.6103134155273438, + "learning_rate": 4.229452054794521e-06, + "loss": 1.6468, + "step": 263 + }, + { + "epoch": 0.2711864406779661, + "grad_norm": 2.5538523197174072, + "learning_rate": 4.246575342465754e-06, + "loss": 1.6294, + "step": 264 + }, + { + "epoch": 0.2722136620441705, + "grad_norm": 2.5841965675354004, + "learning_rate": 4.263698630136986e-06, + "loss": 1.5837, + "step": 265 + }, + { + "epoch": 0.27324088341037495, + "grad_norm": 2.4604523181915283, + "learning_rate": 4.28082191780822e-06, + "loss": 1.7473, + "step": 266 + }, + { + "epoch": 0.27426810477657937, + "grad_norm": 2.6701765060424805, + "learning_rate": 4.297945205479452e-06, + "loss": 1.6539, + "step": 267 + }, + { + "epoch": 0.2752953261427838, + "grad_norm": 2.360069990158081, + "learning_rate": 4.315068493150685e-06, + "loss": 1.6385, + "step": 268 + }, + { + "epoch": 0.2763225475089882, + "grad_norm": 2.4922797679901123, + "learning_rate": 4.3321917808219185e-06, + "loss": 1.6586, + "step": 269 + }, + { + "epoch": 0.2773497688751926, + "grad_norm": 2.5443379878997803, + "learning_rate": 4.349315068493151e-06, + "loss": 1.6318, + "step": 270 + }, + { + "epoch": 0.2773497688751926, + "eval_loss": 1.9062609672546387, + "eval_runtime": 37.92, + "eval_samples_per_second": 26.371, + "eval_steps_per_second": 1.108, + "step": 270 + }, + { + "epoch": 0.278376990241397, + "grad_norm": 2.4064481258392334, + "learning_rate": 4.3664383561643846e-06, + "loss": 1.6178, + "step": 271 + }, + { + "epoch": 0.2794042116076014, + "grad_norm": 2.681637763977051, + "learning_rate": 4.383561643835616e-06, + "loss": 1.7071, + "step": 272 + }, + { + "epoch": 0.28043143297380585, + "grad_norm": 2.5805184841156006, + "learning_rate": 4.40068493150685e-06, + "loss": 1.6553, + "step": 273 + }, + { + "epoch": 0.2814586543400103, + "grad_norm": 2.4169883728027344, + "learning_rate": 4.4178082191780825e-06, + "loss": 1.6249, + "step": 274 + }, + { + "epoch": 0.2824858757062147, + "grad_norm": 2.557570457458496, + "learning_rate": 4.434931506849315e-06, + "loss": 1.5932, + "step": 275 + }, + { + "epoch": 0.2835130970724191, + "grad_norm": 2.3808107376098633, + "learning_rate": 4.4520547945205486e-06, + "loss": 1.6377, + "step": 276 + }, + { + "epoch": 0.28454031843862354, + "grad_norm": 2.402404546737671, + "learning_rate": 4.469178082191781e-06, + "loss": 1.6285, + "step": 277 + }, + { + "epoch": 0.28556753980482796, + "grad_norm": 2.4676456451416016, + "learning_rate": 4.486301369863014e-06, + "loss": 1.6838, + "step": 278 + }, + { + "epoch": 0.28659476117103233, + "grad_norm": 2.659083366394043, + "learning_rate": 4.503424657534247e-06, + "loss": 1.5872, + "step": 279 + }, + { + "epoch": 0.28762198253723675, + "grad_norm": 2.326733350753784, + "learning_rate": 4.52054794520548e-06, + "loss": 1.6142, + "step": 280 + }, + { + "epoch": 0.28762198253723675, + "eval_loss": 1.904759407043457, + "eval_runtime": 37.9651, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 280 + }, + { + "epoch": 0.2886492039034412, + "grad_norm": 2.299504518508911, + "learning_rate": 4.537671232876713e-06, + "loss": 1.6673, + "step": 281 + }, + { + "epoch": 0.2896764252696456, + "grad_norm": 2.4932825565338135, + "learning_rate": 4.554794520547945e-06, + "loss": 1.6586, + "step": 282 + }, + { + "epoch": 0.29070364663585, + "grad_norm": 2.2599050998687744, + "learning_rate": 4.571917808219179e-06, + "loss": 1.6568, + "step": 283 + }, + { + "epoch": 0.29173086800205444, + "grad_norm": 2.435098886489868, + "learning_rate": 4.589041095890411e-06, + "loss": 1.6113, + "step": 284 + }, + { + "epoch": 0.29275808936825887, + "grad_norm": 2.539246082305908, + "learning_rate": 4.606164383561644e-06, + "loss": 1.6306, + "step": 285 + }, + { + "epoch": 0.2937853107344633, + "grad_norm": 2.596198320388794, + "learning_rate": 4.6232876712328774e-06, + "loss": 1.6473, + "step": 286 + }, + { + "epoch": 0.2948125321006677, + "grad_norm": 2.349133014678955, + "learning_rate": 4.64041095890411e-06, + "loss": 1.6042, + "step": 287 + }, + { + "epoch": 0.29583975346687214, + "grad_norm": 2.3572661876678467, + "learning_rate": 4.657534246575343e-06, + "loss": 1.6635, + "step": 288 + }, + { + "epoch": 0.2968669748330765, + "grad_norm": 2.4210307598114014, + "learning_rate": 4.674657534246575e-06, + "loss": 1.624, + "step": 289 + }, + { + "epoch": 0.2978941961992809, + "grad_norm": 2.427403688430786, + "learning_rate": 4.691780821917809e-06, + "loss": 1.6824, + "step": 290 + }, + { + "epoch": 0.2978941961992809, + "eval_loss": 1.8888994455337524, + "eval_runtime": 37.9777, + "eval_samples_per_second": 26.331, + "eval_steps_per_second": 1.106, + "step": 290 + }, + { + "epoch": 0.29892141756548535, + "grad_norm": 2.534029722213745, + "learning_rate": 4.7089041095890414e-06, + "loss": 1.6456, + "step": 291 + }, + { + "epoch": 0.29994863893168977, + "grad_norm": 2.329305648803711, + "learning_rate": 4.726027397260274e-06, + "loss": 1.646, + "step": 292 + }, + { + "epoch": 0.3009758602978942, + "grad_norm": 2.561548948287964, + "learning_rate": 4.7431506849315075e-06, + "loss": 1.6511, + "step": 293 + }, + { + "epoch": 0.3020030816640986, + "grad_norm": 2.707712411880493, + "learning_rate": 4.76027397260274e-06, + "loss": 1.5984, + "step": 294 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 2.403089761734009, + "learning_rate": 4.777397260273973e-06, + "loss": 1.6164, + "step": 295 + }, + { + "epoch": 0.30405752439650746, + "grad_norm": 2.424910306930542, + "learning_rate": 4.7945205479452054e-06, + "loss": 1.6291, + "step": 296 + }, + { + "epoch": 0.3050847457627119, + "grad_norm": 2.3474302291870117, + "learning_rate": 4.811643835616439e-06, + "loss": 1.553, + "step": 297 + }, + { + "epoch": 0.3061119671289163, + "grad_norm": 2.2211227416992188, + "learning_rate": 4.8287671232876716e-06, + "loss": 1.6108, + "step": 298 + }, + { + "epoch": 0.3071391884951207, + "grad_norm": 2.495983600616455, + "learning_rate": 4.845890410958904e-06, + "loss": 1.6135, + "step": 299 + }, + { + "epoch": 0.3081664098613251, + "grad_norm": 2.525495767593384, + "learning_rate": 4.863013698630138e-06, + "loss": 1.6061, + "step": 300 + }, + { + "epoch": 0.3081664098613251, + "eval_loss": 1.8982168436050415, + "eval_runtime": 37.9848, + "eval_samples_per_second": 26.326, + "eval_steps_per_second": 1.106, + "step": 300 + }, + { + "epoch": 0.3091936312275295, + "grad_norm": 2.449064254760742, + "learning_rate": 4.88013698630137e-06, + "loss": 1.5827, + "step": 301 + }, + { + "epoch": 0.31022085259373394, + "grad_norm": 2.288095712661743, + "learning_rate": 4.897260273972603e-06, + "loss": 1.704, + "step": 302 + }, + { + "epoch": 0.31124807395993837, + "grad_norm": 2.5301458835601807, + "learning_rate": 4.914383561643836e-06, + "loss": 1.6342, + "step": 303 + }, + { + "epoch": 0.3122752953261428, + "grad_norm": 2.2793986797332764, + "learning_rate": 4.931506849315069e-06, + "loss": 1.6688, + "step": 304 + }, + { + "epoch": 0.3133025166923472, + "grad_norm": 2.38783597946167, + "learning_rate": 4.948630136986302e-06, + "loss": 1.7014, + "step": 305 + }, + { + "epoch": 0.31432973805855163, + "grad_norm": 2.5224838256835938, + "learning_rate": 4.965753424657534e-06, + "loss": 1.5812, + "step": 306 + }, + { + "epoch": 0.31535695942475606, + "grad_norm": 2.434432029724121, + "learning_rate": 4.982876712328768e-06, + "loss": 1.6186, + "step": 307 + }, + { + "epoch": 0.3163841807909605, + "grad_norm": 2.3708460330963135, + "learning_rate": 5e-06, + "loss": 1.5669, + "step": 308 + }, + { + "epoch": 0.31741140215716485, + "grad_norm": 2.5524990558624268, + "learning_rate": 4.998096688237534e-06, + "loss": 1.6688, + "step": 309 + }, + { + "epoch": 0.31843862352336927, + "grad_norm": 2.322441577911377, + "learning_rate": 4.996193376475067e-06, + "loss": 1.5571, + "step": 310 + }, + { + "epoch": 0.31843862352336927, + "eval_loss": 1.8914265632629395, + "eval_runtime": 37.6505, + "eval_samples_per_second": 26.56, + "eval_steps_per_second": 1.116, + "step": 310 + }, + { + "epoch": 0.3194658448895737, + "grad_norm": 2.5248687267303467, + "learning_rate": 4.9942900647126005e-06, + "loss": 1.5585, + "step": 311 + }, + { + "epoch": 0.3204930662557781, + "grad_norm": 2.2464914321899414, + "learning_rate": 4.992386752950134e-06, + "loss": 1.5921, + "step": 312 + }, + { + "epoch": 0.32152028762198254, + "grad_norm": 2.5355331897735596, + "learning_rate": 4.990483441187667e-06, + "loss": 1.6263, + "step": 313 + }, + { + "epoch": 0.32254750898818696, + "grad_norm": 2.5611143112182617, + "learning_rate": 4.9885801294252e-06, + "loss": 1.6267, + "step": 314 + }, + { + "epoch": 0.3235747303543914, + "grad_norm": 2.364356517791748, + "learning_rate": 4.986676817662733e-06, + "loss": 1.5847, + "step": 315 + }, + { + "epoch": 0.3246019517205958, + "grad_norm": 2.455165147781372, + "learning_rate": 4.9847735059002665e-06, + "loss": 1.661, + "step": 316 + }, + { + "epoch": 0.3256291730868002, + "grad_norm": 2.4312195777893066, + "learning_rate": 4.9828701941378e-06, + "loss": 1.5995, + "step": 317 + }, + { + "epoch": 0.3266563944530046, + "grad_norm": 2.4081499576568604, + "learning_rate": 4.980966882375333e-06, + "loss": 1.5756, + "step": 318 + }, + { + "epoch": 0.327683615819209, + "grad_norm": 2.2608511447906494, + "learning_rate": 4.979063570612867e-06, + "loss": 1.543, + "step": 319 + }, + { + "epoch": 0.32871083718541344, + "grad_norm": 2.6551144123077393, + "learning_rate": 4.9771602588504e-06, + "loss": 1.6467, + "step": 320 + }, + { + "epoch": 0.32871083718541344, + "eval_loss": 1.89058518409729, + "eval_runtime": 37.8192, + "eval_samples_per_second": 26.442, + "eval_steps_per_second": 1.111, + "step": 320 + }, + { + "epoch": 0.32973805855161786, + "grad_norm": 2.3897273540496826, + "learning_rate": 4.975256947087933e-06, + "loss": 1.5607, + "step": 321 + }, + { + "epoch": 0.3307652799178223, + "grad_norm": 2.3869469165802, + "learning_rate": 4.973353635325467e-06, + "loss": 1.6079, + "step": 322 + }, + { + "epoch": 0.3317925012840267, + "grad_norm": 2.159193515777588, + "learning_rate": 4.971450323563e-06, + "loss": 1.5475, + "step": 323 + }, + { + "epoch": 0.33281972265023113, + "grad_norm": 2.3069653511047363, + "learning_rate": 4.969547011800533e-06, + "loss": 1.616, + "step": 324 + }, + { + "epoch": 0.33384694401643555, + "grad_norm": 2.3131422996520996, + "learning_rate": 4.967643700038067e-06, + "loss": 1.6106, + "step": 325 + }, + { + "epoch": 0.33487416538264, + "grad_norm": 2.4398200511932373, + "learning_rate": 4.9657403882756e-06, + "loss": 1.5779, + "step": 326 + }, + { + "epoch": 0.3359013867488444, + "grad_norm": 2.278024435043335, + "learning_rate": 4.9638370765131335e-06, + "loss": 1.5724, + "step": 327 + }, + { + "epoch": 0.33692860811504877, + "grad_norm": 2.4989173412323, + "learning_rate": 4.961933764750667e-06, + "loss": 1.6337, + "step": 328 + }, + { + "epoch": 0.3379558294812532, + "grad_norm": 2.46319580078125, + "learning_rate": 4.9600304529882e-06, + "loss": 1.6519, + "step": 329 + }, + { + "epoch": 0.3389830508474576, + "grad_norm": 2.2703049182891846, + "learning_rate": 4.9581271412257335e-06, + "loss": 1.6081, + "step": 330 + }, + { + "epoch": 0.3389830508474576, + "eval_loss": 1.8755545616149902, + "eval_runtime": 37.9058, + "eval_samples_per_second": 26.381, + "eval_steps_per_second": 1.108, + "step": 330 + }, + { + "epoch": 0.34001027221366203, + "grad_norm": 2.502196788787842, + "learning_rate": 4.956223829463266e-06, + "loss": 1.5221, + "step": 331 + }, + { + "epoch": 0.34103749357986646, + "grad_norm": 2.284280776977539, + "learning_rate": 4.954320517700799e-06, + "loss": 1.6268, + "step": 332 + }, + { + "epoch": 0.3420647149460709, + "grad_norm": 2.3551881313323975, + "learning_rate": 4.952417205938333e-06, + "loss": 1.6078, + "step": 333 + }, + { + "epoch": 0.3430919363122753, + "grad_norm": 2.343780755996704, + "learning_rate": 4.950513894175866e-06, + "loss": 1.591, + "step": 334 + }, + { + "epoch": 0.3441191576784797, + "grad_norm": 2.3524727821350098, + "learning_rate": 4.9486105824133995e-06, + "loss": 1.605, + "step": 335 + }, + { + "epoch": 0.34514637904468415, + "grad_norm": 2.349896192550659, + "learning_rate": 4.946707270650933e-06, + "loss": 1.6106, + "step": 336 + }, + { + "epoch": 0.34617360041088857, + "grad_norm": 2.5154407024383545, + "learning_rate": 4.944803958888466e-06, + "loss": 1.6429, + "step": 337 + }, + { + "epoch": 0.34720082177709294, + "grad_norm": 2.2692947387695312, + "learning_rate": 4.942900647126e-06, + "loss": 1.5865, + "step": 338 + }, + { + "epoch": 0.34822804314329736, + "grad_norm": 2.2672805786132812, + "learning_rate": 4.940997335363533e-06, + "loss": 1.5464, + "step": 339 + }, + { + "epoch": 0.3492552645095018, + "grad_norm": 2.5147039890289307, + "learning_rate": 4.939094023601066e-06, + "loss": 1.5837, + "step": 340 + }, + { + "epoch": 0.3492552645095018, + "eval_loss": 1.8672455549240112, + "eval_runtime": 37.9644, + "eval_samples_per_second": 26.34, + "eval_steps_per_second": 1.106, + "step": 340 + }, + { + "epoch": 0.3502824858757062, + "grad_norm": 2.3416433334350586, + "learning_rate": 4.9371907118386e-06, + "loss": 1.5937, + "step": 341 + }, + { + "epoch": 0.35130970724191063, + "grad_norm": 2.3044960498809814, + "learning_rate": 4.935287400076133e-06, + "loss": 1.6132, + "step": 342 + }, + { + "epoch": 0.35233692860811505, + "grad_norm": 2.4360430240631104, + "learning_rate": 4.933384088313666e-06, + "loss": 1.5731, + "step": 343 + }, + { + "epoch": 0.3533641499743195, + "grad_norm": 2.392050266265869, + "learning_rate": 4.9314807765512e-06, + "loss": 1.5098, + "step": 344 + }, + { + "epoch": 0.3543913713405239, + "grad_norm": 2.3815808296203613, + "learning_rate": 4.929577464788733e-06, + "loss": 1.5827, + "step": 345 + }, + { + "epoch": 0.3554185927067283, + "grad_norm": 2.423077344894409, + "learning_rate": 4.9276741530262665e-06, + "loss": 1.5857, + "step": 346 + }, + { + "epoch": 0.35644581407293274, + "grad_norm": 2.5535898208618164, + "learning_rate": 4.9257708412638e-06, + "loss": 1.5884, + "step": 347 + }, + { + "epoch": 0.3574730354391371, + "grad_norm": 2.2665445804595947, + "learning_rate": 4.923867529501332e-06, + "loss": 1.5561, + "step": 348 + }, + { + "epoch": 0.35850025680534153, + "grad_norm": 2.3357481956481934, + "learning_rate": 4.921964217738866e-06, + "loss": 1.6444, + "step": 349 + }, + { + "epoch": 0.35952747817154596, + "grad_norm": 2.4537885189056396, + "learning_rate": 4.920060905976399e-06, + "loss": 1.6262, + "step": 350 + }, + { + "epoch": 0.35952747817154596, + "eval_loss": 1.8621584177017212, + "eval_runtime": 38.0027, + "eval_samples_per_second": 26.314, + "eval_steps_per_second": 1.105, + "step": 350 + }, + { + "epoch": 0.3605546995377504, + "grad_norm": 2.379754066467285, + "learning_rate": 4.9181575942139325e-06, + "loss": 1.6227, + "step": 351 + }, + { + "epoch": 0.3615819209039548, + "grad_norm": 2.432905435562134, + "learning_rate": 4.916254282451466e-06, + "loss": 1.6554, + "step": 352 + }, + { + "epoch": 0.3626091422701592, + "grad_norm": 2.34236216545105, + "learning_rate": 4.914350970688999e-06, + "loss": 1.5613, + "step": 353 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.3019444942474365, + "learning_rate": 4.9124476589265325e-06, + "loss": 1.5535, + "step": 354 + }, + { + "epoch": 0.36466358500256807, + "grad_norm": 2.3354644775390625, + "learning_rate": 4.910544347164066e-06, + "loss": 1.6384, + "step": 355 + }, + { + "epoch": 0.3656908063687725, + "grad_norm": 2.5154788494110107, + "learning_rate": 4.908641035401599e-06, + "loss": 1.6986, + "step": 356 + }, + { + "epoch": 0.3667180277349769, + "grad_norm": 2.2707178592681885, + "learning_rate": 4.906737723639133e-06, + "loss": 1.5705, + "step": 357 + }, + { + "epoch": 0.3677452491011813, + "grad_norm": 2.3058297634124756, + "learning_rate": 4.904834411876665e-06, + "loss": 1.6033, + "step": 358 + }, + { + "epoch": 0.3687724704673857, + "grad_norm": 2.3923370838165283, + "learning_rate": 4.902931100114199e-06, + "loss": 1.5333, + "step": 359 + }, + { + "epoch": 0.3697996918335901, + "grad_norm": 2.4023196697235107, + "learning_rate": 4.901027788351733e-06, + "loss": 1.5775, + "step": 360 + }, + { + "epoch": 0.3697996918335901, + "eval_loss": 1.867856740951538, + "eval_runtime": 38.0125, + "eval_samples_per_second": 26.307, + "eval_steps_per_second": 1.105, + "step": 360 + }, + { + "epoch": 0.37082691319979455, + "grad_norm": 2.4903581142425537, + "learning_rate": 4.899124476589266e-06, + "loss": 1.5491, + "step": 361 + }, + { + "epoch": 0.371854134565999, + "grad_norm": 2.556683301925659, + "learning_rate": 4.8972211648267994e-06, + "loss": 1.5967, + "step": 362 + }, + { + "epoch": 0.3728813559322034, + "grad_norm": 2.3017377853393555, + "learning_rate": 4.895317853064333e-06, + "loss": 1.5661, + "step": 363 + }, + { + "epoch": 0.3739085772984078, + "grad_norm": 2.286191940307617, + "learning_rate": 4.893414541301866e-06, + "loss": 1.5679, + "step": 364 + }, + { + "epoch": 0.37493579866461224, + "grad_norm": 2.260812520980835, + "learning_rate": 4.891511229539399e-06, + "loss": 1.5965, + "step": 365 + }, + { + "epoch": 0.37596302003081666, + "grad_norm": 2.4569408893585205, + "learning_rate": 4.889607917776932e-06, + "loss": 1.5923, + "step": 366 + }, + { + "epoch": 0.37699024139702103, + "grad_norm": 2.571552038192749, + "learning_rate": 4.887704606014465e-06, + "loss": 1.6021, + "step": 367 + }, + { + "epoch": 0.37801746276322545, + "grad_norm": 2.287754774093628, + "learning_rate": 4.885801294251999e-06, + "loss": 1.5304, + "step": 368 + }, + { + "epoch": 0.3790446841294299, + "grad_norm": 2.317784070968628, + "learning_rate": 4.883897982489532e-06, + "loss": 1.6233, + "step": 369 + }, + { + "epoch": 0.3800719054956343, + "grad_norm": 2.290601968765259, + "learning_rate": 4.8819946707270655e-06, + "loss": 1.624, + "step": 370 + }, + { + "epoch": 0.3800719054956343, + "eval_loss": 1.8604528903961182, + "eval_runtime": 38.0176, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 370 + }, + { + "epoch": 0.3810991268618387, + "grad_norm": 2.28519868850708, + "learning_rate": 4.880091358964599e-06, + "loss": 1.5286, + "step": 371 + }, + { + "epoch": 0.38212634822804314, + "grad_norm": 2.271810293197632, + "learning_rate": 4.878188047202132e-06, + "loss": 1.6879, + "step": 372 + }, + { + "epoch": 0.38315356959424757, + "grad_norm": 2.391848564147949, + "learning_rate": 4.876284735439666e-06, + "loss": 1.6458, + "step": 373 + }, + { + "epoch": 0.384180790960452, + "grad_norm": 2.3787031173706055, + "learning_rate": 4.874381423677199e-06, + "loss": 1.5738, + "step": 374 + }, + { + "epoch": 0.3852080123266564, + "grad_norm": 2.404219150543213, + "learning_rate": 4.8724781119147315e-06, + "loss": 1.6184, + "step": 375 + }, + { + "epoch": 0.38623523369286084, + "grad_norm": 2.372661590576172, + "learning_rate": 4.870574800152265e-06, + "loss": 1.5316, + "step": 376 + }, + { + "epoch": 0.3872624550590652, + "grad_norm": 2.4397103786468506, + "learning_rate": 4.868671488389798e-06, + "loss": 1.5605, + "step": 377 + }, + { + "epoch": 0.3882896764252696, + "grad_norm": 2.371006727218628, + "learning_rate": 4.8667681766273316e-06, + "loss": 1.5184, + "step": 378 + }, + { + "epoch": 0.38931689779147405, + "grad_norm": 2.2621910572052, + "learning_rate": 4.864864864864866e-06, + "loss": 1.5788, + "step": 379 + }, + { + "epoch": 0.39034411915767847, + "grad_norm": 2.2888615131378174, + "learning_rate": 4.862961553102399e-06, + "loss": 1.6244, + "step": 380 + }, + { + "epoch": 0.39034411915767847, + "eval_loss": 1.8586912155151367, + "eval_runtime": 37.9582, + "eval_samples_per_second": 26.345, + "eval_steps_per_second": 1.106, + "step": 380 + }, + { + "epoch": 0.3913713405238829, + "grad_norm": 2.4439220428466797, + "learning_rate": 4.8610582413399325e-06, + "loss": 1.5749, + "step": 381 + }, + { + "epoch": 0.3923985618900873, + "grad_norm": 2.2423763275146484, + "learning_rate": 4.859154929577465e-06, + "loss": 1.5738, + "step": 382 + }, + { + "epoch": 0.39342578325629174, + "grad_norm": 2.267467737197876, + "learning_rate": 4.857251617814998e-06, + "loss": 1.6212, + "step": 383 + }, + { + "epoch": 0.39445300462249616, + "grad_norm": 2.3680620193481445, + "learning_rate": 4.855348306052532e-06, + "loss": 1.6129, + "step": 384 + }, + { + "epoch": 0.3954802259887006, + "grad_norm": 2.3782172203063965, + "learning_rate": 4.853444994290065e-06, + "loss": 1.5605, + "step": 385 + }, + { + "epoch": 0.396507447354905, + "grad_norm": 2.5745720863342285, + "learning_rate": 4.8515416825275985e-06, + "loss": 1.634, + "step": 386 + }, + { + "epoch": 0.3975346687211094, + "grad_norm": 2.38212513923645, + "learning_rate": 4.849638370765132e-06, + "loss": 1.5671, + "step": 387 + }, + { + "epoch": 0.3985618900873138, + "grad_norm": 2.3395450115203857, + "learning_rate": 4.847735059002665e-06, + "loss": 1.5088, + "step": 388 + }, + { + "epoch": 0.3995891114535182, + "grad_norm": 2.474440574645996, + "learning_rate": 4.8458317472401985e-06, + "loss": 1.5997, + "step": 389 + }, + { + "epoch": 0.40061633281972264, + "grad_norm": 2.429133176803589, + "learning_rate": 4.843928435477732e-06, + "loss": 1.5611, + "step": 390 + }, + { + "epoch": 0.40061633281972264, + "eval_loss": 1.8613603115081787, + "eval_runtime": 37.9862, + "eval_samples_per_second": 26.325, + "eval_steps_per_second": 1.106, + "step": 390 + }, + { + "epoch": 0.40164355418592707, + "grad_norm": 2.403207778930664, + "learning_rate": 4.842025123715265e-06, + "loss": 1.5976, + "step": 391 + }, + { + "epoch": 0.4026707755521315, + "grad_norm": 2.383018732070923, + "learning_rate": 4.840121811952798e-06, + "loss": 1.5791, + "step": 392 + }, + { + "epoch": 0.4036979969183359, + "grad_norm": 2.5303306579589844, + "learning_rate": 4.838218500190331e-06, + "loss": 1.6035, + "step": 393 + }, + { + "epoch": 0.40472521828454033, + "grad_norm": 2.391075372695923, + "learning_rate": 4.8363151884278645e-06, + "loss": 1.5795, + "step": 394 + }, + { + "epoch": 0.40575243965074476, + "grad_norm": 2.396934747695923, + "learning_rate": 4.834411876665398e-06, + "loss": 1.5965, + "step": 395 + }, + { + "epoch": 0.4067796610169492, + "grad_norm": 2.4631423950195312, + "learning_rate": 4.832508564902931e-06, + "loss": 1.5372, + "step": 396 + }, + { + "epoch": 0.40780688238315355, + "grad_norm": 2.3457326889038086, + "learning_rate": 4.830605253140465e-06, + "loss": 1.5729, + "step": 397 + }, + { + "epoch": 0.40883410374935797, + "grad_norm": 2.3763277530670166, + "learning_rate": 4.828701941377998e-06, + "loss": 1.5309, + "step": 398 + }, + { + "epoch": 0.4098613251155624, + "grad_norm": 2.442439556121826, + "learning_rate": 4.826798629615531e-06, + "loss": 1.5465, + "step": 399 + }, + { + "epoch": 0.4108885464817668, + "grad_norm": 2.3464581966400146, + "learning_rate": 4.824895317853065e-06, + "loss": 1.5665, + "step": 400 + }, + { + "epoch": 0.4108885464817668, + "eval_loss": 1.8671027421951294, + "eval_runtime": 38.0166, + "eval_samples_per_second": 26.304, + "eval_steps_per_second": 1.105, + "step": 400 + }, + { + "epoch": 0.41191576784797124, + "grad_norm": 2.5441887378692627, + "learning_rate": 4.822992006090598e-06, + "loss": 1.5232, + "step": 401 + }, + { + "epoch": 0.41294298921417566, + "grad_norm": 2.4234976768493652, + "learning_rate": 4.821088694328131e-06, + "loss": 1.5569, + "step": 402 + }, + { + "epoch": 0.4139702105803801, + "grad_norm": 2.3603928089141846, + "learning_rate": 4.819185382565665e-06, + "loss": 1.612, + "step": 403 + }, + { + "epoch": 0.4149974319465845, + "grad_norm": 2.337862491607666, + "learning_rate": 4.817282070803198e-06, + "loss": 1.5882, + "step": 404 + }, + { + "epoch": 0.41602465331278893, + "grad_norm": 2.3805742263793945, + "learning_rate": 4.8153787590407315e-06, + "loss": 1.5204, + "step": 405 + }, + { + "epoch": 0.41705187467899335, + "grad_norm": 2.3335204124450684, + "learning_rate": 4.813475447278265e-06, + "loss": 1.6041, + "step": 406 + }, + { + "epoch": 0.4180790960451977, + "grad_norm": 2.406137466430664, + "learning_rate": 4.811572135515798e-06, + "loss": 1.5348, + "step": 407 + }, + { + "epoch": 0.41910631741140214, + "grad_norm": 2.2823917865753174, + "learning_rate": 4.809668823753332e-06, + "loss": 1.5351, + "step": 408 + }, + { + "epoch": 0.42013353877760656, + "grad_norm": 2.3577091693878174, + "learning_rate": 4.807765511990864e-06, + "loss": 1.5733, + "step": 409 + }, + { + "epoch": 0.421160760143811, + "grad_norm": 2.536691665649414, + "learning_rate": 4.8058622002283975e-06, + "loss": 1.6153, + "step": 410 + }, + { + "epoch": 0.421160760143811, + "eval_loss": 1.8613901138305664, + "eval_runtime": 37.6689, + "eval_samples_per_second": 26.547, + "eval_steps_per_second": 1.115, + "step": 410 + }, + { + "epoch": 0.4221879815100154, + "grad_norm": 2.4020118713378906, + "learning_rate": 4.803958888465931e-06, + "loss": 1.5722, + "step": 411 + }, + { + "epoch": 0.42321520287621983, + "grad_norm": 2.5284810066223145, + "learning_rate": 4.802055576703464e-06, + "loss": 1.6546, + "step": 412 + }, + { + "epoch": 0.42424242424242425, + "grad_norm": 2.270094156265259, + "learning_rate": 4.8001522649409976e-06, + "loss": 1.5345, + "step": 413 + }, + { + "epoch": 0.4252696456086287, + "grad_norm": 2.241417407989502, + "learning_rate": 4.798248953178531e-06, + "loss": 1.5167, + "step": 414 + }, + { + "epoch": 0.4262968669748331, + "grad_norm": 2.3733017444610596, + "learning_rate": 4.796345641416064e-06, + "loss": 1.5532, + "step": 415 + }, + { + "epoch": 0.42732408834103747, + "grad_norm": 2.452587842941284, + "learning_rate": 4.794442329653598e-06, + "loss": 1.6136, + "step": 416 + }, + { + "epoch": 0.4283513097072419, + "grad_norm": 2.417217969894409, + "learning_rate": 4.792539017891131e-06, + "loss": 1.5398, + "step": 417 + }, + { + "epoch": 0.4293785310734463, + "grad_norm": 2.4502453804016113, + "learning_rate": 4.790635706128664e-06, + "loss": 1.5741, + "step": 418 + }, + { + "epoch": 0.43040575243965074, + "grad_norm": 2.2996811866760254, + "learning_rate": 4.788732394366197e-06, + "loss": 1.511, + "step": 419 + }, + { + "epoch": 0.43143297380585516, + "grad_norm": 2.3573555946350098, + "learning_rate": 4.786829082603731e-06, + "loss": 1.5117, + "step": 420 + }, + { + "epoch": 0.43143297380585516, + "eval_loss": 1.8690649271011353, + "eval_runtime": 37.7813, + "eval_samples_per_second": 26.468, + "eval_steps_per_second": 1.112, + "step": 420 + }, + { + "epoch": 0.4324601951720596, + "grad_norm": 2.37233567237854, + "learning_rate": 4.7849257708412645e-06, + "loss": 1.5873, + "step": 421 + }, + { + "epoch": 0.433487416538264, + "grad_norm": 2.146909713745117, + "learning_rate": 4.783022459078798e-06, + "loss": 1.5291, + "step": 422 + }, + { + "epoch": 0.4345146379044684, + "grad_norm": 2.5237205028533936, + "learning_rate": 4.781119147316331e-06, + "loss": 1.5545, + "step": 423 + }, + { + "epoch": 0.43554185927067285, + "grad_norm": 2.356842279434204, + "learning_rate": 4.7792158355538645e-06, + "loss": 1.5635, + "step": 424 + }, + { + "epoch": 0.43656908063687727, + "grad_norm": 2.1963610649108887, + "learning_rate": 4.777312523791398e-06, + "loss": 1.5355, + "step": 425 + }, + { + "epoch": 0.43759630200308164, + "grad_norm": 2.6425886154174805, + "learning_rate": 4.77540921202893e-06, + "loss": 1.5803, + "step": 426 + }, + { + "epoch": 0.43862352336928606, + "grad_norm": 2.4072272777557373, + "learning_rate": 4.773505900266464e-06, + "loss": 1.5649, + "step": 427 + }, + { + "epoch": 0.4396507447354905, + "grad_norm": 2.592374801635742, + "learning_rate": 4.771602588503997e-06, + "loss": 1.5693, + "step": 428 + }, + { + "epoch": 0.4406779661016949, + "grad_norm": 2.4534225463867188, + "learning_rate": 4.7696992767415305e-06, + "loss": 1.591, + "step": 429 + }, + { + "epoch": 0.44170518746789933, + "grad_norm": 2.434457778930664, + "learning_rate": 4.767795964979064e-06, + "loss": 1.5991, + "step": 430 + }, + { + "epoch": 0.44170518746789933, + "eval_loss": 1.8584965467453003, + "eval_runtime": 37.8854, + "eval_samples_per_second": 26.395, + "eval_steps_per_second": 1.109, + "step": 430 + }, + { + "epoch": 0.44273240883410375, + "grad_norm": 2.3150746822357178, + "learning_rate": 4.765892653216597e-06, + "loss": 1.6032, + "step": 431 + }, + { + "epoch": 0.4437596302003082, + "grad_norm": 2.3040223121643066, + "learning_rate": 4.763989341454131e-06, + "loss": 1.5115, + "step": 432 + }, + { + "epoch": 0.4447868515665126, + "grad_norm": 2.413007974624634, + "learning_rate": 4.762086029691664e-06, + "loss": 1.5305, + "step": 433 + }, + { + "epoch": 0.445814072932717, + "grad_norm": 2.526289463043213, + "learning_rate": 4.760182717929197e-06, + "loss": 1.564, + "step": 434 + }, + { + "epoch": 0.44684129429892144, + "grad_norm": 2.4122776985168457, + "learning_rate": 4.758279406166731e-06, + "loss": 1.4918, + "step": 435 + }, + { + "epoch": 0.4478685156651258, + "grad_norm": 2.474274158477783, + "learning_rate": 4.756376094404263e-06, + "loss": 1.5583, + "step": 436 + }, + { + "epoch": 0.44889573703133023, + "grad_norm": 2.5322892665863037, + "learning_rate": 4.7544727826417966e-06, + "loss": 1.6081, + "step": 437 + }, + { + "epoch": 0.44992295839753466, + "grad_norm": 2.391660690307617, + "learning_rate": 4.75256947087933e-06, + "loss": 1.5628, + "step": 438 + }, + { + "epoch": 0.4509501797637391, + "grad_norm": 2.426671028137207, + "learning_rate": 4.750666159116863e-06, + "loss": 1.531, + "step": 439 + }, + { + "epoch": 0.4519774011299435, + "grad_norm": 2.7095694541931152, + "learning_rate": 4.7487628473543975e-06, + "loss": 1.4811, + "step": 440 + }, + { + "epoch": 0.4519774011299435, + "eval_loss": 1.8348888158798218, + "eval_runtime": 37.9114, + "eval_samples_per_second": 26.377, + "eval_steps_per_second": 1.108, + "step": 440 + }, + { + "epoch": 0.4530046224961479, + "grad_norm": 2.5425240993499756, + "learning_rate": 4.746859535591931e-06, + "loss": 1.5453, + "step": 441 + }, + { + "epoch": 0.45403184386235235, + "grad_norm": 2.6848669052124023, + "learning_rate": 4.744956223829464e-06, + "loss": 1.526, + "step": 442 + }, + { + "epoch": 0.45505906522855677, + "grad_norm": 2.634068489074707, + "learning_rate": 4.743052912066997e-06, + "loss": 1.5748, + "step": 443 + }, + { + "epoch": 0.4560862865947612, + "grad_norm": 2.492030382156372, + "learning_rate": 4.74114960030453e-06, + "loss": 1.5858, + "step": 444 + }, + { + "epoch": 0.4571135079609656, + "grad_norm": 2.787740707397461, + "learning_rate": 4.7392462885420635e-06, + "loss": 1.4774, + "step": 445 + }, + { + "epoch": 0.45814072932717, + "grad_norm": 2.395914077758789, + "learning_rate": 4.737342976779597e-06, + "loss": 1.558, + "step": 446 + }, + { + "epoch": 0.4591679506933744, + "grad_norm": 2.6642563343048096, + "learning_rate": 4.73543966501713e-06, + "loss": 1.5558, + "step": 447 + }, + { + "epoch": 0.4601951720595788, + "grad_norm": 2.619943857192993, + "learning_rate": 4.7335363532546636e-06, + "loss": 1.5442, + "step": 448 + }, + { + "epoch": 0.46122239342578325, + "grad_norm": 2.355133056640625, + "learning_rate": 4.731633041492197e-06, + "loss": 1.5613, + "step": 449 + }, + { + "epoch": 0.4622496147919877, + "grad_norm": 2.6572225093841553, + "learning_rate": 4.72972972972973e-06, + "loss": 1.5951, + "step": 450 + }, + { + "epoch": 0.4622496147919877, + "eval_loss": 1.83199143409729, + "eval_runtime": 37.8996, + "eval_samples_per_second": 26.385, + "eval_steps_per_second": 1.108, + "step": 450 + }, + { + "epoch": 0.4632768361581921, + "grad_norm": 2.692873001098633, + "learning_rate": 4.727826417967264e-06, + "loss": 1.5531, + "step": 451 + }, + { + "epoch": 0.4643040575243965, + "grad_norm": 2.3903489112854004, + "learning_rate": 4.725923106204797e-06, + "loss": 1.5411, + "step": 452 + }, + { + "epoch": 0.46533127889060094, + "grad_norm": 2.453263521194458, + "learning_rate": 4.72401979444233e-06, + "loss": 1.5632, + "step": 453 + }, + { + "epoch": 0.46635850025680536, + "grad_norm": 2.5637760162353516, + "learning_rate": 4.722116482679863e-06, + "loss": 1.5629, + "step": 454 + }, + { + "epoch": 0.46738572162300973, + "grad_norm": 2.387033700942993, + "learning_rate": 4.720213170917396e-06, + "loss": 1.5786, + "step": 455 + }, + { + "epoch": 0.46841294298921415, + "grad_norm": 2.4563071727752686, + "learning_rate": 4.71830985915493e-06, + "loss": 1.5321, + "step": 456 + }, + { + "epoch": 0.4694401643554186, + "grad_norm": 2.607451915740967, + "learning_rate": 4.716406547392463e-06, + "loss": 1.5958, + "step": 457 + }, + { + "epoch": 0.470467385721623, + "grad_norm": 2.4846854209899902, + "learning_rate": 4.714503235629996e-06, + "loss": 1.5205, + "step": 458 + }, + { + "epoch": 0.4714946070878274, + "grad_norm": 2.377629041671753, + "learning_rate": 4.71259992386753e-06, + "loss": 1.5403, + "step": 459 + }, + { + "epoch": 0.47252182845403184, + "grad_norm": 2.4470160007476807, + "learning_rate": 4.710696612105063e-06, + "loss": 1.5456, + "step": 460 + }, + { + "epoch": 0.47252182845403184, + "eval_loss": 1.83836829662323, + "eval_runtime": 37.88, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 460 + }, + { + "epoch": 0.47354904982023627, + "grad_norm": 2.383418560028076, + "learning_rate": 4.708793300342596e-06, + "loss": 1.5468, + "step": 461 + }, + { + "epoch": 0.4745762711864407, + "grad_norm": 2.3672895431518555, + "learning_rate": 4.70688998858013e-06, + "loss": 1.5313, + "step": 462 + }, + { + "epoch": 0.4756034925526451, + "grad_norm": 2.415717124938965, + "learning_rate": 4.704986676817663e-06, + "loss": 1.5645, + "step": 463 + }, + { + "epoch": 0.47663071391884954, + "grad_norm": 2.4690067768096924, + "learning_rate": 4.7030833650551965e-06, + "loss": 1.5285, + "step": 464 + }, + { + "epoch": 0.4776579352850539, + "grad_norm": 2.3845672607421875, + "learning_rate": 4.70118005329273e-06, + "loss": 1.596, + "step": 465 + }, + { + "epoch": 0.4786851566512583, + "grad_norm": 2.2801802158355713, + "learning_rate": 4.699276741530263e-06, + "loss": 1.5777, + "step": 466 + }, + { + "epoch": 0.47971237801746275, + "grad_norm": 2.4668991565704346, + "learning_rate": 4.697373429767797e-06, + "loss": 1.5566, + "step": 467 + }, + { + "epoch": 0.48073959938366717, + "grad_norm": 2.446078062057495, + "learning_rate": 4.69547011800533e-06, + "loss": 1.5664, + "step": 468 + }, + { + "epoch": 0.4817668207498716, + "grad_norm": 2.2310681343078613, + "learning_rate": 4.693566806242863e-06, + "loss": 1.5425, + "step": 469 + }, + { + "epoch": 0.482794042116076, + "grad_norm": 2.3938229084014893, + "learning_rate": 4.691663494480397e-06, + "loss": 1.576, + "step": 470 + }, + { + "epoch": 0.482794042116076, + "eval_loss": 1.83778977394104, + "eval_runtime": 37.9276, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 470 + }, + { + "epoch": 0.48382126348228044, + "grad_norm": 2.23827862739563, + "learning_rate": 4.689760182717929e-06, + "loss": 1.4763, + "step": 471 + }, + { + "epoch": 0.48484848484848486, + "grad_norm": 2.3148930072784424, + "learning_rate": 4.6878568709554626e-06, + "loss": 1.5386, + "step": 472 + }, + { + "epoch": 0.4858757062146893, + "grad_norm": 2.407036781311035, + "learning_rate": 4.685953559192996e-06, + "loss": 1.5576, + "step": 473 + }, + { + "epoch": 0.4869029275808937, + "grad_norm": 2.4675374031066895, + "learning_rate": 4.684050247430529e-06, + "loss": 1.4931, + "step": 474 + }, + { + "epoch": 0.4879301489470981, + "grad_norm": 2.4593775272369385, + "learning_rate": 4.682146935668063e-06, + "loss": 1.5348, + "step": 475 + }, + { + "epoch": 0.4889573703133025, + "grad_norm": 2.4169375896453857, + "learning_rate": 4.680243623905596e-06, + "loss": 1.5614, + "step": 476 + }, + { + "epoch": 0.4899845916795069, + "grad_norm": 2.6177303791046143, + "learning_rate": 4.678340312143129e-06, + "loss": 1.5138, + "step": 477 + }, + { + "epoch": 0.49101181304571134, + "grad_norm": 2.6418113708496094, + "learning_rate": 4.676437000380663e-06, + "loss": 1.6377, + "step": 478 + }, + { + "epoch": 0.49203903441191577, + "grad_norm": 2.338491439819336, + "learning_rate": 4.674533688618196e-06, + "loss": 1.5682, + "step": 479 + }, + { + "epoch": 0.4930662557781202, + "grad_norm": 2.5901622772216797, + "learning_rate": 4.6726303768557295e-06, + "loss": 1.4899, + "step": 480 + }, + { + "epoch": 0.4930662557781202, + "eval_loss": 1.8238886594772339, + "eval_runtime": 37.9605, + "eval_samples_per_second": 26.343, + "eval_steps_per_second": 1.106, + "step": 480 + }, + { + "epoch": 0.4940934771443246, + "grad_norm": 2.470827102661133, + "learning_rate": 4.670727065093263e-06, + "loss": 1.5016, + "step": 481 + }, + { + "epoch": 0.49512069851052903, + "grad_norm": 2.371201992034912, + "learning_rate": 4.668823753330796e-06, + "loss": 1.5399, + "step": 482 + }, + { + "epoch": 0.49614791987673346, + "grad_norm": 2.532802104949951, + "learning_rate": 4.6669204415683295e-06, + "loss": 1.5985, + "step": 483 + }, + { + "epoch": 0.4971751412429379, + "grad_norm": 2.6294727325439453, + "learning_rate": 4.665017129805863e-06, + "loss": 1.5802, + "step": 484 + }, + { + "epoch": 0.49820236260914225, + "grad_norm": 2.2164878845214844, + "learning_rate": 4.663113818043396e-06, + "loss": 1.5384, + "step": 485 + }, + { + "epoch": 0.49922958397534667, + "grad_norm": 2.311232805252075, + "learning_rate": 4.66121050628093e-06, + "loss": 1.5785, + "step": 486 + }, + { + "epoch": 0.5002568053415511, + "grad_norm": 2.2916736602783203, + "learning_rate": 4.659307194518463e-06, + "loss": 1.5401, + "step": 487 + }, + { + "epoch": 0.5012840267077555, + "grad_norm": 2.3978800773620605, + "learning_rate": 4.6574038827559955e-06, + "loss": 1.5619, + "step": 488 + }, + { + "epoch": 0.50231124807396, + "grad_norm": 2.438917398452759, + "learning_rate": 4.655500570993529e-06, + "loss": 1.5824, + "step": 489 + }, + { + "epoch": 0.5033384694401644, + "grad_norm": 2.528862237930298, + "learning_rate": 4.653597259231062e-06, + "loss": 1.4802, + "step": 490 + }, + { + "epoch": 0.5033384694401644, + "eval_loss": 1.83504056930542, + "eval_runtime": 37.9357, + "eval_samples_per_second": 26.36, + "eval_steps_per_second": 1.107, + "step": 490 + }, + { + "epoch": 0.5043656908063687, + "grad_norm": 2.54522442817688, + "learning_rate": 4.651693947468596e-06, + "loss": 1.6186, + "step": 491 + }, + { + "epoch": 0.5053929121725732, + "grad_norm": 2.5037031173706055, + "learning_rate": 4.649790635706129e-06, + "loss": 1.5159, + "step": 492 + }, + { + "epoch": 0.5064201335387776, + "grad_norm": 2.4600465297698975, + "learning_rate": 4.647887323943662e-06, + "loss": 1.5378, + "step": 493 + }, + { + "epoch": 0.507447354904982, + "grad_norm": 2.393695831298828, + "learning_rate": 4.645984012181196e-06, + "loss": 1.4485, + "step": 494 + }, + { + "epoch": 0.5084745762711864, + "grad_norm": 2.488626718521118, + "learning_rate": 4.644080700418729e-06, + "loss": 1.57, + "step": 495 + }, + { + "epoch": 0.5095017976373909, + "grad_norm": 2.409858465194702, + "learning_rate": 4.642177388656262e-06, + "loss": 1.5343, + "step": 496 + }, + { + "epoch": 0.5105290190035953, + "grad_norm": 2.5607573986053467, + "learning_rate": 4.640274076893796e-06, + "loss": 1.5825, + "step": 497 + }, + { + "epoch": 0.5115562403697997, + "grad_norm": 2.5953354835510254, + "learning_rate": 4.638370765131328e-06, + "loss": 1.5129, + "step": 498 + }, + { + "epoch": 0.5125834617360041, + "grad_norm": 2.3321897983551025, + "learning_rate": 4.636467453368862e-06, + "loss": 1.4972, + "step": 499 + }, + { + "epoch": 0.5136106831022085, + "grad_norm": 2.332052707672119, + "learning_rate": 4.634564141606396e-06, + "loss": 1.567, + "step": 500 + }, + { + "epoch": 0.5136106831022085, + "eval_loss": 1.824558138847351, + "eval_runtime": 37.9157, + "eval_samples_per_second": 26.374, + "eval_steps_per_second": 1.108, + "step": 500 + }, + { + "epoch": 0.514637904468413, + "grad_norm": 2.431809902191162, + "learning_rate": 4.632660829843929e-06, + "loss": 1.494, + "step": 501 + }, + { + "epoch": 0.5156651258346173, + "grad_norm": 2.7118983268737793, + "learning_rate": 4.630757518081463e-06, + "loss": 1.6289, + "step": 502 + }, + { + "epoch": 0.5166923472008218, + "grad_norm": 2.3549370765686035, + "learning_rate": 4.628854206318996e-06, + "loss": 1.5363, + "step": 503 + }, + { + "epoch": 0.5177195685670262, + "grad_norm": 2.3343000411987305, + "learning_rate": 4.626950894556529e-06, + "loss": 1.5749, + "step": 504 + }, + { + "epoch": 0.5187467899332306, + "grad_norm": 2.4281466007232666, + "learning_rate": 4.625047582794062e-06, + "loss": 1.4958, + "step": 505 + }, + { + "epoch": 0.519774011299435, + "grad_norm": 2.3858845233917236, + "learning_rate": 4.623144271031595e-06, + "loss": 1.5031, + "step": 506 + }, + { + "epoch": 0.5208012326656395, + "grad_norm": 2.3714635372161865, + "learning_rate": 4.6212409592691286e-06, + "loss": 1.5808, + "step": 507 + }, + { + "epoch": 0.5218284540318439, + "grad_norm": 2.5290520191192627, + "learning_rate": 4.619337647506662e-06, + "loss": 1.541, + "step": 508 + }, + { + "epoch": 0.5228556753980482, + "grad_norm": 2.5521368980407715, + "learning_rate": 4.617434335744195e-06, + "loss": 1.4992, + "step": 509 + }, + { + "epoch": 0.5238828967642527, + "grad_norm": 2.436588764190674, + "learning_rate": 4.615531023981729e-06, + "loss": 1.5283, + "step": 510 + }, + { + "epoch": 0.5238828967642527, + "eval_loss": 1.8220165967941284, + "eval_runtime": 37.6123, + "eval_samples_per_second": 26.587, + "eval_steps_per_second": 1.117, + "step": 510 + }, + { + "epoch": 0.5249101181304571, + "grad_norm": 2.5240023136138916, + "learning_rate": 4.613627712219262e-06, + "loss": 1.5426, + "step": 511 + }, + { + "epoch": 0.5259373394966615, + "grad_norm": 2.5292775630950928, + "learning_rate": 4.611724400456795e-06, + "loss": 1.5411, + "step": 512 + }, + { + "epoch": 0.5269645608628659, + "grad_norm": 2.3622210025787354, + "learning_rate": 4.609821088694329e-06, + "loss": 1.5083, + "step": 513 + }, + { + "epoch": 0.5279917822290704, + "grad_norm": 2.545267105102539, + "learning_rate": 4.607917776931862e-06, + "loss": 1.5541, + "step": 514 + }, + { + "epoch": 0.5290190035952748, + "grad_norm": 2.3417961597442627, + "learning_rate": 4.606014465169395e-06, + "loss": 1.5686, + "step": 515 + }, + { + "epoch": 0.5300462249614792, + "grad_norm": 2.2782580852508545, + "learning_rate": 4.604111153406928e-06, + "loss": 1.4761, + "step": 516 + }, + { + "epoch": 0.5310734463276836, + "grad_norm": 2.4437527656555176, + "learning_rate": 4.602207841644461e-06, + "loss": 1.5818, + "step": 517 + }, + { + "epoch": 0.5321006676938881, + "grad_norm": 2.3594319820404053, + "learning_rate": 4.600304529881995e-06, + "loss": 1.5556, + "step": 518 + }, + { + "epoch": 0.5331278890600925, + "grad_norm": 2.325979471206665, + "learning_rate": 4.598401218119528e-06, + "loss": 1.5726, + "step": 519 + }, + { + "epoch": 0.5341551104262968, + "grad_norm": 2.2765676975250244, + "learning_rate": 4.5964979063570614e-06, + "loss": 1.5151, + "step": 520 + }, + { + "epoch": 0.5341551104262968, + "eval_loss": 1.825669288635254, + "eval_runtime": 37.7491, + "eval_samples_per_second": 26.491, + "eval_steps_per_second": 1.113, + "step": 520 + }, + { + "epoch": 0.5351823317925013, + "grad_norm": 2.357647657394409, + "learning_rate": 4.594594594594596e-06, + "loss": 1.5505, + "step": 521 + }, + { + "epoch": 0.5362095531587057, + "grad_norm": 2.3652896881103516, + "learning_rate": 4.592691282832128e-06, + "loss": 1.5621, + "step": 522 + }, + { + "epoch": 0.5372367745249101, + "grad_norm": 2.234096050262451, + "learning_rate": 4.5907879710696615e-06, + "loss": 1.4851, + "step": 523 + }, + { + "epoch": 0.5382639958911145, + "grad_norm": 2.3414828777313232, + "learning_rate": 4.588884659307195e-06, + "loss": 1.5594, + "step": 524 + }, + { + "epoch": 0.539291217257319, + "grad_norm": 2.3934929370880127, + "learning_rate": 4.586981347544728e-06, + "loss": 1.4994, + "step": 525 + }, + { + "epoch": 0.5403184386235234, + "grad_norm": 2.400479555130005, + "learning_rate": 4.585078035782262e-06, + "loss": 1.5346, + "step": 526 + }, + { + "epoch": 0.5413456599897278, + "grad_norm": 2.3817873001098633, + "learning_rate": 4.583174724019795e-06, + "loss": 1.4803, + "step": 527 + }, + { + "epoch": 0.5423728813559322, + "grad_norm": 2.4503185749053955, + "learning_rate": 4.581271412257328e-06, + "loss": 1.4711, + "step": 528 + }, + { + "epoch": 0.5434001027221366, + "grad_norm": 2.452578067779541, + "learning_rate": 4.579368100494862e-06, + "loss": 1.5645, + "step": 529 + }, + { + "epoch": 0.544427324088341, + "grad_norm": 2.3278186321258545, + "learning_rate": 4.577464788732395e-06, + "loss": 1.5428, + "step": 530 + }, + { + "epoch": 0.544427324088341, + "eval_loss": 1.8242785930633545, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 530 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 2.5192365646362305, + "learning_rate": 4.575561476969928e-06, + "loss": 1.5896, + "step": 531 + }, + { + "epoch": 0.5464817668207499, + "grad_norm": 2.308701276779175, + "learning_rate": 4.573658165207461e-06, + "loss": 1.5518, + "step": 532 + }, + { + "epoch": 0.5475089881869543, + "grad_norm": 2.4221107959747314, + "learning_rate": 4.571754853444994e-06, + "loss": 1.5387, + "step": 533 + }, + { + "epoch": 0.5485362095531587, + "grad_norm": 2.457606077194214, + "learning_rate": 4.569851541682528e-06, + "loss": 1.4997, + "step": 534 + }, + { + "epoch": 0.5495634309193631, + "grad_norm": 2.3406503200531006, + "learning_rate": 4.567948229920061e-06, + "loss": 1.4761, + "step": 535 + }, + { + "epoch": 0.5505906522855676, + "grad_norm": 2.555712938308716, + "learning_rate": 4.566044918157594e-06, + "loss": 1.5114, + "step": 536 + }, + { + "epoch": 0.551617873651772, + "grad_norm": 2.373013734817505, + "learning_rate": 4.564141606395128e-06, + "loss": 1.5516, + "step": 537 + }, + { + "epoch": 0.5526450950179764, + "grad_norm": 2.3915255069732666, + "learning_rate": 4.562238294632661e-06, + "loss": 1.5145, + "step": 538 + }, + { + "epoch": 0.5536723163841808, + "grad_norm": 2.3272323608398438, + "learning_rate": 4.5603349828701945e-06, + "loss": 1.5472, + "step": 539 + }, + { + "epoch": 0.5546995377503852, + "grad_norm": 2.416072130203247, + "learning_rate": 4.558431671107728e-06, + "loss": 1.5842, + "step": 540 + }, + { + "epoch": 0.5546995377503852, + "eval_loss": 1.825427532196045, + "eval_runtime": 37.9905, + "eval_samples_per_second": 26.322, + "eval_steps_per_second": 1.106, + "step": 540 + }, + { + "epoch": 0.5557267591165896, + "grad_norm": 2.3225271701812744, + "learning_rate": 4.556528359345261e-06, + "loss": 1.5188, + "step": 541 + }, + { + "epoch": 0.556753980482794, + "grad_norm": 2.2387030124664307, + "learning_rate": 4.5546250475827946e-06, + "loss": 1.4778, + "step": 542 + }, + { + "epoch": 0.5577812018489985, + "grad_norm": 2.3054494857788086, + "learning_rate": 4.552721735820328e-06, + "loss": 1.5229, + "step": 543 + }, + { + "epoch": 0.5588084232152029, + "grad_norm": 2.4956626892089844, + "learning_rate": 4.550818424057861e-06, + "loss": 1.5367, + "step": 544 + }, + { + "epoch": 0.5598356445814073, + "grad_norm": 2.454986810684204, + "learning_rate": 4.548915112295395e-06, + "loss": 1.5444, + "step": 545 + }, + { + "epoch": 0.5608628659476117, + "grad_norm": 2.466810703277588, + "learning_rate": 4.547011800532928e-06, + "loss": 1.4914, + "step": 546 + }, + { + "epoch": 0.5618900873138162, + "grad_norm": 2.51777982711792, + "learning_rate": 4.545108488770461e-06, + "loss": 1.5775, + "step": 547 + }, + { + "epoch": 0.5629173086800205, + "grad_norm": 2.340421199798584, + "learning_rate": 4.543205177007995e-06, + "loss": 1.5185, + "step": 548 + }, + { + "epoch": 0.5639445300462249, + "grad_norm": 2.376772880554199, + "learning_rate": 4.541301865245527e-06, + "loss": 1.5588, + "step": 549 + }, + { + "epoch": 0.5649717514124294, + "grad_norm": 2.4004335403442383, + "learning_rate": 4.539398553483061e-06, + "loss": 1.5283, + "step": 550 + }, + { + "epoch": 0.5649717514124294, + "eval_loss": 1.8294963836669922, + "eval_runtime": 37.9391, + "eval_samples_per_second": 26.358, + "eval_steps_per_second": 1.107, + "step": 550 + }, + { + "epoch": 0.5659989727786338, + "grad_norm": 2.424459934234619, + "learning_rate": 4.537495241720594e-06, + "loss": 1.5576, + "step": 551 + }, + { + "epoch": 0.5670261941448382, + "grad_norm": 2.4413998126983643, + "learning_rate": 4.535591929958127e-06, + "loss": 1.5444, + "step": 552 + }, + { + "epoch": 0.5680534155110426, + "grad_norm": 2.408897876739502, + "learning_rate": 4.533688618195661e-06, + "loss": 1.5269, + "step": 553 + }, + { + "epoch": 0.5690806368772471, + "grad_norm": 2.471024990081787, + "learning_rate": 4.531785306433194e-06, + "loss": 1.5109, + "step": 554 + }, + { + "epoch": 0.5701078582434514, + "grad_norm": 2.374882459640503, + "learning_rate": 4.529881994670727e-06, + "loss": 1.5205, + "step": 555 + }, + { + "epoch": 0.5711350796096559, + "grad_norm": 2.4166030883789062, + "learning_rate": 4.527978682908261e-06, + "loss": 1.5276, + "step": 556 + }, + { + "epoch": 0.5721623009758603, + "grad_norm": 2.2681620121002197, + "learning_rate": 4.526075371145794e-06, + "loss": 1.5374, + "step": 557 + }, + { + "epoch": 0.5731895223420647, + "grad_norm": 2.3933286666870117, + "learning_rate": 4.5241720593833275e-06, + "loss": 1.5349, + "step": 558 + }, + { + "epoch": 0.5742167437082691, + "grad_norm": 2.4086451530456543, + "learning_rate": 4.52226874762086e-06, + "loss": 1.5512, + "step": 559 + }, + { + "epoch": 0.5752439650744735, + "grad_norm": 2.559021472930908, + "learning_rate": 4.520365435858393e-06, + "loss": 1.5789, + "step": 560 + }, + { + "epoch": 0.5752439650744735, + "eval_loss": 1.8243234157562256, + "eval_runtime": 37.934, + "eval_samples_per_second": 26.362, + "eval_steps_per_second": 1.107, + "step": 560 + }, + { + "epoch": 0.576271186440678, + "grad_norm": 2.483494520187378, + "learning_rate": 4.518462124095928e-06, + "loss": 1.5348, + "step": 561 + }, + { + "epoch": 0.5772984078068824, + "grad_norm": 2.4039993286132812, + "learning_rate": 4.516558812333461e-06, + "loss": 1.5201, + "step": 562 + }, + { + "epoch": 0.5783256291730868, + "grad_norm": 2.5080959796905518, + "learning_rate": 4.514655500570994e-06, + "loss": 1.5864, + "step": 563 + }, + { + "epoch": 0.5793528505392912, + "grad_norm": 2.620887279510498, + "learning_rate": 4.512752188808528e-06, + "loss": 1.5901, + "step": 564 + }, + { + "epoch": 0.5803800719054957, + "grad_norm": 2.5691096782684326, + "learning_rate": 4.510848877046061e-06, + "loss": 1.5506, + "step": 565 + }, + { + "epoch": 0.5814072932717, + "grad_norm": 2.3777835369110107, + "learning_rate": 4.5089455652835936e-06, + "loss": 1.5149, + "step": 566 + }, + { + "epoch": 0.5824345146379045, + "grad_norm": 2.2791666984558105, + "learning_rate": 4.507042253521127e-06, + "loss": 1.4771, + "step": 567 + }, + { + "epoch": 0.5834617360041089, + "grad_norm": 2.4953627586364746, + "learning_rate": 4.50513894175866e-06, + "loss": 1.556, + "step": 568 + }, + { + "epoch": 0.5844889573703133, + "grad_norm": 2.3563215732574463, + "learning_rate": 4.503235629996194e-06, + "loss": 1.549, + "step": 569 + }, + { + "epoch": 0.5855161787365177, + "grad_norm": 2.49676513671875, + "learning_rate": 4.501332318233727e-06, + "loss": 1.513, + "step": 570 + }, + { + "epoch": 0.5855161787365177, + "eval_loss": 1.8353043794631958, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 570 + }, + { + "epoch": 0.5865434001027221, + "grad_norm": 2.385275363922119, + "learning_rate": 4.49942900647126e-06, + "loss": 1.5622, + "step": 571 + }, + { + "epoch": 0.5875706214689266, + "grad_norm": 2.3089914321899414, + "learning_rate": 4.497525694708794e-06, + "loss": 1.5426, + "step": 572 + }, + { + "epoch": 0.588597842835131, + "grad_norm": 2.307323455810547, + "learning_rate": 4.495622382946327e-06, + "loss": 1.5088, + "step": 573 + }, + { + "epoch": 0.5896250642013354, + "grad_norm": 2.597288131713867, + "learning_rate": 4.4937190711838605e-06, + "loss": 1.5788, + "step": 574 + }, + { + "epoch": 0.5906522855675398, + "grad_norm": 2.3923256397247314, + "learning_rate": 4.491815759421394e-06, + "loss": 1.5038, + "step": 575 + }, + { + "epoch": 0.5916795069337443, + "grad_norm": 2.4199156761169434, + "learning_rate": 4.489912447658926e-06, + "loss": 1.5057, + "step": 576 + }, + { + "epoch": 0.5927067282999486, + "grad_norm": 2.5643393993377686, + "learning_rate": 4.48800913589646e-06, + "loss": 1.5274, + "step": 577 + }, + { + "epoch": 0.593733949666153, + "grad_norm": 2.5854485034942627, + "learning_rate": 4.486105824133993e-06, + "loss": 1.5341, + "step": 578 + }, + { + "epoch": 0.5947611710323575, + "grad_norm": 2.5363597869873047, + "learning_rate": 4.4842025123715264e-06, + "loss": 1.4953, + "step": 579 + }, + { + "epoch": 0.5957883923985619, + "grad_norm": 2.3880693912506104, + "learning_rate": 4.48229920060906e-06, + "loss": 1.4868, + "step": 580 + }, + { + "epoch": 0.5957883923985619, + "eval_loss": 1.8251410722732544, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 580 + }, + { + "epoch": 0.5968156137647663, + "grad_norm": 2.5272583961486816, + "learning_rate": 4.480395888846594e-06, + "loss": 1.5851, + "step": 581 + }, + { + "epoch": 0.5978428351309707, + "grad_norm": 2.4496943950653076, + "learning_rate": 4.478492577084127e-06, + "loss": 1.4896, + "step": 582 + }, + { + "epoch": 0.5988700564971752, + "grad_norm": 2.3996055126190186, + "learning_rate": 4.47658926532166e-06, + "loss": 1.5253, + "step": 583 + }, + { + "epoch": 0.5998972778633795, + "grad_norm": 2.5964202880859375, + "learning_rate": 4.474685953559193e-06, + "loss": 1.6294, + "step": 584 + }, + { + "epoch": 0.600924499229584, + "grad_norm": 2.5567739009857178, + "learning_rate": 4.472782641796727e-06, + "loss": 1.4834, + "step": 585 + }, + { + "epoch": 0.6019517205957884, + "grad_norm": 2.440291166305542, + "learning_rate": 4.47087933003426e-06, + "loss": 1.4966, + "step": 586 + }, + { + "epoch": 0.6029789419619929, + "grad_norm": 2.525763988494873, + "learning_rate": 4.468976018271793e-06, + "loss": 1.5005, + "step": 587 + }, + { + "epoch": 0.6040061633281972, + "grad_norm": 2.4822962284088135, + "learning_rate": 4.467072706509327e-06, + "loss": 1.5358, + "step": 588 + }, + { + "epoch": 0.6050333846944016, + "grad_norm": 2.591149091720581, + "learning_rate": 4.46516939474686e-06, + "loss": 1.4975, + "step": 589 + }, + { + "epoch": 0.6060606060606061, + "grad_norm": 2.4884238243103027, + "learning_rate": 4.463266082984393e-06, + "loss": 1.6107, + "step": 590 + }, + { + "epoch": 0.6060606060606061, + "eval_loss": 1.8383466005325317, + "eval_runtime": 37.9554, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 590 + }, + { + "epoch": 0.6070878274268104, + "grad_norm": 2.4965732097625732, + "learning_rate": 4.461362771221927e-06, + "loss": 1.4986, + "step": 591 + }, + { + "epoch": 0.6081150487930149, + "grad_norm": 2.544792413711548, + "learning_rate": 4.45945945945946e-06, + "loss": 1.5067, + "step": 592 + }, + { + "epoch": 0.6091422701592193, + "grad_norm": 2.5726170539855957, + "learning_rate": 4.457556147696993e-06, + "loss": 1.5861, + "step": 593 + }, + { + "epoch": 0.6101694915254238, + "grad_norm": 2.493027448654175, + "learning_rate": 4.455652835934526e-06, + "loss": 1.5857, + "step": 594 + }, + { + "epoch": 0.6111967128916281, + "grad_norm": 2.470771074295044, + "learning_rate": 4.453749524172059e-06, + "loss": 1.5419, + "step": 595 + }, + { + "epoch": 0.6122239342578326, + "grad_norm": 2.568052053451538, + "learning_rate": 4.451846212409593e-06, + "loss": 1.5391, + "step": 596 + }, + { + "epoch": 0.613251155624037, + "grad_norm": 2.4346213340759277, + "learning_rate": 4.449942900647126e-06, + "loss": 1.4761, + "step": 597 + }, + { + "epoch": 0.6142783769902413, + "grad_norm": 2.418133497238159, + "learning_rate": 4.4480395888846595e-06, + "loss": 1.5383, + "step": 598 + }, + { + "epoch": 0.6153055983564458, + "grad_norm": 2.4795567989349365, + "learning_rate": 4.446136277122193e-06, + "loss": 1.4777, + "step": 599 + }, + { + "epoch": 0.6163328197226502, + "grad_norm": 2.5122742652893066, + "learning_rate": 4.444232965359726e-06, + "loss": 1.5353, + "step": 600 + }, + { + "epoch": 0.6163328197226502, + "eval_loss": 1.8325241804122925, + "eval_runtime": 37.9546, + "eval_samples_per_second": 26.347, + "eval_steps_per_second": 1.107, + "step": 600 + }, + { + "epoch": 0.6173600410888547, + "grad_norm": 2.604067325592041, + "learning_rate": 4.4423296535972596e-06, + "loss": 1.489, + "step": 601 + }, + { + "epoch": 0.618387262455059, + "grad_norm": 2.3741753101348877, + "learning_rate": 4.440426341834793e-06, + "loss": 1.532, + "step": 602 + }, + { + "epoch": 0.6194144838212635, + "grad_norm": 2.5373036861419678, + "learning_rate": 4.438523030072326e-06, + "loss": 1.561, + "step": 603 + }, + { + "epoch": 0.6204417051874679, + "grad_norm": 2.2958099842071533, + "learning_rate": 4.43661971830986e-06, + "loss": 1.5025, + "step": 604 + }, + { + "epoch": 0.6214689265536724, + "grad_norm": 2.417006015777588, + "learning_rate": 4.434716406547393e-06, + "loss": 1.4977, + "step": 605 + }, + { + "epoch": 0.6224961479198767, + "grad_norm": 2.335296392440796, + "learning_rate": 4.432813094784926e-06, + "loss": 1.5504, + "step": 606 + }, + { + "epoch": 0.6235233692860811, + "grad_norm": 2.4086828231811523, + "learning_rate": 4.43090978302246e-06, + "loss": 1.5217, + "step": 607 + }, + { + "epoch": 0.6245505906522856, + "grad_norm": 2.3816819190979004, + "learning_rate": 4.429006471259993e-06, + "loss": 1.5126, + "step": 608 + }, + { + "epoch": 0.6255778120184899, + "grad_norm": 2.307671546936035, + "learning_rate": 4.4271031594975265e-06, + "loss": 1.4876, + "step": 609 + }, + { + "epoch": 0.6266050333846944, + "grad_norm": 2.355837821960449, + "learning_rate": 4.425199847735059e-06, + "loss": 1.4827, + "step": 610 + }, + { + "epoch": 0.6266050333846944, + "eval_loss": 1.8269455432891846, + "eval_runtime": 37.6459, + "eval_samples_per_second": 26.563, + "eval_steps_per_second": 1.116, + "step": 610 + }, + { + "epoch": 0.6276322547508988, + "grad_norm": 2.558631658554077, + "learning_rate": 4.423296535972592e-06, + "loss": 1.5172, + "step": 611 + }, + { + "epoch": 0.6286594761171033, + "grad_norm": 2.3170175552368164, + "learning_rate": 4.421393224210126e-06, + "loss": 1.4784, + "step": 612 + }, + { + "epoch": 0.6296866974833076, + "grad_norm": 2.4064524173736572, + "learning_rate": 4.419489912447659e-06, + "loss": 1.5616, + "step": 613 + }, + { + "epoch": 0.6307139188495121, + "grad_norm": 2.3719160556793213, + "learning_rate": 4.4175866006851924e-06, + "loss": 1.5292, + "step": 614 + }, + { + "epoch": 0.6317411402157165, + "grad_norm": 2.3808202743530273, + "learning_rate": 4.415683288922726e-06, + "loss": 1.5122, + "step": 615 + }, + { + "epoch": 0.632768361581921, + "grad_norm": 2.4208037853240967, + "learning_rate": 4.413779977160259e-06, + "loss": 1.5088, + "step": 616 + }, + { + "epoch": 0.6337955829481253, + "grad_norm": 2.5558979511260986, + "learning_rate": 4.4118766653977925e-06, + "loss": 1.5678, + "step": 617 + }, + { + "epoch": 0.6348228043143297, + "grad_norm": 2.443908452987671, + "learning_rate": 4.409973353635326e-06, + "loss": 1.5234, + "step": 618 + }, + { + "epoch": 0.6358500256805342, + "grad_norm": 2.420408010482788, + "learning_rate": 4.408070041872859e-06, + "loss": 1.4708, + "step": 619 + }, + { + "epoch": 0.6368772470467385, + "grad_norm": 2.3884904384613037, + "learning_rate": 4.406166730110393e-06, + "loss": 1.4821, + "step": 620 + }, + { + "epoch": 0.6368772470467385, + "eval_loss": 1.8217289447784424, + "eval_runtime": 37.7936, + "eval_samples_per_second": 26.459, + "eval_steps_per_second": 1.111, + "step": 620 + }, + { + "epoch": 0.637904468412943, + "grad_norm": 2.313720703125, + "learning_rate": 4.404263418347925e-06, + "loss": 1.4962, + "step": 621 + }, + { + "epoch": 0.6389316897791474, + "grad_norm": 2.3173866271972656, + "learning_rate": 4.402360106585459e-06, + "loss": 1.4857, + "step": 622 + }, + { + "epoch": 0.6399589111453519, + "grad_norm": 2.541757822036743, + "learning_rate": 4.400456794822993e-06, + "loss": 1.5177, + "step": 623 + }, + { + "epoch": 0.6409861325115562, + "grad_norm": 2.4382357597351074, + "learning_rate": 4.398553483060526e-06, + "loss": 1.4718, + "step": 624 + }, + { + "epoch": 0.6420133538777607, + "grad_norm": 2.2988059520721436, + "learning_rate": 4.396650171298059e-06, + "loss": 1.4883, + "step": 625 + }, + { + "epoch": 0.6430405752439651, + "grad_norm": 2.4817495346069336, + "learning_rate": 4.394746859535593e-06, + "loss": 1.4823, + "step": 626 + }, + { + "epoch": 0.6440677966101694, + "grad_norm": 2.4578585624694824, + "learning_rate": 4.392843547773125e-06, + "loss": 1.4882, + "step": 627 + }, + { + "epoch": 0.6450950179763739, + "grad_norm": 2.184972047805786, + "learning_rate": 4.390940236010659e-06, + "loss": 1.514, + "step": 628 + }, + { + "epoch": 0.6461222393425783, + "grad_norm": 2.4040279388427734, + "learning_rate": 4.389036924248192e-06, + "loss": 1.5343, + "step": 629 + }, + { + "epoch": 0.6471494607087828, + "grad_norm": 2.3676459789276123, + "learning_rate": 4.387133612485725e-06, + "loss": 1.5007, + "step": 630 + }, + { + "epoch": 0.6471494607087828, + "eval_loss": 1.8106164932250977, + "eval_runtime": 37.9383, + "eval_samples_per_second": 26.359, + "eval_steps_per_second": 1.107, + "step": 630 + }, + { + "epoch": 0.6481766820749871, + "grad_norm": 2.234689235687256, + "learning_rate": 4.385230300723259e-06, + "loss": 1.5064, + "step": 631 + }, + { + "epoch": 0.6492039034411916, + "grad_norm": 2.281742572784424, + "learning_rate": 4.383326988960792e-06, + "loss": 1.4777, + "step": 632 + }, + { + "epoch": 0.650231124807396, + "grad_norm": 2.4314322471618652, + "learning_rate": 4.3814236771983255e-06, + "loss": 1.5331, + "step": 633 + }, + { + "epoch": 0.6512583461736005, + "grad_norm": 2.4092588424682617, + "learning_rate": 4.379520365435859e-06, + "loss": 1.5122, + "step": 634 + }, + { + "epoch": 0.6522855675398048, + "grad_norm": 2.517362117767334, + "learning_rate": 4.377617053673392e-06, + "loss": 1.52, + "step": 635 + }, + { + "epoch": 0.6533127889060092, + "grad_norm": 2.4784109592437744, + "learning_rate": 4.3757137419109256e-06, + "loss": 1.5771, + "step": 636 + }, + { + "epoch": 0.6543400102722137, + "grad_norm": 2.32181715965271, + "learning_rate": 4.373810430148459e-06, + "loss": 1.4718, + "step": 637 + }, + { + "epoch": 0.655367231638418, + "grad_norm": 2.5146260261535645, + "learning_rate": 4.3719071183859914e-06, + "loss": 1.5307, + "step": 638 + }, + { + "epoch": 0.6563944530046225, + "grad_norm": 2.2364494800567627, + "learning_rate": 4.370003806623525e-06, + "loss": 1.4803, + "step": 639 + }, + { + "epoch": 0.6574216743708269, + "grad_norm": 2.3735806941986084, + "learning_rate": 4.368100494861058e-06, + "loss": 1.5249, + "step": 640 + }, + { + "epoch": 0.6574216743708269, + "eval_loss": 1.8134880065917969, + "eval_runtime": 38.0376, + "eval_samples_per_second": 26.29, + "eval_steps_per_second": 1.104, + "step": 640 + }, + { + "epoch": 0.6584488957370314, + "grad_norm": 2.4600815773010254, + "learning_rate": 4.3661971830985915e-06, + "loss": 1.4893, + "step": 641 + }, + { + "epoch": 0.6594761171032357, + "grad_norm": 2.4106783866882324, + "learning_rate": 4.364293871336126e-06, + "loss": 1.5163, + "step": 642 + }, + { + "epoch": 0.6605033384694402, + "grad_norm": 2.276624917984009, + "learning_rate": 4.362390559573659e-06, + "loss": 1.482, + "step": 643 + }, + { + "epoch": 0.6615305598356446, + "grad_norm": 2.336341381072998, + "learning_rate": 4.3604872478111925e-06, + "loss": 1.4152, + "step": 644 + }, + { + "epoch": 0.662557781201849, + "grad_norm": 2.3329105377197266, + "learning_rate": 4.358583936048725e-06, + "loss": 1.5411, + "step": 645 + }, + { + "epoch": 0.6635850025680534, + "grad_norm": 2.2877161502838135, + "learning_rate": 4.356680624286258e-06, + "loss": 1.4997, + "step": 646 + }, + { + "epoch": 0.6646122239342578, + "grad_norm": 2.410820484161377, + "learning_rate": 4.354777312523792e-06, + "loss": 1.5446, + "step": 647 + }, + { + "epoch": 0.6656394453004623, + "grad_norm": 2.403038501739502, + "learning_rate": 4.352874000761325e-06, + "loss": 1.4916, + "step": 648 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 2.3261778354644775, + "learning_rate": 4.3509706889988584e-06, + "loss": 1.5561, + "step": 649 + }, + { + "epoch": 0.6676938880328711, + "grad_norm": 2.4512174129486084, + "learning_rate": 4.349067377236392e-06, + "loss": 1.556, + "step": 650 + }, + { + "epoch": 0.6676938880328711, + "eval_loss": 1.8238279819488525, + "eval_runtime": 37.9988, + "eval_samples_per_second": 26.317, + "eval_steps_per_second": 1.105, + "step": 650 + }, + { + "epoch": 0.6687211093990755, + "grad_norm": 2.4523823261260986, + "learning_rate": 4.347164065473925e-06, + "loss": 1.5676, + "step": 651 + }, + { + "epoch": 0.66974833076528, + "grad_norm": 2.3770365715026855, + "learning_rate": 4.3452607537114585e-06, + "loss": 1.4815, + "step": 652 + }, + { + "epoch": 0.6707755521314843, + "grad_norm": 2.2440760135650635, + "learning_rate": 4.343357441948992e-06, + "loss": 1.5141, + "step": 653 + }, + { + "epoch": 0.6718027734976888, + "grad_norm": 2.3341948986053467, + "learning_rate": 4.341454130186525e-06, + "loss": 1.4859, + "step": 654 + }, + { + "epoch": 0.6728299948638932, + "grad_norm": 2.507354497909546, + "learning_rate": 4.339550818424058e-06, + "loss": 1.4734, + "step": 655 + }, + { + "epoch": 0.6738572162300975, + "grad_norm": 2.395954132080078, + "learning_rate": 4.337647506661591e-06, + "loss": 1.5069, + "step": 656 + }, + { + "epoch": 0.674884437596302, + "grad_norm": 2.362211227416992, + "learning_rate": 4.3357441948991245e-06, + "loss": 1.5299, + "step": 657 + }, + { + "epoch": 0.6759116589625064, + "grad_norm": 2.460008144378662, + "learning_rate": 4.333840883136658e-06, + "loss": 1.4795, + "step": 658 + }, + { + "epoch": 0.6769388803287109, + "grad_norm": 2.3578217029571533, + "learning_rate": 4.331937571374191e-06, + "loss": 1.4906, + "step": 659 + }, + { + "epoch": 0.6779661016949152, + "grad_norm": 2.398843288421631, + "learning_rate": 4.3300342596117246e-06, + "loss": 1.4986, + "step": 660 + }, + { + "epoch": 0.6779661016949152, + "eval_loss": 1.8206881284713745, + "eval_runtime": 38.044, + "eval_samples_per_second": 26.285, + "eval_steps_per_second": 1.104, + "step": 660 + }, + { + "epoch": 0.6789933230611197, + "grad_norm": 2.3587260246276855, + "learning_rate": 4.328130947849258e-06, + "loss": 1.461, + "step": 661 + }, + { + "epoch": 0.6800205444273241, + "grad_norm": 2.311952590942383, + "learning_rate": 4.326227636086791e-06, + "loss": 1.5041, + "step": 662 + }, + { + "epoch": 0.6810477657935285, + "grad_norm": 2.376188278198242, + "learning_rate": 4.324324324324325e-06, + "loss": 1.5042, + "step": 663 + }, + { + "epoch": 0.6820749871597329, + "grad_norm": 2.5807230472564697, + "learning_rate": 4.322421012561858e-06, + "loss": 1.4935, + "step": 664 + }, + { + "epoch": 0.6831022085259374, + "grad_norm": 2.368128538131714, + "learning_rate": 4.320517700799391e-06, + "loss": 1.5851, + "step": 665 + }, + { + "epoch": 0.6841294298921418, + "grad_norm": 2.4517524242401123, + "learning_rate": 4.318614389036925e-06, + "loss": 1.5528, + "step": 666 + }, + { + "epoch": 0.6851566512583461, + "grad_norm": 2.5436811447143555, + "learning_rate": 4.316711077274458e-06, + "loss": 1.4633, + "step": 667 + }, + { + "epoch": 0.6861838726245506, + "grad_norm": 2.4287915229797363, + "learning_rate": 4.3148077655119915e-06, + "loss": 1.5454, + "step": 668 + }, + { + "epoch": 0.687211093990755, + "grad_norm": 2.386528730392456, + "learning_rate": 4.312904453749525e-06, + "loss": 1.5116, + "step": 669 + }, + { + "epoch": 0.6882383153569595, + "grad_norm": 2.6602251529693604, + "learning_rate": 4.311001141987058e-06, + "loss": 1.5199, + "step": 670 + }, + { + "epoch": 0.6882383153569595, + "eval_loss": 1.8252663612365723, + "eval_runtime": 38.0791, + "eval_samples_per_second": 26.261, + "eval_steps_per_second": 1.103, + "step": 670 + }, + { + "epoch": 0.6892655367231638, + "grad_norm": 2.351128339767456, + "learning_rate": 4.3090978302245916e-06, + "loss": 1.507, + "step": 671 + }, + { + "epoch": 0.6902927580893683, + "grad_norm": 2.393441677093506, + "learning_rate": 4.307194518462124e-06, + "loss": 1.579, + "step": 672 + }, + { + "epoch": 0.6913199794555727, + "grad_norm": 2.3730578422546387, + "learning_rate": 4.3052912066996574e-06, + "loss": 1.5446, + "step": 673 + }, + { + "epoch": 0.6923472008217771, + "grad_norm": 2.2522945404052734, + "learning_rate": 4.303387894937191e-06, + "loss": 1.4705, + "step": 674 + }, + { + "epoch": 0.6933744221879815, + "grad_norm": 2.384563446044922, + "learning_rate": 4.301484583174724e-06, + "loss": 1.4636, + "step": 675 + }, + { + "epoch": 0.6944016435541859, + "grad_norm": 2.2879116535186768, + "learning_rate": 4.2995812714122575e-06, + "loss": 1.4914, + "step": 676 + }, + { + "epoch": 0.6954288649203904, + "grad_norm": 2.494788646697998, + "learning_rate": 4.297677959649791e-06, + "loss": 1.5126, + "step": 677 + }, + { + "epoch": 0.6964560862865947, + "grad_norm": 2.700012683868408, + "learning_rate": 4.295774647887324e-06, + "loss": 1.5362, + "step": 678 + }, + { + "epoch": 0.6974833076527992, + "grad_norm": 2.5130302906036377, + "learning_rate": 4.293871336124858e-06, + "loss": 1.5075, + "step": 679 + }, + { + "epoch": 0.6985105290190036, + "grad_norm": 2.486236572265625, + "learning_rate": 4.291968024362391e-06, + "loss": 1.5338, + "step": 680 + }, + { + "epoch": 0.6985105290190036, + "eval_loss": 1.8229657411575317, + "eval_runtime": 38.0642, + "eval_samples_per_second": 26.271, + "eval_steps_per_second": 1.103, + "step": 680 + }, + { + "epoch": 0.699537750385208, + "grad_norm": 2.441683530807495, + "learning_rate": 4.290064712599924e-06, + "loss": 1.537, + "step": 681 + }, + { + "epoch": 0.7005649717514124, + "grad_norm": 2.396798610687256, + "learning_rate": 4.288161400837458e-06, + "loss": 1.4553, + "step": 682 + }, + { + "epoch": 0.7015921931176169, + "grad_norm": 2.3636221885681152, + "learning_rate": 4.286258089074991e-06, + "loss": 1.5138, + "step": 683 + }, + { + "epoch": 0.7026194144838213, + "grad_norm": 2.4584708213806152, + "learning_rate": 4.284354777312524e-06, + "loss": 1.4725, + "step": 684 + }, + { + "epoch": 0.7036466358500256, + "grad_norm": 2.850332498550415, + "learning_rate": 4.282451465550058e-06, + "loss": 1.5075, + "step": 685 + }, + { + "epoch": 0.7046738572162301, + "grad_norm": 2.6345887184143066, + "learning_rate": 4.280548153787591e-06, + "loss": 1.5054, + "step": 686 + }, + { + "epoch": 0.7057010785824345, + "grad_norm": 2.3535311222076416, + "learning_rate": 4.2786448420251245e-06, + "loss": 1.4607, + "step": 687 + }, + { + "epoch": 0.706728299948639, + "grad_norm": 2.490431547164917, + "learning_rate": 4.276741530262658e-06, + "loss": 1.4852, + "step": 688 + }, + { + "epoch": 0.7077555213148433, + "grad_norm": 2.4353036880493164, + "learning_rate": 4.27483821850019e-06, + "loss": 1.5503, + "step": 689 + }, + { + "epoch": 0.7087827426810478, + "grad_norm": 2.3356940746307373, + "learning_rate": 4.272934906737724e-06, + "loss": 1.5416, + "step": 690 + }, + { + "epoch": 0.7087827426810478, + "eval_loss": 1.8164218664169312, + "eval_runtime": 38.0592, + "eval_samples_per_second": 26.275, + "eval_steps_per_second": 1.104, + "step": 690 + }, + { + "epoch": 0.7098099640472522, + "grad_norm": 2.4772775173187256, + "learning_rate": 4.271031594975257e-06, + "loss": 1.5392, + "step": 691 + }, + { + "epoch": 0.7108371854134566, + "grad_norm": 2.2921438217163086, + "learning_rate": 4.2691282832127905e-06, + "loss": 1.4935, + "step": 692 + }, + { + "epoch": 0.711864406779661, + "grad_norm": 2.3176305294036865, + "learning_rate": 4.267224971450324e-06, + "loss": 1.4451, + "step": 693 + }, + { + "epoch": 0.7128916281458655, + "grad_norm": 2.6825685501098633, + "learning_rate": 4.265321659687857e-06, + "loss": 1.4878, + "step": 694 + }, + { + "epoch": 0.7139188495120699, + "grad_norm": 2.6141414642333984, + "learning_rate": 4.2634183479253906e-06, + "loss": 1.4695, + "step": 695 + }, + { + "epoch": 0.7149460708782742, + "grad_norm": 2.5179429054260254, + "learning_rate": 4.261515036162924e-06, + "loss": 1.492, + "step": 696 + }, + { + "epoch": 0.7159732922444787, + "grad_norm": 2.4744884967803955, + "learning_rate": 4.259611724400457e-06, + "loss": 1.4752, + "step": 697 + }, + { + "epoch": 0.7170005136106831, + "grad_norm": 2.4911048412323, + "learning_rate": 4.257708412637991e-06, + "loss": 1.423, + "step": 698 + }, + { + "epoch": 0.7180277349768875, + "grad_norm": 2.5116214752197266, + "learning_rate": 4.255805100875523e-06, + "loss": 1.4535, + "step": 699 + }, + { + "epoch": 0.7190549563430919, + "grad_norm": 2.476651430130005, + "learning_rate": 4.2539017891130565e-06, + "loss": 1.5349, + "step": 700 + }, + { + "epoch": 0.7190549563430919, + "eval_loss": 1.8069179058074951, + "eval_runtime": 38.0292, + "eval_samples_per_second": 26.296, + "eval_steps_per_second": 1.104, + "step": 700 + }, + { + "epoch": 0.7200821777092964, + "grad_norm": 2.5234575271606445, + "learning_rate": 4.25199847735059e-06, + "loss": 1.4596, + "step": 701 + }, + { + "epoch": 0.7211093990755008, + "grad_norm": 2.33290433883667, + "learning_rate": 4.250095165588123e-06, + "loss": 1.5147, + "step": 702 + }, + { + "epoch": 0.7221366204417052, + "grad_norm": 2.479883909225464, + "learning_rate": 4.2481918538256575e-06, + "loss": 1.47, + "step": 703 + }, + { + "epoch": 0.7231638418079096, + "grad_norm": 2.432253837585449, + "learning_rate": 4.246288542063191e-06, + "loss": 1.5152, + "step": 704 + }, + { + "epoch": 0.724191063174114, + "grad_norm": 2.358304738998413, + "learning_rate": 4.244385230300724e-06, + "loss": 1.4562, + "step": 705 + }, + { + "epoch": 0.7252182845403184, + "grad_norm": 2.480877161026001, + "learning_rate": 4.242481918538257e-06, + "loss": 1.4906, + "step": 706 + }, + { + "epoch": 0.7262455059065228, + "grad_norm": 2.4287192821502686, + "learning_rate": 4.24057860677579e-06, + "loss": 1.5294, + "step": 707 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 2.4782888889312744, + "learning_rate": 4.2386752950133234e-06, + "loss": 1.4767, + "step": 708 + }, + { + "epoch": 0.7282999486389317, + "grad_norm": 2.4948253631591797, + "learning_rate": 4.236771983250857e-06, + "loss": 1.4664, + "step": 709 + }, + { + "epoch": 0.7293271700051361, + "grad_norm": 2.4625205993652344, + "learning_rate": 4.23486867148839e-06, + "loss": 1.495, + "step": 710 + }, + { + "epoch": 0.7293271700051361, + "eval_loss": 1.8140079975128174, + "eval_runtime": 37.6964, + "eval_samples_per_second": 26.528, + "eval_steps_per_second": 1.114, + "step": 710 + }, + { + "epoch": 0.7303543913713405, + "grad_norm": 2.669027805328369, + "learning_rate": 4.2329653597259235e-06, + "loss": 1.5204, + "step": 711 + }, + { + "epoch": 0.731381612737545, + "grad_norm": 2.5024707317352295, + "learning_rate": 4.231062047963457e-06, + "loss": 1.5283, + "step": 712 + }, + { + "epoch": 0.7324088341037494, + "grad_norm": 2.498258590698242, + "learning_rate": 4.22915873620099e-06, + "loss": 1.5139, + "step": 713 + }, + { + "epoch": 0.7334360554699538, + "grad_norm": 2.4305667877197266, + "learning_rate": 4.227255424438524e-06, + "loss": 1.5337, + "step": 714 + }, + { + "epoch": 0.7344632768361582, + "grad_norm": 2.3153419494628906, + "learning_rate": 4.225352112676057e-06, + "loss": 1.4646, + "step": 715 + }, + { + "epoch": 0.7354904982023626, + "grad_norm": 2.2835583686828613, + "learning_rate": 4.2234488009135895e-06, + "loss": 1.4328, + "step": 716 + }, + { + "epoch": 0.736517719568567, + "grad_norm": 2.401418924331665, + "learning_rate": 4.221545489151123e-06, + "loss": 1.4685, + "step": 717 + }, + { + "epoch": 0.7375449409347714, + "grad_norm": 2.454864978790283, + "learning_rate": 4.219642177388656e-06, + "loss": 1.4925, + "step": 718 + }, + { + "epoch": 0.7385721623009759, + "grad_norm": 2.2332873344421387, + "learning_rate": 4.2177388656261896e-06, + "loss": 1.4481, + "step": 719 + }, + { + "epoch": 0.7395993836671803, + "grad_norm": 2.46628475189209, + "learning_rate": 4.215835553863723e-06, + "loss": 1.4173, + "step": 720 + }, + { + "epoch": 0.7395993836671803, + "eval_loss": 1.817592740058899, + "eval_runtime": 37.8086, + "eval_samples_per_second": 26.449, + "eval_steps_per_second": 1.111, + "step": 720 + }, + { + "epoch": 0.7406266050333847, + "grad_norm": 2.346245765686035, + "learning_rate": 4.213932242101256e-06, + "loss": 1.4688, + "step": 721 + }, + { + "epoch": 0.7416538263995891, + "grad_norm": 2.2708640098571777, + "learning_rate": 4.21202893033879e-06, + "loss": 1.4937, + "step": 722 + }, + { + "epoch": 0.7426810477657936, + "grad_norm": 2.3716516494750977, + "learning_rate": 4.210125618576323e-06, + "loss": 1.4869, + "step": 723 + }, + { + "epoch": 0.743708269131998, + "grad_norm": 2.443058490753174, + "learning_rate": 4.208222306813856e-06, + "loss": 1.542, + "step": 724 + }, + { + "epoch": 0.7447354904982023, + "grad_norm": 2.327270984649658, + "learning_rate": 4.20631899505139e-06, + "loss": 1.3933, + "step": 725 + }, + { + "epoch": 0.7457627118644068, + "grad_norm": 2.3734512329101562, + "learning_rate": 4.204415683288923e-06, + "loss": 1.5109, + "step": 726 + }, + { + "epoch": 0.7467899332306112, + "grad_norm": 2.3631174564361572, + "learning_rate": 4.2025123715264565e-06, + "loss": 1.474, + "step": 727 + }, + { + "epoch": 0.7478171545968156, + "grad_norm": 2.306068181991577, + "learning_rate": 4.20060905976399e-06, + "loss": 1.42, + "step": 728 + }, + { + "epoch": 0.74884437596302, + "grad_norm": 2.391889810562134, + "learning_rate": 4.198705748001523e-06, + "loss": 1.5178, + "step": 729 + }, + { + "epoch": 0.7498715973292245, + "grad_norm": 2.3151705265045166, + "learning_rate": 4.1968024362390566e-06, + "loss": 1.4808, + "step": 730 + }, + { + "epoch": 0.7498715973292245, + "eval_loss": 1.8157070875167847, + "eval_runtime": 37.8799, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 730 + }, + { + "epoch": 0.7508988186954288, + "grad_norm": 2.336479902267456, + "learning_rate": 4.19489912447659e-06, + "loss": 1.5438, + "step": 731 + }, + { + "epoch": 0.7519260400616333, + "grad_norm": 2.338242292404175, + "learning_rate": 4.192995812714123e-06, + "loss": 1.4869, + "step": 732 + }, + { + "epoch": 0.7529532614278377, + "grad_norm": 2.268650770187378, + "learning_rate": 4.191092500951656e-06, + "loss": 1.4787, + "step": 733 + }, + { + "epoch": 0.7539804827940421, + "grad_norm": 2.3010005950927734, + "learning_rate": 4.189189189189189e-06, + "loss": 1.4711, + "step": 734 + }, + { + "epoch": 0.7550077041602465, + "grad_norm": 2.54770827293396, + "learning_rate": 4.1872858774267225e-06, + "loss": 1.5351, + "step": 735 + }, + { + "epoch": 0.7560349255264509, + "grad_norm": 2.3412787914276123, + "learning_rate": 4.185382565664256e-06, + "loss": 1.4655, + "step": 736 + }, + { + "epoch": 0.7570621468926554, + "grad_norm": 2.266413927078247, + "learning_rate": 4.183479253901789e-06, + "loss": 1.4402, + "step": 737 + }, + { + "epoch": 0.7580893682588598, + "grad_norm": 2.2785308361053467, + "learning_rate": 4.181575942139323e-06, + "loss": 1.4874, + "step": 738 + }, + { + "epoch": 0.7591165896250642, + "grad_norm": 2.2710752487182617, + "learning_rate": 4.179672630376856e-06, + "loss": 1.4644, + "step": 739 + }, + { + "epoch": 0.7601438109912686, + "grad_norm": 2.2460498809814453, + "learning_rate": 4.177769318614389e-06, + "loss": 1.5267, + "step": 740 + }, + { + "epoch": 0.7601438109912686, + "eval_loss": 1.8187947273254395, + "eval_runtime": 37.9464, + "eval_samples_per_second": 26.353, + "eval_steps_per_second": 1.107, + "step": 740 + }, + { + "epoch": 0.7611710323574731, + "grad_norm": 2.3589370250701904, + "learning_rate": 4.175866006851923e-06, + "loss": 1.5115, + "step": 741 + }, + { + "epoch": 0.7621982537236774, + "grad_norm": 2.2818758487701416, + "learning_rate": 4.173962695089456e-06, + "loss": 1.4945, + "step": 742 + }, + { + "epoch": 0.7632254750898819, + "grad_norm": 2.338085889816284, + "learning_rate": 4.1720593833269894e-06, + "loss": 1.5076, + "step": 743 + }, + { + "epoch": 0.7642526964560863, + "grad_norm": 2.3874809741973877, + "learning_rate": 4.170156071564523e-06, + "loss": 1.5665, + "step": 744 + }, + { + "epoch": 0.7652799178222907, + "grad_norm": 2.2418739795684814, + "learning_rate": 4.168252759802056e-06, + "loss": 1.4827, + "step": 745 + }, + { + "epoch": 0.7663071391884951, + "grad_norm": 2.441315174102783, + "learning_rate": 4.1663494480395895e-06, + "loss": 1.5171, + "step": 746 + }, + { + "epoch": 0.7673343605546995, + "grad_norm": 2.3201963901519775, + "learning_rate": 4.164446136277123e-06, + "loss": 1.5153, + "step": 747 + }, + { + "epoch": 0.768361581920904, + "grad_norm": 2.3395135402679443, + "learning_rate": 4.162542824514656e-06, + "loss": 1.4911, + "step": 748 + }, + { + "epoch": 0.7693888032871083, + "grad_norm": 2.3455941677093506, + "learning_rate": 4.16063951275219e-06, + "loss": 1.4493, + "step": 749 + }, + { + "epoch": 0.7704160246533128, + "grad_norm": 2.3737382888793945, + "learning_rate": 4.158736200989722e-06, + "loss": 1.4893, + "step": 750 + }, + { + "epoch": 0.7704160246533128, + "eval_loss": 1.8153141736984253, + "eval_runtime": 37.9216, + "eval_samples_per_second": 26.37, + "eval_steps_per_second": 1.108, + "step": 750 + }, + { + "epoch": 0.7714432460195172, + "grad_norm": 2.503803014755249, + "learning_rate": 4.1568328892272555e-06, + "loss": 1.499, + "step": 751 + }, + { + "epoch": 0.7724704673857217, + "grad_norm": 2.469083547592163, + "learning_rate": 4.154929577464789e-06, + "loss": 1.4984, + "step": 752 + }, + { + "epoch": 0.773497688751926, + "grad_norm": 2.387991428375244, + "learning_rate": 4.153026265702322e-06, + "loss": 1.4283, + "step": 753 + }, + { + "epoch": 0.7745249101181304, + "grad_norm": 2.3355393409729004, + "learning_rate": 4.1511229539398556e-06, + "loss": 1.4869, + "step": 754 + }, + { + "epoch": 0.7755521314843349, + "grad_norm": 2.629345417022705, + "learning_rate": 4.149219642177389e-06, + "loss": 1.4885, + "step": 755 + }, + { + "epoch": 0.7765793528505393, + "grad_norm": 2.458944797515869, + "learning_rate": 4.147316330414922e-06, + "loss": 1.415, + "step": 756 + }, + { + "epoch": 0.7776065742167437, + "grad_norm": 2.40051531791687, + "learning_rate": 4.145413018652456e-06, + "loss": 1.5129, + "step": 757 + }, + { + "epoch": 0.7786337955829481, + "grad_norm": 2.5249011516571045, + "learning_rate": 4.143509706889989e-06, + "loss": 1.5002, + "step": 758 + }, + { + "epoch": 0.7796610169491526, + "grad_norm": 2.413422107696533, + "learning_rate": 4.141606395127522e-06, + "loss": 1.4837, + "step": 759 + }, + { + "epoch": 0.7806882383153569, + "grad_norm": 2.3438029289245605, + "learning_rate": 4.139703083365056e-06, + "loss": 1.5213, + "step": 760 + }, + { + "epoch": 0.7806882383153569, + "eval_loss": 1.8037596940994263, + "eval_runtime": 37.9229, + "eval_samples_per_second": 26.369, + "eval_steps_per_second": 1.108, + "step": 760 + }, + { + "epoch": 0.7817154596815614, + "grad_norm": 2.326828718185425, + "learning_rate": 4.137799771602588e-06, + "loss": 1.5031, + "step": 761 + }, + { + "epoch": 0.7827426810477658, + "grad_norm": 2.4902493953704834, + "learning_rate": 4.135896459840122e-06, + "loss": 1.4602, + "step": 762 + }, + { + "epoch": 0.7837699024139703, + "grad_norm": 2.488004684448242, + "learning_rate": 4.133993148077656e-06, + "loss": 1.4764, + "step": 763 + }, + { + "epoch": 0.7847971237801746, + "grad_norm": 2.5515761375427246, + "learning_rate": 4.132089836315189e-06, + "loss": 1.4946, + "step": 764 + }, + { + "epoch": 0.785824345146379, + "grad_norm": 2.556278705596924, + "learning_rate": 4.1301865245527226e-06, + "loss": 1.4338, + "step": 765 + }, + { + "epoch": 0.7868515665125835, + "grad_norm": 2.3017969131469727, + "learning_rate": 4.128283212790256e-06, + "loss": 1.4053, + "step": 766 + }, + { + "epoch": 0.7878787878787878, + "grad_norm": 2.4371304512023926, + "learning_rate": 4.1263799010277884e-06, + "loss": 1.5238, + "step": 767 + }, + { + "epoch": 0.7889060092449923, + "grad_norm": 2.6256046295166016, + "learning_rate": 4.124476589265322e-06, + "loss": 1.4672, + "step": 768 + }, + { + "epoch": 0.7899332306111967, + "grad_norm": 2.3725669384002686, + "learning_rate": 4.122573277502855e-06, + "loss": 1.4677, + "step": 769 + }, + { + "epoch": 0.7909604519774012, + "grad_norm": 2.2928454875946045, + "learning_rate": 4.1206699657403885e-06, + "loss": 1.4927, + "step": 770 + }, + { + "epoch": 0.7909604519774012, + "eval_loss": 1.792726755142212, + "eval_runtime": 37.9273, + "eval_samples_per_second": 26.366, + "eval_steps_per_second": 1.107, + "step": 770 + }, + { + "epoch": 0.7919876733436055, + "grad_norm": 2.315674066543579, + "learning_rate": 4.118766653977922e-06, + "loss": 1.4632, + "step": 771 + }, + { + "epoch": 0.79301489470981, + "grad_norm": 2.418849468231201, + "learning_rate": 4.116863342215455e-06, + "loss": 1.446, + "step": 772 + }, + { + "epoch": 0.7940421160760144, + "grad_norm": 2.3209879398345947, + "learning_rate": 4.114960030452989e-06, + "loss": 1.4755, + "step": 773 + }, + { + "epoch": 0.7950693374422187, + "grad_norm": 2.2322115898132324, + "learning_rate": 4.113056718690522e-06, + "loss": 1.4149, + "step": 774 + }, + { + "epoch": 0.7960965588084232, + "grad_norm": 2.6517558097839355, + "learning_rate": 4.111153406928055e-06, + "loss": 1.5323, + "step": 775 + }, + { + "epoch": 0.7971237801746276, + "grad_norm": 2.4602737426757812, + "learning_rate": 4.109250095165589e-06, + "loss": 1.5179, + "step": 776 + }, + { + "epoch": 0.7981510015408321, + "grad_norm": 2.4015536308288574, + "learning_rate": 4.107346783403122e-06, + "loss": 1.4711, + "step": 777 + }, + { + "epoch": 0.7991782229070364, + "grad_norm": 2.4367337226867676, + "learning_rate": 4.105443471640655e-06, + "loss": 1.4425, + "step": 778 + }, + { + "epoch": 0.8002054442732409, + "grad_norm": 2.5161020755767822, + "learning_rate": 4.103540159878188e-06, + "loss": 1.4832, + "step": 779 + }, + { + "epoch": 0.8012326656394453, + "grad_norm": 2.483154773712158, + "learning_rate": 4.101636848115721e-06, + "loss": 1.4508, + "step": 780 + }, + { + "epoch": 0.8012326656394453, + "eval_loss": 1.792397379875183, + "eval_runtime": 37.8937, + "eval_samples_per_second": 26.39, + "eval_steps_per_second": 1.108, + "step": 780 + }, + { + "epoch": 0.8022598870056498, + "grad_norm": 2.4649806022644043, + "learning_rate": 4.099733536353255e-06, + "loss": 1.4919, + "step": 781 + }, + { + "epoch": 0.8032871083718541, + "grad_norm": 2.578634262084961, + "learning_rate": 4.097830224590788e-06, + "loss": 1.565, + "step": 782 + }, + { + "epoch": 0.8043143297380585, + "grad_norm": 2.444427967071533, + "learning_rate": 4.095926912828322e-06, + "loss": 1.5309, + "step": 783 + }, + { + "epoch": 0.805341551104263, + "grad_norm": 2.32051420211792, + "learning_rate": 4.094023601065855e-06, + "loss": 1.4777, + "step": 784 + }, + { + "epoch": 0.8063687724704673, + "grad_norm": 2.3207180500030518, + "learning_rate": 4.092120289303388e-06, + "loss": 1.4863, + "step": 785 + }, + { + "epoch": 0.8073959938366718, + "grad_norm": 2.1404120922088623, + "learning_rate": 4.0902169775409215e-06, + "loss": 1.437, + "step": 786 + }, + { + "epoch": 0.8084232152028762, + "grad_norm": 2.2980856895446777, + "learning_rate": 4.088313665778455e-06, + "loss": 1.4571, + "step": 787 + }, + { + "epoch": 0.8094504365690807, + "grad_norm": 2.2555243968963623, + "learning_rate": 4.086410354015988e-06, + "loss": 1.4451, + "step": 788 + }, + { + "epoch": 0.810477657935285, + "grad_norm": 2.3837077617645264, + "learning_rate": 4.0845070422535216e-06, + "loss": 1.4641, + "step": 789 + }, + { + "epoch": 0.8115048793014895, + "grad_norm": 2.5199060440063477, + "learning_rate": 4.082603730491055e-06, + "loss": 1.5048, + "step": 790 + }, + { + "epoch": 0.8115048793014895, + "eval_loss": 1.8019492626190186, + "eval_runtime": 37.8805, + "eval_samples_per_second": 26.399, + "eval_steps_per_second": 1.109, + "step": 790 + }, + { + "epoch": 0.8125321006676939, + "grad_norm": 2.4010138511657715, + "learning_rate": 4.080700418728588e-06, + "loss": 1.4881, + "step": 791 + }, + { + "epoch": 0.8135593220338984, + "grad_norm": 2.342094898223877, + "learning_rate": 4.078797106966122e-06, + "loss": 1.3792, + "step": 792 + }, + { + "epoch": 0.8145865434001027, + "grad_norm": 2.276772975921631, + "learning_rate": 4.076893795203655e-06, + "loss": 1.4539, + "step": 793 + }, + { + "epoch": 0.8156137647663071, + "grad_norm": 2.569275140762329, + "learning_rate": 4.074990483441188e-06, + "loss": 1.5059, + "step": 794 + }, + { + "epoch": 0.8166409861325116, + "grad_norm": 2.345761299133301, + "learning_rate": 4.073087171678721e-06, + "loss": 1.494, + "step": 795 + }, + { + "epoch": 0.8176682074987159, + "grad_norm": 2.4425504207611084, + "learning_rate": 4.071183859916254e-06, + "loss": 1.471, + "step": 796 + }, + { + "epoch": 0.8186954288649204, + "grad_norm": 2.5758635997772217, + "learning_rate": 4.069280548153788e-06, + "loss": 1.4378, + "step": 797 + }, + { + "epoch": 0.8197226502311248, + "grad_norm": 2.4320778846740723, + "learning_rate": 4.067377236391321e-06, + "loss": 1.474, + "step": 798 + }, + { + "epoch": 0.8207498715973293, + "grad_norm": 2.466978073120117, + "learning_rate": 4.065473924628854e-06, + "loss": 1.4997, + "step": 799 + }, + { + "epoch": 0.8217770929635336, + "grad_norm": 2.462489128112793, + "learning_rate": 4.063570612866388e-06, + "loss": 1.5, + "step": 800 + }, + { + "epoch": 0.8217770929635336, + "eval_loss": 1.783312201499939, + "eval_runtime": 37.8832, + "eval_samples_per_second": 26.397, + "eval_steps_per_second": 1.109, + "step": 800 + } + ], + "logging_steps": 1, + "max_steps": 2919, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.493847527489536e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/training_args.bin b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..01f19febcdb70137f8174ed0cddf6d18dec7535a --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4beb369b5d9b2e313f485f5348c5624fda1578b4b4b839a45918f8f15a4fbd4b +size 6011 diff --git a/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/zero_to_fp32.py b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B-FTP/babylm_reverse_full_10M_seed0/runs/checkpoint-800/zero_to_fp32.py @@ -0,0 +1,604 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: python zero_to_fp32.py . pytorch_model.bin + +import argparse +import torch +import glob +import math +import os +import re +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + + total_files = len(files) + state_dicts = [] + for f in files: + state_dict = torch.load(f, map_location=device) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + if zero_stage <= 2: + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + elif zero_stage == 3: + # if there is more than one param group, there will be multiple flattened tensors - one + # flattened tensor per group - for simplicity merge them into a single tensor + # + # XXX: could make the script more memory efficient for when there are multiple groups - it + # will require matching the sub-lists of param_shapes for each param group flattened tensor + + fp32_flat_groups = [ + torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts)) + ] + + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = fp32_flat_groups[0].numel() * world_size + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + for name, shape in param_shapes.items(): + + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # XXX: memory usage doubles here + state_dict[name] = torch.cat( + tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)), + 0).narrow(0, 0, unpartitioned_numel).view(shape) + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + + Returns: + - pytorch ``state_dict`` + + Note: this approach may not work if your application doesn't have sufficient free CPU memory and + you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters) + print(f"Saving fp32 state dict to {output_file}") + torch.save(state_dict, output_file) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info(f"Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info(f"Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument( + "output_file", + type=str, + help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_file, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters)