diff --git a/.gitattributes b/.gitattributes index 7b1565142805c267c99b7b8b399aefabb3fc8db0..d348beb74cf2699cf0315976bbee2cefa58069c7 100644 --- a/.gitattributes +++ b/.gitattributes @@ -571,3 +571,12 @@ checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-900/tokeni checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-100/tokenizer.json filter=lfs diff=lfs merge=lfs -text checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1700/tokenizer.json filter=lfs diff=lfs merge=lfs -text checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..fad444f8143e1badd2a8d15a75e619abd485ceab --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1100/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..39b6620f3f8d750109fe44053f80f2e221c6e290 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1100/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..04d30c4b88bb88b6860f56816d3db2c6c7742594 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1100/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..b0a59b3d69f7c73d7ad02040ef9ddf0af64846ba --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:500aa1e2983f7f455a3e8210712a5fb824056d0fb864457f35db87cedc4f15d7 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1100/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..9f4723485ff30930b59224e8afe0cb0b2b3c51bc --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1200/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..ca7b87736ad96984e06c88dc58b2e5efcdb20f2d --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5d60eb151191f050cb4b3590ac23ee7ff247807e742892633ea85518fbb160a +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1200/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..3429b5bbf5e65631c22532fed743338bede54d6b --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1300/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..cbba424cb2109da26bffbb01960eb52b2dc47b0d --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1300/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..b47aeb757c57446497f8f5d1181a50c3faf2dd79 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1300/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..7a30969f5d2b2999f9a40b5f8aa0f3c8ac393445 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ed4432bd76a70895d1eed9ffb9bcf3d55a555c7ee8371297518a9bf289187660 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1300/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..466c494b53a5cf3030ac585b29d1563c49b789d4 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1400/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..262d37671147896b5dfa4ab77f0e9578ea0d968c --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1400/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..b31f6e0c1e981c4c1dee9a4626e2a6c6ae5abb8e --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1400/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..ebaa878c7fbdf50f50d94d0f589fb9d9c78a45b7 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:444ee29b4620487bb6a50bae5508a70cfe811b44ca4f82431f50157295bb1f92 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1400/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..491d7aa6873ac345ad7d2d8238b3751df06748aa --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1500/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..0c1653c04de59e53b4d05673f04ed4b2c7039c73 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1500/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..4cf5bfa682e62dbe5c2e496536a84208dc8433d0 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1500/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..a94b34e1aa5087ed1129ca130601f722e8d7959d --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d470a1566fade71dbfe3de9e5fe45feb7e321a52bff45349224c90061e1661d9 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1500/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..5509f0e1ece25901639e38b63337f0f28e33b98d --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1600/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..bdaddc4981c690246dfa665b10da40b38a679bc3 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1600/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..fd1ca9202bb0004e7be63108d043e05db918daf3 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-1600/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..9206ddeeffec149ed3d4e16f994082d5aebc21d0 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ff2f11dbfd060ce016b21218b873a40ef74c7e31eecd585532ac412097f6fdc2 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1600/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..0b4ae0642e5f92f74c658999b194b5ef1fe11845 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8eeb05b2464265890d03970b06f167c82600ae380c5af274feadb43b529af8da +size 16567 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..7f17a148b6fb83bbdca37275527834750b348cea --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ba564af3daac2ee124ab61d7e40b0c6a31626027ac994a15a45b4c2ff14ad78 +size 16567 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..af9aae0e9d9650fb5e117bc169005317e2f3c5b1 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7e9ef8a3b92a4a920eec9610cc2d3ae5914e60cee13c62730935b1fecc7ec2d +size 16567 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f70248c467716c25b5bb51158c6e363983a8b666 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9e9dac98de75514ef8a2170470529bcca9086f59ac0a6a3251d108795f3e16a +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-1900/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..d5b3eaa75e01e0564b8641c8f89a308c359a8a6c --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-400/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..f31c91e6fc0dcc1eb5c101a4c40e35f5b315d26f --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-400/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..c1492af50e3afb56d1aecd7362d9e876ea1c01da --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-400/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..207bc81dcbfdea224b4e9a71bb746a20689aab5b --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5340761043b7ad61c7ba658f6fc46681f4d609c5ac3728de2c2013a9e0ce4e79 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-400/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..d5f37008b65b2b6b774ba8ca3e08ea7f2886beeb --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-600/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..43698f33ac8bdb91266f12645eda8b36787da828 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-600/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..90d0b3f139f4cb1368a680fad6f12df4c66f078f --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-600/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..aeba18b3f74be376a645e91e43000677785c02d2 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:030ba726537291b026d46325eab79e8e9916823e9059bf1870c5c7ab595e53e3 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-600/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/rng_state_0.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..10b2f5e4f6c9f32023941f5b8cf39645a86289fd --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-800/rng_state_1.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..1f46eec58ec6affe22de910a8d659fe43a6018f4 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-800/rng_state_2.pth b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..89cef672a6b7edc3804190d221d1a77cd6a2d146 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_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/babylm_hop_control_10M_seed0/runs/checkpoint-800/scheduler.pt b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..69bf883712a5fc210130703c570147f01420987e --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b764de8f0a7386632694b8b1ee7ba7aeae3190be43916b4b86dbeaf03deefaa2 +size 627 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/tokenizer.json b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..f563348f854ed1af8a8e9ff3391872f8d994ff83 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2119a48316b293375cf214819b7f053f5f12171eb3ee8dccfce26510f91e2c +size 17210462 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/training_args.bin b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d28b7c159ad5567421a03464d70e00a1e2d43361 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_control_10M_seed0/runs/checkpoint-800/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b717b63c99e15a7061bfb9ad0d8db9718415d88d6419d3c3dc4217194b1c01 +size 6011 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-100/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-100/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..deccf1a1ed1cabf2c883d5eec2c3903b83d37cb3 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-100/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22161289a9777b5d57f4f55f90963425b884b52601ca9c049dc1ccf909285183 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1000/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8f2751e2602ac0b795130e3bec8a0334c238ccd3 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8342679c2955c880b9b5628ca7cd082be6b0329af3eada87f66e490fb4098bba +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1100/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1100/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d7a4ed9ce6dd8553bb040d48fdf988a1371a858c --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1100/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:58ca12daa79bc9fdf0aa3a30061686ac1527bb63432a41a1136e47ec3dcebc83 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1200/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1200/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5ebe59e905495ee2e286ac21fa177a942746ca53 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1200/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c6210823d3ac3cf0c1ada216f8f648ce45367f36033020bea63258d35e0a6386 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1400/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1400/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..288a6cb0241ae5705d5d71e9c794193fdf0b55c0 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1400/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0879851eaeb57cc468b730cd0cf04a7da6f347c4ec8b849081ee8cf65a007d4e +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1600/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1600/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ff160aabba23da8cdc4930802363ef76a9ae03be --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1600/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:421fe64331759a6fda5d4b4c30be0975afd0732acbae5ce767673eb13000f131 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1700/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1700/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5f530d119a1fa8d356eb827af8488bba1a61c538 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-1700/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fea7000f4d3079ac59e55ec9a6aa0d7936614ed3ae961f11c432045ba4e30d5 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-200/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-200/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3fe6f1c55f9777cb887b97e0e80542e1446365c4 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-200/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9cde1c5853ec22e6fbd5cb5db63c9a77784eb9a3f45e865c710da3cb171f14fc +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2100/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2100/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a84512055dcfb480bbee5d974653699c36ff8166 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2100/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f04810d94397bb0b3e4fac5d8e112fcf21240b4abbf48a4ccb184bdeaba64ee4 +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2200/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2200/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..66257f0fa9e84b277f77b4965928cfd22d2638ec --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2200/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d2f004a39f845d41ba0cb724761038fb93a19953d154a9afe3ce1b3aecd1a89c +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2300/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2300/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..06070d426d27a2107d9c57f6ca1a6c7f16ee3873 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2300/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:879e831df4a11128162ac9263b355c8347d4c7a05ee1b8ecbabf13186a6c65ef +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2376/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2376/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5bbf4b60541e54dee2c113df4433e31c5c26bcf6 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-2376/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6ea0157e2099402d64495f793a161ff64fef02c995b5d4ac53f94a698572d44 +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-300/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-300/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3598a02f7fba84f394a439e1c1fe08885d08987e --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-300/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:79265f4c03cc9dd838baaf1c23245ee4beb0f655a3026b55c937e52976404a9c +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-500/model-00002-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-500/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5693382a059272196edac6a9d99237f1fa19c6ed --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-500/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:33ba1de65fe10b9ec5a251dd70dd271a6c92bbf6929a876583e131f72c790934 +size 2247734920 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-600/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-600/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..94cb2ffddc37169b9c41e55e91991dee1ed39bdb --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-600/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7dc95db6defb71dc60ecfc082a8b65915e5714d61d22443f0ff302aca7765906 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-900/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-900/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a5ba52ad52cde7fdf340ea73c5c7a14483f300ef --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_hop_words4_10M_seed0/runs/checkpoint-900/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c6451fe949f92a213dce7187f29391712323c440d4134af9ce8a8491f29bb94 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2200/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2200/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..992737068c893a97f342d0dc2d1b264936ae44bb --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2200/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a788dff9289ae4c10266b0dffeeacf4c6e743446ffb5e3710112f07efa2ab95 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2300/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2300/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cf7b8878a45fda6928c3a7196ef3504ead8e1f06 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2300/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b77c69d6fd2700bf78c800e1442812488923a5f7ecb7522d960ad4197855128c +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2500/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2500/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c99c467ce8c181785a6b8a15a22a12a9b11768b2 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2500/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662dca6050ebd663d53ec6fbf8bb33cac3b5702b1db8bb9ed09de9c676f6f288 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2600/model-00001-of-00002.safetensors b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2600/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ba7124a9759b22c315867954a7b490c4e7f5b03f --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_reverse_full_10M_seed0/runs/checkpoint-2600/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1a3b7e0b99294386826724c4271ce3315f708e599cac59d341606b963594a463 +size 4965798912 diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/latest new file mode 100644 index 0000000000000000000000000000000000000000..22cd5c3402316b70299aed2025d7943595f5d495 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/latest @@ -0,0 +1 @@ +global_step1100 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..eccb6752e8f8c7517bc4d13307403b5d747bed80 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/trainer_state.json @@ -0,0 +1,8613 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.2054794520547945, + "eval_steps": 10, + "global_step": 1100, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6008102893829346, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7848, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.630127429962158, + "learning_rate": 4.362307067424858e-06, + "loss": 1.7135, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.720945358276367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7897, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.898836851119995, + "learning_rate": 4.358245329000813e-06, + "loss": 1.7434, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.6548235416412354, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7683, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.7218399047851562, + "learning_rate": 4.354183590576767e-06, + "loss": 1.7729, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.5972483158111572, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7517, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.6741726398468018, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7354, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.898418426513672, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7843, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.3964333534240723, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.808, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.9525212049484253, + "eval_runtime": 36.8297, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.9952592849731445, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7856, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.81740140914917, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7613, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6271541118621826, + "learning_rate": 4.339967506092607e-06, + "loss": 1.8409, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.540395498275757, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.7765, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.629131317138672, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7634, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.5796725749969482, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7414, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6980209350585938, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.8048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.658919095993042, + "learning_rate": 4.329813160032495e-06, + "loss": 1.7725, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5237927436828613, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7559, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.5822298526763916, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7129, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9492865800857544, + "eval_runtime": 37.0012, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.762810468673706, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7895, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.48427414894104, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7667, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.790052890777588, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7548, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.927957773208618, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.7489, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.4964418411254883, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7947, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.4845147132873535, + "learning_rate": 4.313566206336312e-06, + "loss": 1.7886, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.687361717224121, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7968, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.568202018737793, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7332, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.496429443359375, + "learning_rate": 4.307473598700244e-06, + "loss": 1.7583, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6063425540924072, + "learning_rate": 4.305442729488221e-06, + "loss": 1.779, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.9421359300613403, + "eval_runtime": 37.0712, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.5371274948120117, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7133, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.5601038932800293, + "learning_rate": 4.301380991064176e-06, + "loss": 1.8089, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.6244118213653564, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7983, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.4798192977905273, + "learning_rate": 4.29731925264013e-06, + "loss": 1.8221, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.483376979827881, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.8032, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.716745615005493, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7583, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.6685945987701416, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7365, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.465810537338257, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7682, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.7629129886627197, + "learning_rate": 4.287164906580017e-06, + "loss": 1.809, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5273563861846924, + "learning_rate": 4.285134037367994e-06, + "loss": 1.7464, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.9359965324401855, + "eval_runtime": 37.1317, + "eval_samples_per_second": 26.931, + "eval_steps_per_second": 1.131, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.514230966567993, + "learning_rate": 4.283103168155971e-06, + "loss": 1.7365, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.704042911529541, + "learning_rate": 4.281072298943948e-06, + "loss": 1.7268, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.4716262817382812, + "learning_rate": 4.279041429731926e-06, + "loss": 1.7212, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5695507526397705, + "learning_rate": 4.277010560519903e-06, + "loss": 1.7398, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.762845039367676, + "learning_rate": 4.27497969130788e-06, + "loss": 1.7158, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.571054220199585, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7685, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.908184289932251, + "learning_rate": 4.270917952883835e-06, + "loss": 1.7865, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.705479621887207, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7922, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.4418110847473145, + "learning_rate": 4.266856214459789e-06, + "loss": 1.7312, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.4619884490966797, + "learning_rate": 4.264825345247766e-06, + "loss": 1.8343, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.9352803230285645, + "eval_runtime": 37.0958, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5386202335357666, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6902, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9012045860290527, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7552, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.8028385639190674, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.6872, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.4505386352539062, + "learning_rate": 4.256701868399675e-06, + "loss": 1.75, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7551822662353516, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7612, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.596270799636841, + "learning_rate": 4.252640129975629e-06, + "loss": 1.7694, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.387662172317505, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.7776, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.7208738327026367, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7375, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.5317776203155518, + "learning_rate": 4.246547522339562e-06, + "loss": 1.6922, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.4803197383880615, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.741, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.9322715997695923, + "eval_runtime": 37.0962, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.5688276290893555, + "learning_rate": 4.242485783915517e-06, + "loss": 1.7518, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.4652810096740723, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6853, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.541903257369995, + "learning_rate": 4.23842404549147e-06, + "loss": 1.6841, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.377857208251953, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7312, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.418323040008545, + "learning_rate": 4.234362307067425e-06, + "loss": 1.6892, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5374410152435303, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7719, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7119228839874268, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.7819, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.5333240032196045, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7758, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.528696060180664, + "learning_rate": 4.226238830219334e-06, + "loss": 1.75, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.539039134979248, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7498, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.934200644493103, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.5384178161621094, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7615, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.601658821105957, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7645, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5359764099121094, + "learning_rate": 4.218115353371243e-06, + "loss": 1.7549, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.641118049621582, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7487, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.4146382808685303, + "learning_rate": 4.214053614947198e-06, + "loss": 1.7102, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4245245456695557, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7842, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.509824752807617, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6876, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.636444330215454, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7681, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.476858615875244, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7049, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.75732159614563, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7833, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.931411623954773, + "eval_runtime": 37.0994, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.5437941551208496, + "learning_rate": 4.201868399675061e-06, + "loss": 1.7751, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.5010673999786377, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7715, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.6352853775024414, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6798, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.588247776031494, + "learning_rate": 4.195775792038993e-06, + "loss": 1.7575, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.349921226501465, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7245, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.394432783126831, + "learning_rate": 4.191714053614948e-06, + "loss": 1.6778, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.883906602859497, + "learning_rate": 4.189683184402924e-06, + "loss": 1.7701, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.537799596786499, + "learning_rate": 4.187652315190902e-06, + "loss": 1.7286, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.575700283050537, + "learning_rate": 4.185621445978879e-06, + "loss": 1.7812, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.592252492904663, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7176, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.9259952306747437, + "eval_runtime": 37.144, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.511845111846924, + "learning_rate": 4.181559707554833e-06, + "loss": 1.7564, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.575446367263794, + "learning_rate": 4.179528838342811e-06, + "loss": 1.7943, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.379391670227051, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7098, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.701812982559204, + "learning_rate": 4.175467099918766e-06, + "loss": 1.7292, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.486640453338623, + "learning_rate": 4.173436230706742e-06, + "loss": 1.7963, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.561983823776245, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.7153, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.5709176063537598, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7412, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.3308839797973633, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7188, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.592470169067383, + "learning_rate": 4.165312753858652e-06, + "loss": 1.7599, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.3747010231018066, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7361, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.9229093790054321, + "eval_runtime": 37.1417, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.4062509536743164, + "learning_rate": 4.161251015434607e-06, + "loss": 1.7256, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.4951181411743164, + "learning_rate": 4.159220146222583e-06, + "loss": 1.7622, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.3453004360198975, + "learning_rate": 4.157189277010561e-06, + "loss": 1.6969, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.382671594619751, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7109, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4277336597442627, + "learning_rate": 4.153127538586516e-06, + "loss": 1.7471, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.4476678371429443, + "learning_rate": 4.151096669374492e-06, + "loss": 1.7483, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.451099395751953, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7366, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.618217706680298, + "learning_rate": 4.147034930950447e-06, + "loss": 1.7282, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.414400100708008, + "learning_rate": 4.145004061738425e-06, + "loss": 1.7409, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.535353422164917, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.7346, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.9240931272506714, + "eval_runtime": 36.844, + "eval_samples_per_second": 27.141, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.5075416564941406, + "learning_rate": 4.140942323314379e-06, + "loss": 1.6668, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.4136979579925537, + "learning_rate": 4.138911454102356e-06, + "loss": 1.696, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.4363577365875244, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7174, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.4578073024749756, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7621, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.370976686477661, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7443, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.363755702972412, + "learning_rate": 4.130787977254265e-06, + "loss": 1.7399, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.44704532623291, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7481, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.5463340282440186, + "learning_rate": 4.12672623883022e-06, + "loss": 1.7646, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7088072299957275, + "learning_rate": 4.124695369618197e-06, + "loss": 1.8108, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.463271141052246, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7311, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.918715238571167, + "eval_runtime": 37.0468, + "eval_samples_per_second": 26.993, + "eval_steps_per_second": 1.134, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6165266036987305, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.7295, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.3432133197784424, + "learning_rate": 4.118602761982129e-06, + "loss": 1.7512, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.552751302719116, + "learning_rate": 4.116571892770106e-06, + "loss": 1.6976, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7586560249328613, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6708, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.5326449871063232, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7824, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.7658956050872803, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7093, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.6868369579315186, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7475, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.543748140335083, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.7714, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.7235589027404785, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7593, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.697038173675537, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6296, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.91779363155365, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 2.6746180057525635, + "learning_rate": 4.100324939073924e-06, + "loss": 1.7971, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.6281094551086426, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7212, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.446025848388672, + "learning_rate": 4.096263200649879e-06, + "loss": 1.7324, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.5460824966430664, + "learning_rate": 4.094232331437856e-06, + "loss": 1.7435, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.5419771671295166, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6979, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.5545520782470703, + "learning_rate": 4.090170593013811e-06, + "loss": 1.7042, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.8353254795074463, + "learning_rate": 4.088139723801787e-06, + "loss": 1.7313, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.6521668434143066, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6901, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.6611177921295166, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7064, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.411482572555542, + "learning_rate": 4.08204711616572e-06, + "loss": 1.729, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.9176616668701172, + "eval_runtime": 37.1603, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.689110279083252, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7707, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.544018268585205, + "learning_rate": 4.077985377741674e-06, + "loss": 1.6724, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.55307936668396, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7677, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.3896875381469727, + "learning_rate": 4.073923639317629e-06, + "loss": 1.7729, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.381887197494507, + "learning_rate": 4.071892770105605e-06, + "loss": 1.7737, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.4873931407928467, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6928, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.4908432960510254, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6674, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.4346399307250977, + "learning_rate": 4.065800162469538e-06, + "loss": 1.6947, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6459081172943115, + "learning_rate": 4.063769293257514e-06, + "loss": 1.701, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.6092610359191895, + "learning_rate": 4.061738424045492e-06, + "loss": 1.7425, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.919799566268921, + "eval_runtime": 37.1632, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.4443421363830566, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7267, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.8252694606781006, + "learning_rate": 4.057676685621446e-06, + "loss": 1.7412, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.3391194343566895, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.7916, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.57979679107666, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7438, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.398383855819702, + "learning_rate": 4.051584077985378e-06, + "loss": 1.6606, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.394498348236084, + "learning_rate": 4.049553208773355e-06, + "loss": 1.736, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5070464611053467, + "learning_rate": 4.047522339561333e-06, + "loss": 1.7559, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.439722776412964, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7133, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.569152593612671, + "learning_rate": 4.043460601137287e-06, + "loss": 1.7123, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.637657642364502, + "learning_rate": 4.041429731925264e-06, + "loss": 1.8267, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.9148659706115723, + "eval_runtime": 37.2248, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.441269636154175, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7106, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.582172155380249, + "learning_rate": 4.037367993501219e-06, + "loss": 1.7167, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.5083518028259277, + "learning_rate": 4.035337124289196e-06, + "loss": 1.8031, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.504112958908081, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.694, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.5068695545196533, + "learning_rate": 4.031275385865151e-06, + "loss": 1.6798, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.600541114807129, + "learning_rate": 4.029244516653128e-06, + "loss": 1.7346, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.503020763397217, + "learning_rate": 4.027213647441105e-06, + "loss": 1.7507, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.7774016857147217, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6737, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.5771589279174805, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6943, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.3841609954833984, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7142, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.9146945476531982, + "eval_runtime": 37.1701, + "eval_samples_per_second": 26.903, + "eval_steps_per_second": 1.13, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.466207981109619, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.7192, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.72037935256958, + "learning_rate": 4.017059301380992e-06, + "loss": 1.8082, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.36811900138855, + "learning_rate": 4.015028432168969e-06, + "loss": 1.6964, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.4803504943847656, + "learning_rate": 4.012997562956945e-06, + "loss": 1.779, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.707174777984619, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6715, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.462543249130249, + "learning_rate": 4.0089358245329e-06, + "loss": 1.7117, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.481444835662842, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7007, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.4432222843170166, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.703, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.611387014389038, + "learning_rate": 4.002843216896833e-06, + "loss": 1.7722, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.456305742263794, + "learning_rate": 4.000812347684809e-06, + "loss": 1.7615, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.9089828729629517, + "eval_runtime": 37.1812, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4287772178649902, + "learning_rate": 3.998781478472787e-06, + "loss": 1.7098, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5659544467926025, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7143, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.467067003250122, + "learning_rate": 3.994719740048742e-06, + "loss": 1.6668, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.5660688877105713, + "learning_rate": 3.992688870836718e-06, + "loss": 1.8185, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.6185719966888428, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7356, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.5389695167541504, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6973, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.499234914779663, + "learning_rate": 3.98659626320065e-06, + "loss": 1.7539, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.6246273517608643, + "learning_rate": 3.984565393988627e-06, + "loss": 1.7055, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6127917766571045, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7177, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.4941787719726562, + "learning_rate": 3.980503655564582e-06, + "loss": 1.728, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.9063301086425781, + "eval_runtime": 37.3284, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.4694740772247314, + "learning_rate": 3.978472786352559e-06, + "loss": 1.7384, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.8288261890411377, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7409, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.518359661102295, + "learning_rate": 3.974411047928514e-06, + "loss": 1.714, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4874095916748047, + "learning_rate": 3.972380178716491e-06, + "loss": 1.781, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.523772716522217, + "learning_rate": 3.970349309504468e-06, + "loss": 1.7474, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.406132459640503, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.7486, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.388207197189331, + "learning_rate": 3.966287571080423e-06, + "loss": 1.67, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.5564754009246826, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6803, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.531890869140625, + "learning_rate": 3.962225832656377e-06, + "loss": 1.7368, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.404284954071045, + "learning_rate": 3.960194963444355e-06, + "loss": 1.7247, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.9008547067642212, + "eval_runtime": 37.1962, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.3903231620788574, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6465, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7126, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7174, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.6466639041900635, + "learning_rate": 3.954102355808286e-06, + "loss": 1.7166, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.530284881591797, + "learning_rate": 3.952071486596264e-06, + "loss": 1.7155, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.601516008377075, + "learning_rate": 3.950040617384241e-06, + "loss": 1.7157, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.6146240234375, + "learning_rate": 3.948009748172218e-06, + "loss": 1.6931, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.4904634952545166, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.7594, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.345689296722412, + "learning_rate": 3.943948009748173e-06, + "loss": 1.7175, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4469082355499268, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7161, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.9021154642105103, + "eval_runtime": 36.8772, + "eval_samples_per_second": 27.117, + "eval_steps_per_second": 1.139, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.473752737045288, + "learning_rate": 3.939886271324127e-06, + "loss": 1.7688, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.461033582687378, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6752, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.419179916381836, + "learning_rate": 3.935824532900082e-06, + "loss": 1.6685, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.543102502822876, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7348, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.4932174682617188, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.7572, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.4563698768615723, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6691, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.566558837890625, + "learning_rate": 3.927701056051991e-06, + "loss": 1.697, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.5550761222839355, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.7194, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.51362681388855, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6804, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.448714017868042, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6575, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.904768943786621, + "eval_runtime": 37.0644, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.582794666290283, + "learning_rate": 3.9195775792039e-06, + "loss": 1.717, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.613327980041504, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.6927, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.522038221359253, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7391, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.372134208679199, + "learning_rate": 3.913484971567831e-06, + "loss": 1.7448, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.246591329574585, + "learning_rate": 3.911454102355808e-06, + "loss": 1.6531, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.48276948928833, + "learning_rate": 3.909423233143786e-06, + "loss": 1.7268, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.7920989990234375, + "learning_rate": 3.907392363931763e-06, + "loss": 1.7155, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.5430562496185303, + "learning_rate": 3.90536149471974e-06, + "loss": 1.7866, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.5819969177246094, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.8024, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.726024866104126, + "learning_rate": 3.901299756295695e-06, + "loss": 1.8467, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.899052381515503, + "eval_runtime": 37.192, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.6913323402404785, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6976, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.5307703018188477, + "learning_rate": 3.897238017871649e-06, + "loss": 1.7072, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.4976935386657715, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.7183, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.435443878173828, + "learning_rate": 3.893176279447604e-06, + "loss": 1.6845, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.704554319381714, + "learning_rate": 3.891145410235581e-06, + "loss": 1.78, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.4770138263702393, + "learning_rate": 3.889114541023558e-06, + "loss": 1.7436, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.472780227661133, + "learning_rate": 3.887083671811536e-06, + "loss": 1.7007, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.4693336486816406, + "learning_rate": 3.885052802599513e-06, + "loss": 1.6487, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.458834648132324, + "learning_rate": 3.88302193338749e-06, + "loss": 1.7895, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.4430108070373535, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.72, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.9009006023406982, + "eval_runtime": 37.1906, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.6104867458343506, + "learning_rate": 3.878960194963445e-06, + "loss": 1.6938, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.5191142559051514, + "learning_rate": 3.876929325751422e-06, + "loss": 1.7528, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.3788414001464844, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6446, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.485631227493286, + "learning_rate": 3.872867587327377e-06, + "loss": 1.7087, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.3721091747283936, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6752, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4648489952087402, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6883, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.447054862976074, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.7109, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.70007061958313, + "learning_rate": 3.864744110479286e-06, + "loss": 1.7342, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.6336557865142822, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7308, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.807671070098877, + "learning_rate": 3.86068237205524e-06, + "loss": 1.7519, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8990603685379028, + "eval_runtime": 37.1643, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.7177038192749023, + "learning_rate": 3.858651502843217e-06, + "loss": 1.7065, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.570364236831665, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6529, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.590329170227051, + "learning_rate": 3.854589764419171e-06, + "loss": 1.6955, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.542722702026367, + "learning_rate": 3.852558895207149e-06, + "loss": 1.7803, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.6095776557922363, + "learning_rate": 3.850528025995126e-06, + "loss": 1.7267, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.550161838531494, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6767, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.5079615116119385, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6897, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.5113067626953125, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.7024, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.559811592102051, + "learning_rate": 3.842404549147035e-06, + "loss": 1.695, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.507901191711426, + "learning_rate": 3.840373679935013e-06, + "loss": 1.661, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8952044248580933, + "eval_runtime": 37.2303, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.288231134414673, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.7095, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.260040044784546, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6858, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.5524179935455322, + "learning_rate": 3.834281072298944e-06, + "loss": 1.7362, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.419013738632202, + "learning_rate": 3.832250203086921e-06, + "loss": 1.6803, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.5345773696899414, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7435, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.5771822929382324, + "learning_rate": 3.828188464662876e-06, + "loss": 1.7932, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.4880053997039795, + "learning_rate": 3.826157595450853e-06, + "loss": 1.711, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.596400022506714, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6569, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.620776891708374, + "learning_rate": 3.822095857026808e-06, + "loss": 1.7159, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.512397527694702, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6832, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8950546979904175, + "eval_runtime": 37.1526, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5925326347351074, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6977, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.3980817794799805, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.6997, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.4825663566589355, + "learning_rate": 3.813972380178717e-06, + "loss": 1.7063, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.535006284713745, + "learning_rate": 3.811941510966694e-06, + "loss": 1.7457, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.450657606124878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.6521, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.3942458629608154, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6194, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5781140327453613, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.7105, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.5196564197540283, + "learning_rate": 3.803818034118603e-06, + "loss": 1.6741, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4819812774658203, + "learning_rate": 3.80178716490658e-06, + "loss": 1.7559, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.6393206119537354, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.7421, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.8945845365524292, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5286684036254883, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6933, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.690138339996338, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.6804, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.7789907455444336, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7839, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.5196211338043213, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7564, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.550898790359497, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.758, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.617361545562744, + "learning_rate": 3.787571080422421e-06, + "loss": 1.7497, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.533339500427246, + "learning_rate": 3.785540211210398e-06, + "loss": 1.7336, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.475037097930908, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.712, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.485506534576416, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.6985, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.712912082672119, + "learning_rate": 3.77944760357433e-06, + "loss": 1.812, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8921257257461548, + "eval_runtime": 37.1309, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.6280291080474854, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.7504, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6162853240966797, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6793, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.665316581726074, + "learning_rate": 3.773354995938262e-06, + "loss": 1.7333, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.531297445297241, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7298, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.543412446975708, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.7677, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.3839635848999023, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.7039, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.3417305946350098, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6206, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.4434802532196045, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.713, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.5223472118377686, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6587, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.577479124069214, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.7032, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8891643285751343, + "eval_runtime": 37.1123, + "eval_samples_per_second": 26.945, + "eval_steps_per_second": 1.132, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.468871593475342, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.732, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.5907723903656006, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.6866, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.4454426765441895, + "learning_rate": 3.753046303818034e-06, + "loss": 1.746, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.4651715755462646, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7375, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6342053413391113, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.7693, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.454977512359619, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6796, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.440626859664917, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7075, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.4905781745910645, + "learning_rate": 3.742891957757921e-06, + "loss": 1.7422, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6843647956848145, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7107, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.5906362533569336, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.7545, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8889163732528687, + "eval_runtime": 36.8136, + "eval_samples_per_second": 27.164, + "eval_steps_per_second": 1.141, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.5185422897338867, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6857, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.4425783157348633, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.687, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.5200576782226562, + "learning_rate": 3.732737611697807e-06, + "loss": 1.643, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.4664227962493896, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.5965, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.495962381362915, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6663, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.5607051849365234, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.6388, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 2.9295718669891357, + "learning_rate": 3.724614134849716e-06, + "loss": 1.7541, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.5632426738739014, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6359, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.6059279441833496, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.7427, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.6244208812713623, + "learning_rate": 3.718521527213648e-06, + "loss": 1.7097, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8882938623428345, + "eval_runtime": 36.9853, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.434684991836548, + "learning_rate": 3.716490658001625e-06, + "loss": 1.6875, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.6481709480285645, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5726, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.5398213863372803, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.6881, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.684122085571289, + "learning_rate": 3.710398050365557e-06, + "loss": 1.691, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.712214946746826, + "learning_rate": 3.708367181153534e-06, + "loss": 1.653, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.8427517414093018, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.6853, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.676410436630249, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.645, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.5397610664367676, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6171, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.606250286102295, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6729, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.3617212772369385, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.6546, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8915373086929321, + "eval_runtime": 37.094, + "eval_samples_per_second": 26.959, + "eval_steps_per_second": 1.132, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.6635043621063232, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.628, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.724837303161621, + "learning_rate": 3.694151096669375e-06, + "loss": 1.7537, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.56447434425354, + "learning_rate": 3.692120227457352e-06, + "loss": 1.6638, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6624584197998047, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.6806, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.7147905826568604, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5887, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.683840751647949, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.6664, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.8141531944274902, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6744, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.6172802448272705, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6866, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.7853286266326904, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.7034, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.7121260166168213, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6573, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.883426308631897, + "eval_runtime": 37.1964, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.4775121212005615, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6651, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.5361502170562744, + "learning_rate": 3.673842404549147e-06, + "loss": 1.6524, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.7969970703125, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6598, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.88427996635437, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.6821, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.9321346282958984, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.7777, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.802886724472046, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6808, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.85810923576355, + "learning_rate": 3.663688058489034e-06, + "loss": 1.7375, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.6501998901367188, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.6662, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.550652265548706, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6084, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.5003185272216797, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.681, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8800761699676514, + "eval_runtime": 37.2014, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.4262983798980713, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6608, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6474270820617676, + "learning_rate": 3.65353371242892e-06, + "loss": 1.6972, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.5103557109832764, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5581, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.529756546020508, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.693, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.578362464904785, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6302, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8062336444854736, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6855, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.935694932937622, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6861, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5260093212127686, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6534, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.7627487182617188, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.6833, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.9553534984588623, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6887, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.8832038640975952, + "eval_runtime": 37.1828, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.5526084899902344, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.7162, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.761218786239624, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6823, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.672189950942993, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6489, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.7650046348571777, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.7201, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.6351473331451416, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.6164, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5959131717681885, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.603, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.525513172149658, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6671, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.5529232025146484, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6446, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6846, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6786, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.885197639465332, + "eval_runtime": 37.2374, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.50577449798584, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6892, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.57772159576416, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5863, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.5455222129821777, + "learning_rate": 3.612916328188465e-06, + "loss": 1.692, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.6154329776763916, + "learning_rate": 3.610885458976442e-06, + "loss": 1.7027, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.5776023864746094, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.6353, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.5480730533599854, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.6358, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.5939579010009766, + "learning_rate": 3.604792851340374e-06, + "loss": 1.7099, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.65395188331604, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6763, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.5302748680114746, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.613, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.669433832168579, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.6584, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8816337585449219, + "eval_runtime": 37.2174, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.859063148498535, + "learning_rate": 3.596669374492283e-06, + "loss": 1.7096, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.531301259994507, + "learning_rate": 3.59463850528026e-06, + "loss": 1.6168, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.5920488834381104, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6953, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.698284387588501, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6602, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.762387752532959, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.7073, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.554579019546509, + "learning_rate": 3.586515028432169e-06, + "loss": 1.7151, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.644576072692871, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.6088, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.6375787258148193, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6434, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9475791454315186, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6936, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.5753467082977295, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5915, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8799077272415161, + "eval_runtime": 37.2335, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8163156509399414, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6824, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.6741368770599365, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.7066, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.5235214233398438, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6779, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.620927095413208, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6392, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7678003311157227, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6768, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.5900607109069824, + "learning_rate": 3.566206336311942e-06, + "loss": 1.618, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.4440128803253174, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5981, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.686119318008423, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6649, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.6912872791290283, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.6426, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.7249274253845215, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.6639, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8750032186508179, + "eval_runtime": 37.2319, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.750321626663208, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.7014, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.593047618865967, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6557, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.744931936264038, + "learning_rate": 3.551990251827783e-06, + "loss": 1.6257, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.4940085411071777, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6546, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.6905856132507324, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.6395, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.6744513511657715, + "learning_rate": 3.545897644191714e-06, + "loss": 1.6159, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6766908168792725, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6754, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.743103504180908, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.6273, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.886472463607788, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6836, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.823476791381836, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6475, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8762013912200928, + "eval_runtime": 36.8594, + "eval_samples_per_second": 27.13, + "eval_steps_per_second": 1.139, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.741985559463501, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6765, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.6321418285369873, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6997, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7359395027160645, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6529, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.610210657119751, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.68, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.7770233154296875, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6836, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.7146646976470947, + "learning_rate": 3.525588952071487e-06, + "loss": 1.699, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.782172679901123, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6941, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.577692747116089, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.6303, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.721348285675049, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.6381, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.825709342956543, + "learning_rate": 3.517465475223396e-06, + "loss": 1.6144, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8759764432907104, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.8446385860443115, + "learning_rate": 3.515434606011373e-06, + "loss": 1.685, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.70577335357666, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.7276, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.6624722480773926, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6735, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.7956180572509766, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6413, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.792559862136841, + "learning_rate": 3.507311129163282e-06, + "loss": 1.6035, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.7113680839538574, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6461, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.656461715698242, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.662, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.8650670051574707, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5952, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.8043570518493652, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6738, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.607752799987793, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6828, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8729931116104126, + "eval_runtime": 37.2275, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.5819315910339355, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.6108, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6494336128234863, + "learning_rate": 3.493095044679123e-06, + "loss": 1.7348, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.552762508392334, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5534, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.6580469608306885, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6675, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.922295331954956, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5819, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.7051825523376465, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6573, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.738743782043457, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6939, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.51552677154541, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.665, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.578549385070801, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6642, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.7047011852264404, + "learning_rate": 3.476848090982941e-06, + "loss": 1.6298, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8726528882980347, + "eval_runtime": 37.25, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.722877025604248, + "learning_rate": 3.474817221770918e-06, + "loss": 1.7522, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.464240074157715, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6876, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.6737945079803467, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.6453, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.7960164546966553, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.607, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.6985723972320557, + "learning_rate": 3.466693744922827e-06, + "loss": 1.6385, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.7561540603637695, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6753, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.828885555267334, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6402, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.653183698654175, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6889, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.5803189277648926, + "learning_rate": 3.458570268074736e-06, + "loss": 1.654, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.5868985652923584, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6366, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8656586408615112, + "eval_runtime": 37.2805, + "eval_samples_per_second": 26.824, + "eval_steps_per_second": 1.127, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.648036241531372, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.6019, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.5482382774353027, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6597, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.700294017791748, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.6075, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.9346179962158203, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6558, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.893852710723877, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6395, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.789057970046997, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6833, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.745492935180664, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.6266, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.7930359840393066, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5781, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.770575761795044, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6518, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 2.7367143630981445, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6982, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8708553314208984, + "eval_runtime": 37.2843, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.89182448387146, + "learning_rate": 3.434199837530463e-06, + "loss": 1.6587, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.5572729110717773, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5974, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8172996044158936, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6941, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.6480889320373535, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6902, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.654735803604126, + "learning_rate": 3.426076360682372e-06, + "loss": 1.7083, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.4182465076446533, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6494, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.6808230876922607, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5782, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.648282527923584, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6695, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.5920112133026123, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.7234, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.690538167953491, + "learning_rate": 3.415922014622259e-06, + "loss": 1.686, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8717143535614014, + "eval_runtime": 37.2714, + "eval_samples_per_second": 26.83, + "eval_steps_per_second": 1.127, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.672515392303467, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6883, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.8012726306915283, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6762, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.7551519870758057, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6694, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.687445878982544, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.6267, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.674445390701294, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6944, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.797832489013672, + "learning_rate": 3.403736799350122e-06, + "loss": 1.6735, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7843174934387207, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.6183, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.678809404373169, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.7336, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.748838424682617, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6499, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.5785276889801025, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6589, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8695456981658936, + "eval_runtime": 37.2493, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.7307844161987305, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5989, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.83357572555542, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6568, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.6178860664367676, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6654, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.699328899383545, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6657, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.7418158054351807, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6351, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.5589001178741455, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.6351, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.492795944213867, + "learning_rate": 3.381397238017872e-06, + "loss": 1.6188, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 2.8586132526397705, + "learning_rate": 3.379366368805849e-06, + "loss": 1.7127, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7773218154907227, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6529, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.6956539154052734, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.6566, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8692638874053955, + "eval_runtime": 37.2317, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.7406537532806396, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6482, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.746194362640381, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.6343, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.767970323562622, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6577, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.730640172958374, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5941, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.5661444664001465, + "learning_rate": 3.36515028432169e-06, + "loss": 1.6362, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.5915091037750244, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6526, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.7135071754455566, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6999, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.591201066970825, + "learning_rate": 3.359057676685622e-06, + "loss": 1.684, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.564439296722412, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5815, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.4480631351470947, + "learning_rate": 3.354995938261576e-06, + "loss": 1.6284, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8628013134002686, + "eval_runtime": 37.23, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1100 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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": 3.429040350298112e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1100/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4ad2097a74bd73cf481b1a8db7900e36a3f94d3c --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1200/trainer_state.json @@ -0,0 +1,9393 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.3150684931506849, + "eval_steps": 10, + "global_step": 1200, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6008102893829346, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7848, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.630127429962158, + "learning_rate": 4.362307067424858e-06, + "loss": 1.7135, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.720945358276367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7897, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.898836851119995, + "learning_rate": 4.358245329000813e-06, + "loss": 1.7434, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.6548235416412354, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7683, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.7218399047851562, + "learning_rate": 4.354183590576767e-06, + "loss": 1.7729, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.5972483158111572, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7517, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.6741726398468018, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7354, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.898418426513672, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7843, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.3964333534240723, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.808, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.9525212049484253, + "eval_runtime": 36.8297, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.9952592849731445, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7856, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.81740140914917, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7613, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6271541118621826, + "learning_rate": 4.339967506092607e-06, + "loss": 1.8409, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.540395498275757, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.7765, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.629131317138672, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7634, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.5796725749969482, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7414, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6980209350585938, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.8048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.658919095993042, + "learning_rate": 4.329813160032495e-06, + "loss": 1.7725, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5237927436828613, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7559, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.5822298526763916, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7129, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9492865800857544, + "eval_runtime": 37.0012, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.762810468673706, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7895, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.48427414894104, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7667, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.790052890777588, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7548, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.927957773208618, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.7489, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.4964418411254883, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7947, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.4845147132873535, + "learning_rate": 4.313566206336312e-06, + "loss": 1.7886, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.687361717224121, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7968, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.568202018737793, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7332, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.496429443359375, + "learning_rate": 4.307473598700244e-06, + "loss": 1.7583, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6063425540924072, + "learning_rate": 4.305442729488221e-06, + "loss": 1.779, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.9421359300613403, + "eval_runtime": 37.0712, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.5371274948120117, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7133, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.5601038932800293, + "learning_rate": 4.301380991064176e-06, + "loss": 1.8089, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.6244118213653564, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7983, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.4798192977905273, + "learning_rate": 4.29731925264013e-06, + "loss": 1.8221, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.483376979827881, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.8032, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.716745615005493, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7583, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.6685945987701416, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7365, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.465810537338257, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7682, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.7629129886627197, + "learning_rate": 4.287164906580017e-06, + "loss": 1.809, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5273563861846924, + "learning_rate": 4.285134037367994e-06, + "loss": 1.7464, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.9359965324401855, + "eval_runtime": 37.1317, + "eval_samples_per_second": 26.931, + "eval_steps_per_second": 1.131, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.514230966567993, + "learning_rate": 4.283103168155971e-06, + "loss": 1.7365, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.704042911529541, + "learning_rate": 4.281072298943948e-06, + "loss": 1.7268, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.4716262817382812, + "learning_rate": 4.279041429731926e-06, + "loss": 1.7212, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5695507526397705, + "learning_rate": 4.277010560519903e-06, + "loss": 1.7398, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.762845039367676, + "learning_rate": 4.27497969130788e-06, + "loss": 1.7158, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.571054220199585, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7685, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.908184289932251, + "learning_rate": 4.270917952883835e-06, + "loss": 1.7865, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.705479621887207, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7922, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.4418110847473145, + "learning_rate": 4.266856214459789e-06, + "loss": 1.7312, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.4619884490966797, + "learning_rate": 4.264825345247766e-06, + "loss": 1.8343, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.9352803230285645, + "eval_runtime": 37.0958, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5386202335357666, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6902, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9012045860290527, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7552, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.8028385639190674, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.6872, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.4505386352539062, + "learning_rate": 4.256701868399675e-06, + "loss": 1.75, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7551822662353516, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7612, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.596270799636841, + "learning_rate": 4.252640129975629e-06, + "loss": 1.7694, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.387662172317505, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.7776, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.7208738327026367, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7375, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.5317776203155518, + "learning_rate": 4.246547522339562e-06, + "loss": 1.6922, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.4803197383880615, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.741, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.9322715997695923, + "eval_runtime": 37.0962, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.5688276290893555, + "learning_rate": 4.242485783915517e-06, + "loss": 1.7518, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.4652810096740723, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6853, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.541903257369995, + "learning_rate": 4.23842404549147e-06, + "loss": 1.6841, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.377857208251953, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7312, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.418323040008545, + "learning_rate": 4.234362307067425e-06, + "loss": 1.6892, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5374410152435303, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7719, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7119228839874268, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.7819, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.5333240032196045, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7758, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.528696060180664, + "learning_rate": 4.226238830219334e-06, + "loss": 1.75, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.539039134979248, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7498, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.934200644493103, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.5384178161621094, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7615, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.601658821105957, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7645, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5359764099121094, + "learning_rate": 4.218115353371243e-06, + "loss": 1.7549, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.641118049621582, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7487, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.4146382808685303, + "learning_rate": 4.214053614947198e-06, + "loss": 1.7102, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4245245456695557, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7842, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.509824752807617, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6876, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.636444330215454, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7681, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.476858615875244, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7049, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.75732159614563, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7833, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.931411623954773, + "eval_runtime": 37.0994, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.5437941551208496, + "learning_rate": 4.201868399675061e-06, + "loss": 1.7751, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.5010673999786377, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7715, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.6352853775024414, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6798, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.588247776031494, + "learning_rate": 4.195775792038993e-06, + "loss": 1.7575, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.349921226501465, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7245, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.394432783126831, + "learning_rate": 4.191714053614948e-06, + "loss": 1.6778, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.883906602859497, + "learning_rate": 4.189683184402924e-06, + "loss": 1.7701, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.537799596786499, + "learning_rate": 4.187652315190902e-06, + "loss": 1.7286, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.575700283050537, + "learning_rate": 4.185621445978879e-06, + "loss": 1.7812, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.592252492904663, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7176, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.9259952306747437, + "eval_runtime": 37.144, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.511845111846924, + "learning_rate": 4.181559707554833e-06, + "loss": 1.7564, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.575446367263794, + "learning_rate": 4.179528838342811e-06, + "loss": 1.7943, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.379391670227051, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7098, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.701812982559204, + "learning_rate": 4.175467099918766e-06, + "loss": 1.7292, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.486640453338623, + "learning_rate": 4.173436230706742e-06, + "loss": 1.7963, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.561983823776245, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.7153, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.5709176063537598, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7412, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.3308839797973633, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7188, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.592470169067383, + "learning_rate": 4.165312753858652e-06, + "loss": 1.7599, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.3747010231018066, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7361, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.9229093790054321, + "eval_runtime": 37.1417, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.4062509536743164, + "learning_rate": 4.161251015434607e-06, + "loss": 1.7256, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.4951181411743164, + "learning_rate": 4.159220146222583e-06, + "loss": 1.7622, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.3453004360198975, + "learning_rate": 4.157189277010561e-06, + "loss": 1.6969, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.382671594619751, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7109, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4277336597442627, + "learning_rate": 4.153127538586516e-06, + "loss": 1.7471, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.4476678371429443, + "learning_rate": 4.151096669374492e-06, + "loss": 1.7483, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.451099395751953, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7366, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.618217706680298, + "learning_rate": 4.147034930950447e-06, + "loss": 1.7282, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.414400100708008, + "learning_rate": 4.145004061738425e-06, + "loss": 1.7409, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.535353422164917, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.7346, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.9240931272506714, + "eval_runtime": 36.844, + "eval_samples_per_second": 27.141, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.5075416564941406, + "learning_rate": 4.140942323314379e-06, + "loss": 1.6668, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.4136979579925537, + "learning_rate": 4.138911454102356e-06, + "loss": 1.696, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.4363577365875244, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7174, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.4578073024749756, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7621, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.370976686477661, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7443, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.363755702972412, + "learning_rate": 4.130787977254265e-06, + "loss": 1.7399, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.44704532623291, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7481, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.5463340282440186, + "learning_rate": 4.12672623883022e-06, + "loss": 1.7646, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7088072299957275, + "learning_rate": 4.124695369618197e-06, + "loss": 1.8108, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.463271141052246, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7311, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.918715238571167, + "eval_runtime": 37.0468, + "eval_samples_per_second": 26.993, + "eval_steps_per_second": 1.134, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6165266036987305, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.7295, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.3432133197784424, + "learning_rate": 4.118602761982129e-06, + "loss": 1.7512, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.552751302719116, + "learning_rate": 4.116571892770106e-06, + "loss": 1.6976, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7586560249328613, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6708, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.5326449871063232, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7824, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.7658956050872803, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7093, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.6868369579315186, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7475, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.543748140335083, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.7714, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.7235589027404785, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7593, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.697038173675537, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6296, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.91779363155365, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 2.6746180057525635, + "learning_rate": 4.100324939073924e-06, + "loss": 1.7971, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.6281094551086426, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7212, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.446025848388672, + "learning_rate": 4.096263200649879e-06, + "loss": 1.7324, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.5460824966430664, + "learning_rate": 4.094232331437856e-06, + "loss": 1.7435, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.5419771671295166, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6979, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.5545520782470703, + "learning_rate": 4.090170593013811e-06, + "loss": 1.7042, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.8353254795074463, + "learning_rate": 4.088139723801787e-06, + "loss": 1.7313, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.6521668434143066, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6901, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.6611177921295166, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7064, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.411482572555542, + "learning_rate": 4.08204711616572e-06, + "loss": 1.729, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.9176616668701172, + "eval_runtime": 37.1603, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.689110279083252, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7707, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.544018268585205, + "learning_rate": 4.077985377741674e-06, + "loss": 1.6724, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.55307936668396, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7677, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.3896875381469727, + "learning_rate": 4.073923639317629e-06, + "loss": 1.7729, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.381887197494507, + "learning_rate": 4.071892770105605e-06, + "loss": 1.7737, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.4873931407928467, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6928, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.4908432960510254, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6674, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.4346399307250977, + "learning_rate": 4.065800162469538e-06, + "loss": 1.6947, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6459081172943115, + "learning_rate": 4.063769293257514e-06, + "loss": 1.701, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.6092610359191895, + "learning_rate": 4.061738424045492e-06, + "loss": 1.7425, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.919799566268921, + "eval_runtime": 37.1632, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.4443421363830566, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7267, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.8252694606781006, + "learning_rate": 4.057676685621446e-06, + "loss": 1.7412, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.3391194343566895, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.7916, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.57979679107666, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7438, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.398383855819702, + "learning_rate": 4.051584077985378e-06, + "loss": 1.6606, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.394498348236084, + "learning_rate": 4.049553208773355e-06, + "loss": 1.736, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5070464611053467, + "learning_rate": 4.047522339561333e-06, + "loss": 1.7559, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.439722776412964, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7133, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.569152593612671, + "learning_rate": 4.043460601137287e-06, + "loss": 1.7123, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.637657642364502, + "learning_rate": 4.041429731925264e-06, + "loss": 1.8267, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.9148659706115723, + "eval_runtime": 37.2248, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.441269636154175, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7106, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.582172155380249, + "learning_rate": 4.037367993501219e-06, + "loss": 1.7167, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.5083518028259277, + "learning_rate": 4.035337124289196e-06, + "loss": 1.8031, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.504112958908081, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.694, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.5068695545196533, + "learning_rate": 4.031275385865151e-06, + "loss": 1.6798, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.600541114807129, + "learning_rate": 4.029244516653128e-06, + "loss": 1.7346, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.503020763397217, + "learning_rate": 4.027213647441105e-06, + "loss": 1.7507, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.7774016857147217, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6737, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.5771589279174805, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6943, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.3841609954833984, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7142, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.9146945476531982, + "eval_runtime": 37.1701, + "eval_samples_per_second": 26.903, + "eval_steps_per_second": 1.13, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.466207981109619, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.7192, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.72037935256958, + "learning_rate": 4.017059301380992e-06, + "loss": 1.8082, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.36811900138855, + "learning_rate": 4.015028432168969e-06, + "loss": 1.6964, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.4803504943847656, + "learning_rate": 4.012997562956945e-06, + "loss": 1.779, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.707174777984619, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6715, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.462543249130249, + "learning_rate": 4.0089358245329e-06, + "loss": 1.7117, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.481444835662842, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7007, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.4432222843170166, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.703, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.611387014389038, + "learning_rate": 4.002843216896833e-06, + "loss": 1.7722, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.456305742263794, + "learning_rate": 4.000812347684809e-06, + "loss": 1.7615, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.9089828729629517, + "eval_runtime": 37.1812, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4287772178649902, + "learning_rate": 3.998781478472787e-06, + "loss": 1.7098, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5659544467926025, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7143, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.467067003250122, + "learning_rate": 3.994719740048742e-06, + "loss": 1.6668, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.5660688877105713, + "learning_rate": 3.992688870836718e-06, + "loss": 1.8185, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.6185719966888428, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7356, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.5389695167541504, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6973, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.499234914779663, + "learning_rate": 3.98659626320065e-06, + "loss": 1.7539, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.6246273517608643, + "learning_rate": 3.984565393988627e-06, + "loss": 1.7055, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6127917766571045, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7177, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.4941787719726562, + "learning_rate": 3.980503655564582e-06, + "loss": 1.728, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.9063301086425781, + "eval_runtime": 37.3284, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.4694740772247314, + "learning_rate": 3.978472786352559e-06, + "loss": 1.7384, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.8288261890411377, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7409, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.518359661102295, + "learning_rate": 3.974411047928514e-06, + "loss": 1.714, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4874095916748047, + "learning_rate": 3.972380178716491e-06, + "loss": 1.781, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.523772716522217, + "learning_rate": 3.970349309504468e-06, + "loss": 1.7474, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.406132459640503, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.7486, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.388207197189331, + "learning_rate": 3.966287571080423e-06, + "loss": 1.67, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.5564754009246826, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6803, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.531890869140625, + "learning_rate": 3.962225832656377e-06, + "loss": 1.7368, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.404284954071045, + "learning_rate": 3.960194963444355e-06, + "loss": 1.7247, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.9008547067642212, + "eval_runtime": 37.1962, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.3903231620788574, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6465, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7126, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7174, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.6466639041900635, + "learning_rate": 3.954102355808286e-06, + "loss": 1.7166, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.530284881591797, + "learning_rate": 3.952071486596264e-06, + "loss": 1.7155, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.601516008377075, + "learning_rate": 3.950040617384241e-06, + "loss": 1.7157, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.6146240234375, + "learning_rate": 3.948009748172218e-06, + "loss": 1.6931, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.4904634952545166, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.7594, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.345689296722412, + "learning_rate": 3.943948009748173e-06, + "loss": 1.7175, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4469082355499268, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7161, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.9021154642105103, + "eval_runtime": 36.8772, + "eval_samples_per_second": 27.117, + "eval_steps_per_second": 1.139, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.473752737045288, + "learning_rate": 3.939886271324127e-06, + "loss": 1.7688, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.461033582687378, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6752, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.419179916381836, + "learning_rate": 3.935824532900082e-06, + "loss": 1.6685, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.543102502822876, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7348, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.4932174682617188, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.7572, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.4563698768615723, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6691, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.566558837890625, + "learning_rate": 3.927701056051991e-06, + "loss": 1.697, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.5550761222839355, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.7194, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.51362681388855, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6804, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.448714017868042, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6575, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.904768943786621, + "eval_runtime": 37.0644, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.582794666290283, + "learning_rate": 3.9195775792039e-06, + "loss": 1.717, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.613327980041504, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.6927, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.522038221359253, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7391, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.372134208679199, + "learning_rate": 3.913484971567831e-06, + "loss": 1.7448, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.246591329574585, + "learning_rate": 3.911454102355808e-06, + "loss": 1.6531, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.48276948928833, + "learning_rate": 3.909423233143786e-06, + "loss": 1.7268, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.7920989990234375, + "learning_rate": 3.907392363931763e-06, + "loss": 1.7155, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.5430562496185303, + "learning_rate": 3.90536149471974e-06, + "loss": 1.7866, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.5819969177246094, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.8024, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.726024866104126, + "learning_rate": 3.901299756295695e-06, + "loss": 1.8467, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.899052381515503, + "eval_runtime": 37.192, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.6913323402404785, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6976, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.5307703018188477, + "learning_rate": 3.897238017871649e-06, + "loss": 1.7072, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.4976935386657715, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.7183, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.435443878173828, + "learning_rate": 3.893176279447604e-06, + "loss": 1.6845, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.704554319381714, + "learning_rate": 3.891145410235581e-06, + "loss": 1.78, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.4770138263702393, + "learning_rate": 3.889114541023558e-06, + "loss": 1.7436, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.472780227661133, + "learning_rate": 3.887083671811536e-06, + "loss": 1.7007, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.4693336486816406, + "learning_rate": 3.885052802599513e-06, + "loss": 1.6487, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.458834648132324, + "learning_rate": 3.88302193338749e-06, + "loss": 1.7895, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.4430108070373535, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.72, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.9009006023406982, + "eval_runtime": 37.1906, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.6104867458343506, + "learning_rate": 3.878960194963445e-06, + "loss": 1.6938, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.5191142559051514, + "learning_rate": 3.876929325751422e-06, + "loss": 1.7528, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.3788414001464844, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6446, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.485631227493286, + "learning_rate": 3.872867587327377e-06, + "loss": 1.7087, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.3721091747283936, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6752, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4648489952087402, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6883, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.447054862976074, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.7109, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.70007061958313, + "learning_rate": 3.864744110479286e-06, + "loss": 1.7342, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.6336557865142822, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7308, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.807671070098877, + "learning_rate": 3.86068237205524e-06, + "loss": 1.7519, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8990603685379028, + "eval_runtime": 37.1643, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.7177038192749023, + "learning_rate": 3.858651502843217e-06, + "loss": 1.7065, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.570364236831665, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6529, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.590329170227051, + "learning_rate": 3.854589764419171e-06, + "loss": 1.6955, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.542722702026367, + "learning_rate": 3.852558895207149e-06, + "loss": 1.7803, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.6095776557922363, + "learning_rate": 3.850528025995126e-06, + "loss": 1.7267, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.550161838531494, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6767, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.5079615116119385, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6897, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.5113067626953125, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.7024, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.559811592102051, + "learning_rate": 3.842404549147035e-06, + "loss": 1.695, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.507901191711426, + "learning_rate": 3.840373679935013e-06, + "loss": 1.661, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8952044248580933, + "eval_runtime": 37.2303, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.288231134414673, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.7095, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.260040044784546, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6858, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.5524179935455322, + "learning_rate": 3.834281072298944e-06, + "loss": 1.7362, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.419013738632202, + "learning_rate": 3.832250203086921e-06, + "loss": 1.6803, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.5345773696899414, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7435, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.5771822929382324, + "learning_rate": 3.828188464662876e-06, + "loss": 1.7932, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.4880053997039795, + "learning_rate": 3.826157595450853e-06, + "loss": 1.711, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.596400022506714, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6569, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.620776891708374, + "learning_rate": 3.822095857026808e-06, + "loss": 1.7159, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.512397527694702, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6832, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8950546979904175, + "eval_runtime": 37.1526, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5925326347351074, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6977, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.3980817794799805, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.6997, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.4825663566589355, + "learning_rate": 3.813972380178717e-06, + "loss": 1.7063, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.535006284713745, + "learning_rate": 3.811941510966694e-06, + "loss": 1.7457, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.450657606124878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.6521, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.3942458629608154, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6194, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5781140327453613, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.7105, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.5196564197540283, + "learning_rate": 3.803818034118603e-06, + "loss": 1.6741, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4819812774658203, + "learning_rate": 3.80178716490658e-06, + "loss": 1.7559, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.6393206119537354, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.7421, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.8945845365524292, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5286684036254883, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6933, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.690138339996338, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.6804, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.7789907455444336, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7839, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.5196211338043213, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7564, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.550898790359497, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.758, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.617361545562744, + "learning_rate": 3.787571080422421e-06, + "loss": 1.7497, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.533339500427246, + "learning_rate": 3.785540211210398e-06, + "loss": 1.7336, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.475037097930908, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.712, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.485506534576416, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.6985, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.712912082672119, + "learning_rate": 3.77944760357433e-06, + "loss": 1.812, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8921257257461548, + "eval_runtime": 37.1309, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.6280291080474854, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.7504, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6162853240966797, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6793, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.665316581726074, + "learning_rate": 3.773354995938262e-06, + "loss": 1.7333, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.531297445297241, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7298, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.543412446975708, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.7677, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.3839635848999023, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.7039, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.3417305946350098, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6206, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.4434802532196045, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.713, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.5223472118377686, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6587, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.577479124069214, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.7032, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8891643285751343, + "eval_runtime": 37.1123, + "eval_samples_per_second": 26.945, + "eval_steps_per_second": 1.132, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.468871593475342, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.732, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.5907723903656006, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.6866, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.4454426765441895, + "learning_rate": 3.753046303818034e-06, + "loss": 1.746, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.4651715755462646, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7375, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6342053413391113, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.7693, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.454977512359619, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6796, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.440626859664917, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7075, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.4905781745910645, + "learning_rate": 3.742891957757921e-06, + "loss": 1.7422, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6843647956848145, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7107, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.5906362533569336, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.7545, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8889163732528687, + "eval_runtime": 36.8136, + "eval_samples_per_second": 27.164, + "eval_steps_per_second": 1.141, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.5185422897338867, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6857, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.4425783157348633, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.687, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.5200576782226562, + "learning_rate": 3.732737611697807e-06, + "loss": 1.643, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.4664227962493896, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.5965, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.495962381362915, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6663, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.5607051849365234, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.6388, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 2.9295718669891357, + "learning_rate": 3.724614134849716e-06, + "loss": 1.7541, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.5632426738739014, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6359, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.6059279441833496, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.7427, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.6244208812713623, + "learning_rate": 3.718521527213648e-06, + "loss": 1.7097, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8882938623428345, + "eval_runtime": 36.9853, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.434684991836548, + "learning_rate": 3.716490658001625e-06, + "loss": 1.6875, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.6481709480285645, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5726, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.5398213863372803, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.6881, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.684122085571289, + "learning_rate": 3.710398050365557e-06, + "loss": 1.691, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.712214946746826, + "learning_rate": 3.708367181153534e-06, + "loss": 1.653, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.8427517414093018, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.6853, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.676410436630249, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.645, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.5397610664367676, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6171, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.606250286102295, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6729, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.3617212772369385, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.6546, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8915373086929321, + "eval_runtime": 37.094, + "eval_samples_per_second": 26.959, + "eval_steps_per_second": 1.132, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.6635043621063232, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.628, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.724837303161621, + "learning_rate": 3.694151096669375e-06, + "loss": 1.7537, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.56447434425354, + "learning_rate": 3.692120227457352e-06, + "loss": 1.6638, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6624584197998047, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.6806, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.7147905826568604, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5887, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.683840751647949, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.6664, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.8141531944274902, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6744, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.6172802448272705, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6866, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.7853286266326904, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.7034, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.7121260166168213, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6573, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.883426308631897, + "eval_runtime": 37.1964, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.4775121212005615, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6651, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.5361502170562744, + "learning_rate": 3.673842404549147e-06, + "loss": 1.6524, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.7969970703125, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6598, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.88427996635437, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.6821, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.9321346282958984, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.7777, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.802886724472046, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6808, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.85810923576355, + "learning_rate": 3.663688058489034e-06, + "loss": 1.7375, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.6501998901367188, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.6662, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.550652265548706, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6084, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.5003185272216797, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.681, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8800761699676514, + "eval_runtime": 37.2014, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.4262983798980713, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6608, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6474270820617676, + "learning_rate": 3.65353371242892e-06, + "loss": 1.6972, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.5103557109832764, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5581, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.529756546020508, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.693, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.578362464904785, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6302, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8062336444854736, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6855, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.935694932937622, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6861, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5260093212127686, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6534, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.7627487182617188, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.6833, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.9553534984588623, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6887, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.8832038640975952, + "eval_runtime": 37.1828, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.5526084899902344, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.7162, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.761218786239624, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6823, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.672189950942993, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6489, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.7650046348571777, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.7201, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.6351473331451416, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.6164, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5959131717681885, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.603, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.525513172149658, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6671, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.5529232025146484, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6446, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6846, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6786, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.885197639465332, + "eval_runtime": 37.2374, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.50577449798584, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6892, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.57772159576416, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5863, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.5455222129821777, + "learning_rate": 3.612916328188465e-06, + "loss": 1.692, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.6154329776763916, + "learning_rate": 3.610885458976442e-06, + "loss": 1.7027, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.5776023864746094, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.6353, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.5480730533599854, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.6358, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.5939579010009766, + "learning_rate": 3.604792851340374e-06, + "loss": 1.7099, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.65395188331604, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6763, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.5302748680114746, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.613, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.669433832168579, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.6584, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8816337585449219, + "eval_runtime": 37.2174, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.859063148498535, + "learning_rate": 3.596669374492283e-06, + "loss": 1.7096, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.531301259994507, + "learning_rate": 3.59463850528026e-06, + "loss": 1.6168, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.5920488834381104, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6953, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.698284387588501, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6602, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.762387752532959, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.7073, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.554579019546509, + "learning_rate": 3.586515028432169e-06, + "loss": 1.7151, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.644576072692871, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.6088, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.6375787258148193, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6434, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9475791454315186, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6936, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.5753467082977295, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5915, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8799077272415161, + "eval_runtime": 37.2335, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8163156509399414, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6824, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.6741368770599365, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.7066, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.5235214233398438, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6779, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.620927095413208, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6392, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7678003311157227, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6768, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.5900607109069824, + "learning_rate": 3.566206336311942e-06, + "loss": 1.618, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.4440128803253174, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5981, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.686119318008423, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6649, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.6912872791290283, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.6426, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.7249274253845215, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.6639, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8750032186508179, + "eval_runtime": 37.2319, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.750321626663208, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.7014, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.593047618865967, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6557, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.744931936264038, + "learning_rate": 3.551990251827783e-06, + "loss": 1.6257, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.4940085411071777, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6546, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.6905856132507324, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.6395, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.6744513511657715, + "learning_rate": 3.545897644191714e-06, + "loss": 1.6159, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6766908168792725, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6754, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.743103504180908, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.6273, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.886472463607788, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6836, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.823476791381836, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6475, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8762013912200928, + "eval_runtime": 36.8594, + "eval_samples_per_second": 27.13, + "eval_steps_per_second": 1.139, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.741985559463501, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6765, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.6321418285369873, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6997, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7359395027160645, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6529, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.610210657119751, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.68, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.7770233154296875, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6836, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.7146646976470947, + "learning_rate": 3.525588952071487e-06, + "loss": 1.699, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.782172679901123, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6941, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.577692747116089, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.6303, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.721348285675049, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.6381, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.825709342956543, + "learning_rate": 3.517465475223396e-06, + "loss": 1.6144, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8759764432907104, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.8446385860443115, + "learning_rate": 3.515434606011373e-06, + "loss": 1.685, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.70577335357666, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.7276, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.6624722480773926, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6735, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.7956180572509766, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6413, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.792559862136841, + "learning_rate": 3.507311129163282e-06, + "loss": 1.6035, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.7113680839538574, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6461, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.656461715698242, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.662, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.8650670051574707, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5952, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.8043570518493652, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6738, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.607752799987793, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6828, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8729931116104126, + "eval_runtime": 37.2275, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.5819315910339355, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.6108, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6494336128234863, + "learning_rate": 3.493095044679123e-06, + "loss": 1.7348, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.552762508392334, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5534, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.6580469608306885, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6675, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.922295331954956, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5819, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.7051825523376465, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6573, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.738743782043457, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6939, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.51552677154541, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.665, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.578549385070801, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6642, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.7047011852264404, + "learning_rate": 3.476848090982941e-06, + "loss": 1.6298, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8726528882980347, + "eval_runtime": 37.25, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.722877025604248, + "learning_rate": 3.474817221770918e-06, + "loss": 1.7522, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.464240074157715, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6876, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.6737945079803467, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.6453, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.7960164546966553, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.607, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.6985723972320557, + "learning_rate": 3.466693744922827e-06, + "loss": 1.6385, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.7561540603637695, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6753, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.828885555267334, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6402, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.653183698654175, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6889, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.5803189277648926, + "learning_rate": 3.458570268074736e-06, + "loss": 1.654, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.5868985652923584, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6366, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8656586408615112, + "eval_runtime": 37.2805, + "eval_samples_per_second": 26.824, + "eval_steps_per_second": 1.127, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.648036241531372, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.6019, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.5482382774353027, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6597, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.700294017791748, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.6075, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.9346179962158203, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6558, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.893852710723877, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6395, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.789057970046997, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6833, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.745492935180664, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.6266, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.7930359840393066, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5781, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.770575761795044, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6518, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 2.7367143630981445, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6982, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8708553314208984, + "eval_runtime": 37.2843, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.89182448387146, + "learning_rate": 3.434199837530463e-06, + "loss": 1.6587, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.5572729110717773, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5974, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8172996044158936, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6941, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.6480889320373535, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6902, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.654735803604126, + "learning_rate": 3.426076360682372e-06, + "loss": 1.7083, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.4182465076446533, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6494, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.6808230876922607, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5782, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.648282527923584, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6695, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.5920112133026123, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.7234, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.690538167953491, + "learning_rate": 3.415922014622259e-06, + "loss": 1.686, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8717143535614014, + "eval_runtime": 37.2714, + "eval_samples_per_second": 26.83, + "eval_steps_per_second": 1.127, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.672515392303467, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6883, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.8012726306915283, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6762, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.7551519870758057, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6694, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.687445878982544, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.6267, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.674445390701294, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6944, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.797832489013672, + "learning_rate": 3.403736799350122e-06, + "loss": 1.6735, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7843174934387207, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.6183, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.678809404373169, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.7336, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.748838424682617, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6499, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.5785276889801025, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6589, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8695456981658936, + "eval_runtime": 37.2493, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.7307844161987305, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5989, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.83357572555542, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6568, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.6178860664367676, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6654, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.699328899383545, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6657, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.7418158054351807, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6351, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.5589001178741455, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.6351, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.492795944213867, + "learning_rate": 3.381397238017872e-06, + "loss": 1.6188, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 2.8586132526397705, + "learning_rate": 3.379366368805849e-06, + "loss": 1.7127, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7773218154907227, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6529, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.6956539154052734, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.6566, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8692638874053955, + "eval_runtime": 37.2317, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.7406537532806396, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6482, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.746194362640381, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.6343, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.767970323562622, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6577, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.730640172958374, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5941, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.5661444664001465, + "learning_rate": 3.36515028432169e-06, + "loss": 1.6362, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.5915091037750244, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6526, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.7135071754455566, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6999, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.591201066970825, + "learning_rate": 3.359057676685622e-06, + "loss": 1.684, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.564439296722412, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5815, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.4480631351470947, + "learning_rate": 3.354995938261576e-06, + "loss": 1.6284, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8628013134002686, + "eval_runtime": 37.23, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.5392649173736572, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5841, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6359004974365234, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.6278, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.803671360015869, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6602, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.6031086444854736, + "learning_rate": 3.346872461413485e-06, + "loss": 1.6021, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.687382459640503, + "learning_rate": 3.344841592201463e-06, + "loss": 1.6207, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.6388001441955566, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.6194, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.603121042251587, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6532, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.8053836822509766, + "learning_rate": 3.338748984565394e-06, + "loss": 1.68, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.5472421646118164, + "learning_rate": 3.336718115353372e-06, + "loss": 1.7164, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.5142838954925537, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6441, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.8628681898117065, + "eval_runtime": 36.8298, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.6917803287506104, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.6235, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.6062285900115967, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6488, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.5503156185150146, + "learning_rate": 3.32859463850528e-06, + "loss": 1.639, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.433518648147583, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5951, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.600661516189575, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.6148, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.5558395385742188, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.6334, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.6363396644592285, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5858, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.819617748260498, + "learning_rate": 3.318440292445167e-06, + "loss": 1.6274, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 2.9121081829071045, + "learning_rate": 3.316409423233144e-06, + "loss": 1.6081, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.734858989715576, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6889, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8644160032272339, + "eval_runtime": 37.0389, + "eval_samples_per_second": 26.999, + "eval_steps_per_second": 1.134, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.6940014362335205, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6463, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.82663631439209, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6474, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.634434223175049, + "learning_rate": 3.308285946385053e-06, + "loss": 1.7172, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.6721079349517822, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5795, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.786130666732788, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6634, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7698960304260254, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.687, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.6305580139160156, + "learning_rate": 3.300162469536962e-06, + "loss": 1.6372, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.5221288204193115, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6746, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.583643674850464, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6538, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.5978968143463135, + "learning_rate": 3.294069861900894e-06, + "loss": 1.7177, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8665071725845337, + "eval_runtime": 37.1194, + "eval_samples_per_second": 26.94, + "eval_steps_per_second": 1.131, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.6446492671966553, + "learning_rate": 3.292038992688871e-06, + "loss": 1.657, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.646174192428589, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.682, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.671532392501831, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.7033, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.616450309753418, + "learning_rate": 3.285946385052803e-06, + "loss": 1.586, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.552800416946411, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6599, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.5700533390045166, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5593, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.675553560256958, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.7097, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.6610019207000732, + "learning_rate": 3.277822908204712e-06, + "loss": 1.6408, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.55435848236084, + "learning_rate": 3.275792038992689e-06, + "loss": 1.5909, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.811556816101074, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6552, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.8618569374084473, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.7060115337371826, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.6144, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.6817731857299805, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6481, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.8206212520599365, + "learning_rate": 3.267668562144598e-06, + "loss": 1.6294, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.834332227706909, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.6103, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.754049062728882, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6788, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.6089437007904053, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6469, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.6095848083496094, + "learning_rate": 3.259545085296507e-06, + "loss": 1.7122, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.606787919998169, + "learning_rate": 3.257514216084484e-06, + "loss": 1.5851, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.643245220184326, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.7042, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.8400895595550537, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6526, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8640165328979492, + "eval_runtime": 37.162, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.519524335861206, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.6268, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.565365791320801, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6562, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.6364638805389404, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.6143, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.6235382556915283, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.64, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.7358779907226562, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6509, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.6720035076141357, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5902, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5960681438446045, + "learning_rate": 3.23923639317628e-06, + "loss": 1.6378, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.6655092239379883, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.6102, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.7873294353485107, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6558, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.746189594268799, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.7116, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8605948686599731, + "eval_runtime": 37.1486, + "eval_samples_per_second": 26.919, + "eval_steps_per_second": 1.131, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5058770179748535, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.6156, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.8002450466156006, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6884, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.696186065673828, + "learning_rate": 3.227051177904143e-06, + "loss": 1.6403, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.5473484992980957, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.6375, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.6391875743865967, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6597, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.647244453430176, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5666, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849031448364258, + "learning_rate": 3.218927701056052e-06, + "loss": 1.6211, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.7565410137176514, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5686, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.938270092010498, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.7046, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.6462385654449463, + "learning_rate": 3.212835093419984e-06, + "loss": 1.602, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.8577098846435547, + "eval_runtime": 37.1448, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.679330587387085, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.636, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.6410343647003174, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6904, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6398510932922363, + "learning_rate": 3.206742485783916e-06, + "loss": 1.597, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.4636199474334717, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5823, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.646580457687378, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6591, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.8582923412323, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.6731, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.5474820137023926, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5656, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6609606742858887, + "learning_rate": 3.196588139723802e-06, + "loss": 1.6374, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.651947021484375, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6557, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.6957671642303467, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.6436, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.8583322763442993, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.5442283153533936, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6424, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.593792200088501, + "learning_rate": 3.188464662875711e-06, + "loss": 1.7012, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.5021846294403076, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.6268, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.7368624210357666, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5953, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.5100302696228027, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5866, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.5767033100128174, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6621, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.4694652557373047, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6554, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.606158971786499, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.6238, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.857424020767212, + "learning_rate": 3.174248578391552e-06, + "loss": 1.626, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.6120405197143555, + "learning_rate": 3.172217709179529e-06, + "loss": 1.6134, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8599685430526733, + "eval_runtime": 37.1373, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.61995530128479, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.6303, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6253163814544678, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5886, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.627302885055542, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6526, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.8007607460021973, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.6581, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.551227331161499, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5921, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.7288761138916016, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6478, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5698401927948, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6552, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.796701192855835, + "learning_rate": 3.155970755483347e-06, + "loss": 1.6391, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.6523337364196777, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6495, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.5528745651245117, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6396, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8551454544067383, + "eval_runtime": 37.1721, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 1200 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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": 3.740771291234304e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/latest new file mode 100644 index 0000000000000000000000000000000000000000..d71c4c79bfb8e74842ee005fbfc27bedb903c39b --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/latest @@ -0,0 +1 @@ +global_step1300 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..c36fd791e8ed6b0839e8328968d8749d0ef1d77c --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/trainer_state.json @@ -0,0 +1,10173 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.4246575342465753, + "eval_steps": 10, + "global_step": 1300, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6008102893829346, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7848, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.630127429962158, + "learning_rate": 4.362307067424858e-06, + "loss": 1.7135, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.720945358276367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7897, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.898836851119995, + "learning_rate": 4.358245329000813e-06, + "loss": 1.7434, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.6548235416412354, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7683, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.7218399047851562, + "learning_rate": 4.354183590576767e-06, + "loss": 1.7729, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.5972483158111572, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7517, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.6741726398468018, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7354, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.898418426513672, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7843, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.3964333534240723, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.808, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.9525212049484253, + "eval_runtime": 36.8297, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.9952592849731445, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7856, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.81740140914917, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7613, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6271541118621826, + "learning_rate": 4.339967506092607e-06, + "loss": 1.8409, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.540395498275757, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.7765, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.629131317138672, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7634, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.5796725749969482, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7414, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6980209350585938, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.8048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.658919095993042, + "learning_rate": 4.329813160032495e-06, + "loss": 1.7725, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5237927436828613, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7559, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.5822298526763916, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7129, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9492865800857544, + "eval_runtime": 37.0012, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.762810468673706, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7895, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.48427414894104, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7667, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.790052890777588, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7548, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.927957773208618, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.7489, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.4964418411254883, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7947, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.4845147132873535, + "learning_rate": 4.313566206336312e-06, + "loss": 1.7886, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.687361717224121, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7968, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.568202018737793, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7332, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.496429443359375, + "learning_rate": 4.307473598700244e-06, + "loss": 1.7583, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6063425540924072, + "learning_rate": 4.305442729488221e-06, + "loss": 1.779, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.9421359300613403, + "eval_runtime": 37.0712, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.5371274948120117, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7133, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.5601038932800293, + "learning_rate": 4.301380991064176e-06, + "loss": 1.8089, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.6244118213653564, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7983, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.4798192977905273, + "learning_rate": 4.29731925264013e-06, + "loss": 1.8221, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.483376979827881, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.8032, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.716745615005493, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7583, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.6685945987701416, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7365, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.465810537338257, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7682, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.7629129886627197, + "learning_rate": 4.287164906580017e-06, + "loss": 1.809, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5273563861846924, + "learning_rate": 4.285134037367994e-06, + "loss": 1.7464, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.9359965324401855, + "eval_runtime": 37.1317, + "eval_samples_per_second": 26.931, + "eval_steps_per_second": 1.131, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.514230966567993, + "learning_rate": 4.283103168155971e-06, + "loss": 1.7365, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.704042911529541, + "learning_rate": 4.281072298943948e-06, + "loss": 1.7268, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.4716262817382812, + "learning_rate": 4.279041429731926e-06, + "loss": 1.7212, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5695507526397705, + "learning_rate": 4.277010560519903e-06, + "loss": 1.7398, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.762845039367676, + "learning_rate": 4.27497969130788e-06, + "loss": 1.7158, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.571054220199585, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7685, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.908184289932251, + "learning_rate": 4.270917952883835e-06, + "loss": 1.7865, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.705479621887207, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7922, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.4418110847473145, + "learning_rate": 4.266856214459789e-06, + "loss": 1.7312, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.4619884490966797, + "learning_rate": 4.264825345247766e-06, + "loss": 1.8343, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.9352803230285645, + "eval_runtime": 37.0958, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5386202335357666, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6902, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9012045860290527, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7552, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.8028385639190674, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.6872, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.4505386352539062, + "learning_rate": 4.256701868399675e-06, + "loss": 1.75, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7551822662353516, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7612, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.596270799636841, + "learning_rate": 4.252640129975629e-06, + "loss": 1.7694, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.387662172317505, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.7776, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.7208738327026367, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7375, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.5317776203155518, + "learning_rate": 4.246547522339562e-06, + "loss": 1.6922, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.4803197383880615, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.741, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.9322715997695923, + "eval_runtime": 37.0962, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.5688276290893555, + "learning_rate": 4.242485783915517e-06, + "loss": 1.7518, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.4652810096740723, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6853, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.541903257369995, + "learning_rate": 4.23842404549147e-06, + "loss": 1.6841, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.377857208251953, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7312, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.418323040008545, + "learning_rate": 4.234362307067425e-06, + "loss": 1.6892, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5374410152435303, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7719, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7119228839874268, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.7819, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.5333240032196045, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7758, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.528696060180664, + "learning_rate": 4.226238830219334e-06, + "loss": 1.75, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.539039134979248, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7498, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.934200644493103, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.5384178161621094, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7615, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.601658821105957, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7645, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5359764099121094, + "learning_rate": 4.218115353371243e-06, + "loss": 1.7549, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.641118049621582, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7487, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.4146382808685303, + "learning_rate": 4.214053614947198e-06, + "loss": 1.7102, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4245245456695557, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7842, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.509824752807617, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6876, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.636444330215454, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7681, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.476858615875244, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7049, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.75732159614563, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7833, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.931411623954773, + "eval_runtime": 37.0994, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.5437941551208496, + "learning_rate": 4.201868399675061e-06, + "loss": 1.7751, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.5010673999786377, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7715, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.6352853775024414, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6798, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.588247776031494, + "learning_rate": 4.195775792038993e-06, + "loss": 1.7575, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.349921226501465, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7245, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.394432783126831, + "learning_rate": 4.191714053614948e-06, + "loss": 1.6778, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.883906602859497, + "learning_rate": 4.189683184402924e-06, + "loss": 1.7701, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.537799596786499, + "learning_rate": 4.187652315190902e-06, + "loss": 1.7286, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.575700283050537, + "learning_rate": 4.185621445978879e-06, + "loss": 1.7812, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.592252492904663, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7176, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.9259952306747437, + "eval_runtime": 37.144, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.511845111846924, + "learning_rate": 4.181559707554833e-06, + "loss": 1.7564, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.575446367263794, + "learning_rate": 4.179528838342811e-06, + "loss": 1.7943, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.379391670227051, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7098, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.701812982559204, + "learning_rate": 4.175467099918766e-06, + "loss": 1.7292, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.486640453338623, + "learning_rate": 4.173436230706742e-06, + "loss": 1.7963, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.561983823776245, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.7153, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.5709176063537598, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7412, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.3308839797973633, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7188, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.592470169067383, + "learning_rate": 4.165312753858652e-06, + "loss": 1.7599, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.3747010231018066, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7361, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.9229093790054321, + "eval_runtime": 37.1417, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.4062509536743164, + "learning_rate": 4.161251015434607e-06, + "loss": 1.7256, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.4951181411743164, + "learning_rate": 4.159220146222583e-06, + "loss": 1.7622, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.3453004360198975, + "learning_rate": 4.157189277010561e-06, + "loss": 1.6969, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.382671594619751, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7109, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4277336597442627, + "learning_rate": 4.153127538586516e-06, + "loss": 1.7471, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.4476678371429443, + "learning_rate": 4.151096669374492e-06, + "loss": 1.7483, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.451099395751953, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7366, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.618217706680298, + "learning_rate": 4.147034930950447e-06, + "loss": 1.7282, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.414400100708008, + "learning_rate": 4.145004061738425e-06, + "loss": 1.7409, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.535353422164917, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.7346, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.9240931272506714, + "eval_runtime": 36.844, + "eval_samples_per_second": 27.141, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.5075416564941406, + "learning_rate": 4.140942323314379e-06, + "loss": 1.6668, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.4136979579925537, + "learning_rate": 4.138911454102356e-06, + "loss": 1.696, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.4363577365875244, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7174, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.4578073024749756, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7621, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.370976686477661, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7443, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.363755702972412, + "learning_rate": 4.130787977254265e-06, + "loss": 1.7399, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.44704532623291, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7481, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.5463340282440186, + "learning_rate": 4.12672623883022e-06, + "loss": 1.7646, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7088072299957275, + "learning_rate": 4.124695369618197e-06, + "loss": 1.8108, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.463271141052246, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7311, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.918715238571167, + "eval_runtime": 37.0468, + "eval_samples_per_second": 26.993, + "eval_steps_per_second": 1.134, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6165266036987305, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.7295, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.3432133197784424, + "learning_rate": 4.118602761982129e-06, + "loss": 1.7512, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.552751302719116, + "learning_rate": 4.116571892770106e-06, + "loss": 1.6976, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7586560249328613, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6708, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.5326449871063232, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7824, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.7658956050872803, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7093, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.6868369579315186, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7475, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.543748140335083, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.7714, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.7235589027404785, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7593, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.697038173675537, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6296, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.91779363155365, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 2.6746180057525635, + "learning_rate": 4.100324939073924e-06, + "loss": 1.7971, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.6281094551086426, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7212, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.446025848388672, + "learning_rate": 4.096263200649879e-06, + "loss": 1.7324, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.5460824966430664, + "learning_rate": 4.094232331437856e-06, + "loss": 1.7435, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.5419771671295166, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6979, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.5545520782470703, + "learning_rate": 4.090170593013811e-06, + "loss": 1.7042, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.8353254795074463, + "learning_rate": 4.088139723801787e-06, + "loss": 1.7313, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.6521668434143066, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6901, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.6611177921295166, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7064, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.411482572555542, + "learning_rate": 4.08204711616572e-06, + "loss": 1.729, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.9176616668701172, + "eval_runtime": 37.1603, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.689110279083252, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7707, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.544018268585205, + "learning_rate": 4.077985377741674e-06, + "loss": 1.6724, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.55307936668396, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7677, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.3896875381469727, + "learning_rate": 4.073923639317629e-06, + "loss": 1.7729, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.381887197494507, + "learning_rate": 4.071892770105605e-06, + "loss": 1.7737, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.4873931407928467, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6928, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.4908432960510254, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6674, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.4346399307250977, + "learning_rate": 4.065800162469538e-06, + "loss": 1.6947, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6459081172943115, + "learning_rate": 4.063769293257514e-06, + "loss": 1.701, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.6092610359191895, + "learning_rate": 4.061738424045492e-06, + "loss": 1.7425, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.919799566268921, + "eval_runtime": 37.1632, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.4443421363830566, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7267, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.8252694606781006, + "learning_rate": 4.057676685621446e-06, + "loss": 1.7412, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.3391194343566895, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.7916, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.57979679107666, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7438, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.398383855819702, + "learning_rate": 4.051584077985378e-06, + "loss": 1.6606, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.394498348236084, + "learning_rate": 4.049553208773355e-06, + "loss": 1.736, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5070464611053467, + "learning_rate": 4.047522339561333e-06, + "loss": 1.7559, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.439722776412964, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7133, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.569152593612671, + "learning_rate": 4.043460601137287e-06, + "loss": 1.7123, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.637657642364502, + "learning_rate": 4.041429731925264e-06, + "loss": 1.8267, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.9148659706115723, + "eval_runtime": 37.2248, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.441269636154175, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7106, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.582172155380249, + "learning_rate": 4.037367993501219e-06, + "loss": 1.7167, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.5083518028259277, + "learning_rate": 4.035337124289196e-06, + "loss": 1.8031, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.504112958908081, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.694, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.5068695545196533, + "learning_rate": 4.031275385865151e-06, + "loss": 1.6798, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.600541114807129, + "learning_rate": 4.029244516653128e-06, + "loss": 1.7346, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.503020763397217, + "learning_rate": 4.027213647441105e-06, + "loss": 1.7507, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.7774016857147217, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6737, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.5771589279174805, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6943, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.3841609954833984, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7142, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.9146945476531982, + "eval_runtime": 37.1701, + "eval_samples_per_second": 26.903, + "eval_steps_per_second": 1.13, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.466207981109619, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.7192, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.72037935256958, + "learning_rate": 4.017059301380992e-06, + "loss": 1.8082, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.36811900138855, + "learning_rate": 4.015028432168969e-06, + "loss": 1.6964, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.4803504943847656, + "learning_rate": 4.012997562956945e-06, + "loss": 1.779, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.707174777984619, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6715, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.462543249130249, + "learning_rate": 4.0089358245329e-06, + "loss": 1.7117, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.481444835662842, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7007, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.4432222843170166, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.703, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.611387014389038, + "learning_rate": 4.002843216896833e-06, + "loss": 1.7722, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.456305742263794, + "learning_rate": 4.000812347684809e-06, + "loss": 1.7615, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.9089828729629517, + "eval_runtime": 37.1812, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4287772178649902, + "learning_rate": 3.998781478472787e-06, + "loss": 1.7098, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5659544467926025, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7143, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.467067003250122, + "learning_rate": 3.994719740048742e-06, + "loss": 1.6668, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.5660688877105713, + "learning_rate": 3.992688870836718e-06, + "loss": 1.8185, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.6185719966888428, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7356, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.5389695167541504, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6973, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.499234914779663, + "learning_rate": 3.98659626320065e-06, + "loss": 1.7539, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.6246273517608643, + "learning_rate": 3.984565393988627e-06, + "loss": 1.7055, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6127917766571045, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7177, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.4941787719726562, + "learning_rate": 3.980503655564582e-06, + "loss": 1.728, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.9063301086425781, + "eval_runtime": 37.3284, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.4694740772247314, + "learning_rate": 3.978472786352559e-06, + "loss": 1.7384, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.8288261890411377, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7409, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.518359661102295, + "learning_rate": 3.974411047928514e-06, + "loss": 1.714, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4874095916748047, + "learning_rate": 3.972380178716491e-06, + "loss": 1.781, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.523772716522217, + "learning_rate": 3.970349309504468e-06, + "loss": 1.7474, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.406132459640503, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.7486, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.388207197189331, + "learning_rate": 3.966287571080423e-06, + "loss": 1.67, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.5564754009246826, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6803, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.531890869140625, + "learning_rate": 3.962225832656377e-06, + "loss": 1.7368, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.404284954071045, + "learning_rate": 3.960194963444355e-06, + "loss": 1.7247, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.9008547067642212, + "eval_runtime": 37.1962, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.3903231620788574, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6465, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7126, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7174, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.6466639041900635, + "learning_rate": 3.954102355808286e-06, + "loss": 1.7166, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.530284881591797, + "learning_rate": 3.952071486596264e-06, + "loss": 1.7155, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.601516008377075, + "learning_rate": 3.950040617384241e-06, + "loss": 1.7157, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.6146240234375, + "learning_rate": 3.948009748172218e-06, + "loss": 1.6931, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.4904634952545166, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.7594, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.345689296722412, + "learning_rate": 3.943948009748173e-06, + "loss": 1.7175, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4469082355499268, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7161, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.9021154642105103, + "eval_runtime": 36.8772, + "eval_samples_per_second": 27.117, + "eval_steps_per_second": 1.139, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.473752737045288, + "learning_rate": 3.939886271324127e-06, + "loss": 1.7688, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.461033582687378, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6752, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.419179916381836, + "learning_rate": 3.935824532900082e-06, + "loss": 1.6685, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.543102502822876, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7348, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.4932174682617188, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.7572, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.4563698768615723, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6691, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.566558837890625, + "learning_rate": 3.927701056051991e-06, + "loss": 1.697, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.5550761222839355, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.7194, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.51362681388855, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6804, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.448714017868042, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6575, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.904768943786621, + "eval_runtime": 37.0644, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.582794666290283, + "learning_rate": 3.9195775792039e-06, + "loss": 1.717, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.613327980041504, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.6927, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.522038221359253, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7391, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.372134208679199, + "learning_rate": 3.913484971567831e-06, + "loss": 1.7448, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.246591329574585, + "learning_rate": 3.911454102355808e-06, + "loss": 1.6531, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.48276948928833, + "learning_rate": 3.909423233143786e-06, + "loss": 1.7268, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.7920989990234375, + "learning_rate": 3.907392363931763e-06, + "loss": 1.7155, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.5430562496185303, + "learning_rate": 3.90536149471974e-06, + "loss": 1.7866, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.5819969177246094, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.8024, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.726024866104126, + "learning_rate": 3.901299756295695e-06, + "loss": 1.8467, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.899052381515503, + "eval_runtime": 37.192, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.6913323402404785, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6976, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.5307703018188477, + "learning_rate": 3.897238017871649e-06, + "loss": 1.7072, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.4976935386657715, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.7183, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.435443878173828, + "learning_rate": 3.893176279447604e-06, + "loss": 1.6845, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.704554319381714, + "learning_rate": 3.891145410235581e-06, + "loss": 1.78, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.4770138263702393, + "learning_rate": 3.889114541023558e-06, + "loss": 1.7436, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.472780227661133, + "learning_rate": 3.887083671811536e-06, + "loss": 1.7007, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.4693336486816406, + "learning_rate": 3.885052802599513e-06, + "loss": 1.6487, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.458834648132324, + "learning_rate": 3.88302193338749e-06, + "loss": 1.7895, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.4430108070373535, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.72, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.9009006023406982, + "eval_runtime": 37.1906, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.6104867458343506, + "learning_rate": 3.878960194963445e-06, + "loss": 1.6938, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.5191142559051514, + "learning_rate": 3.876929325751422e-06, + "loss": 1.7528, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.3788414001464844, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6446, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.485631227493286, + "learning_rate": 3.872867587327377e-06, + "loss": 1.7087, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.3721091747283936, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6752, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4648489952087402, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6883, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.447054862976074, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.7109, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.70007061958313, + "learning_rate": 3.864744110479286e-06, + "loss": 1.7342, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.6336557865142822, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7308, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.807671070098877, + "learning_rate": 3.86068237205524e-06, + "loss": 1.7519, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8990603685379028, + "eval_runtime": 37.1643, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.7177038192749023, + "learning_rate": 3.858651502843217e-06, + "loss": 1.7065, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.570364236831665, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6529, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.590329170227051, + "learning_rate": 3.854589764419171e-06, + "loss": 1.6955, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.542722702026367, + "learning_rate": 3.852558895207149e-06, + "loss": 1.7803, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.6095776557922363, + "learning_rate": 3.850528025995126e-06, + "loss": 1.7267, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.550161838531494, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6767, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.5079615116119385, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6897, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.5113067626953125, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.7024, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.559811592102051, + "learning_rate": 3.842404549147035e-06, + "loss": 1.695, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.507901191711426, + "learning_rate": 3.840373679935013e-06, + "loss": 1.661, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8952044248580933, + "eval_runtime": 37.2303, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.288231134414673, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.7095, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.260040044784546, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6858, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.5524179935455322, + "learning_rate": 3.834281072298944e-06, + "loss": 1.7362, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.419013738632202, + "learning_rate": 3.832250203086921e-06, + "loss": 1.6803, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.5345773696899414, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7435, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.5771822929382324, + "learning_rate": 3.828188464662876e-06, + "loss": 1.7932, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.4880053997039795, + "learning_rate": 3.826157595450853e-06, + "loss": 1.711, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.596400022506714, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6569, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.620776891708374, + "learning_rate": 3.822095857026808e-06, + "loss": 1.7159, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.512397527694702, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6832, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8950546979904175, + "eval_runtime": 37.1526, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5925326347351074, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6977, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.3980817794799805, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.6997, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.4825663566589355, + "learning_rate": 3.813972380178717e-06, + "loss": 1.7063, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.535006284713745, + "learning_rate": 3.811941510966694e-06, + "loss": 1.7457, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.450657606124878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.6521, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.3942458629608154, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6194, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5781140327453613, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.7105, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.5196564197540283, + "learning_rate": 3.803818034118603e-06, + "loss": 1.6741, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4819812774658203, + "learning_rate": 3.80178716490658e-06, + "loss": 1.7559, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.6393206119537354, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.7421, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.8945845365524292, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5286684036254883, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6933, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.690138339996338, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.6804, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.7789907455444336, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7839, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.5196211338043213, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7564, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.550898790359497, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.758, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.617361545562744, + "learning_rate": 3.787571080422421e-06, + "loss": 1.7497, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.533339500427246, + "learning_rate": 3.785540211210398e-06, + "loss": 1.7336, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.475037097930908, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.712, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.485506534576416, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.6985, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.712912082672119, + "learning_rate": 3.77944760357433e-06, + "loss": 1.812, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8921257257461548, + "eval_runtime": 37.1309, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.6280291080474854, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.7504, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6162853240966797, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6793, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.665316581726074, + "learning_rate": 3.773354995938262e-06, + "loss": 1.7333, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.531297445297241, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7298, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.543412446975708, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.7677, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.3839635848999023, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.7039, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.3417305946350098, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6206, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.4434802532196045, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.713, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.5223472118377686, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6587, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.577479124069214, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.7032, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8891643285751343, + "eval_runtime": 37.1123, + "eval_samples_per_second": 26.945, + "eval_steps_per_second": 1.132, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.468871593475342, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.732, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.5907723903656006, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.6866, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.4454426765441895, + "learning_rate": 3.753046303818034e-06, + "loss": 1.746, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.4651715755462646, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7375, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6342053413391113, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.7693, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.454977512359619, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6796, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.440626859664917, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7075, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.4905781745910645, + "learning_rate": 3.742891957757921e-06, + "loss": 1.7422, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6843647956848145, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7107, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.5906362533569336, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.7545, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8889163732528687, + "eval_runtime": 36.8136, + "eval_samples_per_second": 27.164, + "eval_steps_per_second": 1.141, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.5185422897338867, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6857, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.4425783157348633, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.687, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.5200576782226562, + "learning_rate": 3.732737611697807e-06, + "loss": 1.643, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.4664227962493896, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.5965, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.495962381362915, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6663, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.5607051849365234, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.6388, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 2.9295718669891357, + "learning_rate": 3.724614134849716e-06, + "loss": 1.7541, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.5632426738739014, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6359, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.6059279441833496, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.7427, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.6244208812713623, + "learning_rate": 3.718521527213648e-06, + "loss": 1.7097, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8882938623428345, + "eval_runtime": 36.9853, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.434684991836548, + "learning_rate": 3.716490658001625e-06, + "loss": 1.6875, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.6481709480285645, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5726, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.5398213863372803, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.6881, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.684122085571289, + "learning_rate": 3.710398050365557e-06, + "loss": 1.691, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.712214946746826, + "learning_rate": 3.708367181153534e-06, + "loss": 1.653, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.8427517414093018, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.6853, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.676410436630249, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.645, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.5397610664367676, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6171, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.606250286102295, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6729, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.3617212772369385, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.6546, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8915373086929321, + "eval_runtime": 37.094, + "eval_samples_per_second": 26.959, + "eval_steps_per_second": 1.132, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.6635043621063232, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.628, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.724837303161621, + "learning_rate": 3.694151096669375e-06, + "loss": 1.7537, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.56447434425354, + "learning_rate": 3.692120227457352e-06, + "loss": 1.6638, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6624584197998047, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.6806, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.7147905826568604, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5887, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.683840751647949, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.6664, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.8141531944274902, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6744, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.6172802448272705, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6866, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.7853286266326904, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.7034, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.7121260166168213, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6573, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.883426308631897, + "eval_runtime": 37.1964, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.4775121212005615, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6651, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.5361502170562744, + "learning_rate": 3.673842404549147e-06, + "loss": 1.6524, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.7969970703125, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6598, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.88427996635437, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.6821, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.9321346282958984, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.7777, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.802886724472046, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6808, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.85810923576355, + "learning_rate": 3.663688058489034e-06, + "loss": 1.7375, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.6501998901367188, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.6662, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.550652265548706, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6084, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.5003185272216797, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.681, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8800761699676514, + "eval_runtime": 37.2014, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.4262983798980713, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6608, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6474270820617676, + "learning_rate": 3.65353371242892e-06, + "loss": 1.6972, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.5103557109832764, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5581, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.529756546020508, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.693, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.578362464904785, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6302, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8062336444854736, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6855, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.935694932937622, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6861, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5260093212127686, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6534, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.7627487182617188, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.6833, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.9553534984588623, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6887, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.8832038640975952, + "eval_runtime": 37.1828, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.5526084899902344, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.7162, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.761218786239624, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6823, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.672189950942993, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6489, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.7650046348571777, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.7201, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.6351473331451416, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.6164, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5959131717681885, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.603, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.525513172149658, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6671, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.5529232025146484, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6446, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6846, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6786, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.885197639465332, + "eval_runtime": 37.2374, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.50577449798584, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6892, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.57772159576416, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5863, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.5455222129821777, + "learning_rate": 3.612916328188465e-06, + "loss": 1.692, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.6154329776763916, + "learning_rate": 3.610885458976442e-06, + "loss": 1.7027, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.5776023864746094, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.6353, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.5480730533599854, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.6358, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.5939579010009766, + "learning_rate": 3.604792851340374e-06, + "loss": 1.7099, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.65395188331604, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6763, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.5302748680114746, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.613, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.669433832168579, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.6584, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8816337585449219, + "eval_runtime": 37.2174, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.859063148498535, + "learning_rate": 3.596669374492283e-06, + "loss": 1.7096, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.531301259994507, + "learning_rate": 3.59463850528026e-06, + "loss": 1.6168, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.5920488834381104, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6953, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.698284387588501, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6602, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.762387752532959, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.7073, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.554579019546509, + "learning_rate": 3.586515028432169e-06, + "loss": 1.7151, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.644576072692871, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.6088, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.6375787258148193, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6434, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9475791454315186, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6936, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.5753467082977295, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5915, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8799077272415161, + "eval_runtime": 37.2335, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8163156509399414, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6824, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.6741368770599365, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.7066, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.5235214233398438, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6779, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.620927095413208, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6392, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7678003311157227, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6768, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.5900607109069824, + "learning_rate": 3.566206336311942e-06, + "loss": 1.618, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.4440128803253174, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5981, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.686119318008423, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6649, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.6912872791290283, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.6426, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.7249274253845215, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.6639, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8750032186508179, + "eval_runtime": 37.2319, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.750321626663208, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.7014, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.593047618865967, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6557, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.744931936264038, + "learning_rate": 3.551990251827783e-06, + "loss": 1.6257, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.4940085411071777, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6546, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.6905856132507324, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.6395, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.6744513511657715, + "learning_rate": 3.545897644191714e-06, + "loss": 1.6159, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6766908168792725, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6754, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.743103504180908, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.6273, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.886472463607788, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6836, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.823476791381836, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6475, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8762013912200928, + "eval_runtime": 36.8594, + "eval_samples_per_second": 27.13, + "eval_steps_per_second": 1.139, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.741985559463501, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6765, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.6321418285369873, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6997, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7359395027160645, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6529, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.610210657119751, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.68, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.7770233154296875, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6836, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.7146646976470947, + "learning_rate": 3.525588952071487e-06, + "loss": 1.699, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.782172679901123, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6941, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.577692747116089, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.6303, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.721348285675049, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.6381, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.825709342956543, + "learning_rate": 3.517465475223396e-06, + "loss": 1.6144, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8759764432907104, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.8446385860443115, + "learning_rate": 3.515434606011373e-06, + "loss": 1.685, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.70577335357666, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.7276, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.6624722480773926, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6735, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.7956180572509766, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6413, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.792559862136841, + "learning_rate": 3.507311129163282e-06, + "loss": 1.6035, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.7113680839538574, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6461, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.656461715698242, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.662, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.8650670051574707, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5952, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.8043570518493652, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6738, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.607752799987793, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6828, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8729931116104126, + "eval_runtime": 37.2275, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.5819315910339355, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.6108, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6494336128234863, + "learning_rate": 3.493095044679123e-06, + "loss": 1.7348, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.552762508392334, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5534, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.6580469608306885, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6675, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.922295331954956, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5819, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.7051825523376465, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6573, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.738743782043457, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6939, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.51552677154541, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.665, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.578549385070801, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6642, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.7047011852264404, + "learning_rate": 3.476848090982941e-06, + "loss": 1.6298, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8726528882980347, + "eval_runtime": 37.25, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.722877025604248, + "learning_rate": 3.474817221770918e-06, + "loss": 1.7522, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.464240074157715, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6876, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.6737945079803467, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.6453, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.7960164546966553, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.607, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.6985723972320557, + "learning_rate": 3.466693744922827e-06, + "loss": 1.6385, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.7561540603637695, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6753, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.828885555267334, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6402, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.653183698654175, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6889, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.5803189277648926, + "learning_rate": 3.458570268074736e-06, + "loss": 1.654, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.5868985652923584, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6366, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8656586408615112, + "eval_runtime": 37.2805, + "eval_samples_per_second": 26.824, + "eval_steps_per_second": 1.127, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.648036241531372, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.6019, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.5482382774353027, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6597, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.700294017791748, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.6075, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.9346179962158203, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6558, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.893852710723877, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6395, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.789057970046997, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6833, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.745492935180664, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.6266, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.7930359840393066, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5781, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.770575761795044, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6518, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 2.7367143630981445, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6982, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8708553314208984, + "eval_runtime": 37.2843, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.89182448387146, + "learning_rate": 3.434199837530463e-06, + "loss": 1.6587, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.5572729110717773, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5974, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8172996044158936, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6941, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.6480889320373535, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6902, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.654735803604126, + "learning_rate": 3.426076360682372e-06, + "loss": 1.7083, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.4182465076446533, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6494, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.6808230876922607, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5782, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.648282527923584, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6695, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.5920112133026123, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.7234, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.690538167953491, + "learning_rate": 3.415922014622259e-06, + "loss": 1.686, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8717143535614014, + "eval_runtime": 37.2714, + "eval_samples_per_second": 26.83, + "eval_steps_per_second": 1.127, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.672515392303467, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6883, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.8012726306915283, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6762, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.7551519870758057, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6694, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.687445878982544, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.6267, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.674445390701294, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6944, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.797832489013672, + "learning_rate": 3.403736799350122e-06, + "loss": 1.6735, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7843174934387207, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.6183, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.678809404373169, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.7336, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.748838424682617, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6499, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.5785276889801025, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6589, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8695456981658936, + "eval_runtime": 37.2493, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.7307844161987305, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5989, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.83357572555542, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6568, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.6178860664367676, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6654, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.699328899383545, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6657, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.7418158054351807, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6351, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.5589001178741455, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.6351, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.492795944213867, + "learning_rate": 3.381397238017872e-06, + "loss": 1.6188, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 2.8586132526397705, + "learning_rate": 3.379366368805849e-06, + "loss": 1.7127, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7773218154907227, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6529, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.6956539154052734, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.6566, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8692638874053955, + "eval_runtime": 37.2317, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.7406537532806396, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6482, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.746194362640381, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.6343, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.767970323562622, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6577, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.730640172958374, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5941, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.5661444664001465, + "learning_rate": 3.36515028432169e-06, + "loss": 1.6362, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.5915091037750244, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6526, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.7135071754455566, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6999, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.591201066970825, + "learning_rate": 3.359057676685622e-06, + "loss": 1.684, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.564439296722412, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5815, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.4480631351470947, + "learning_rate": 3.354995938261576e-06, + "loss": 1.6284, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8628013134002686, + "eval_runtime": 37.23, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.5392649173736572, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5841, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6359004974365234, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.6278, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.803671360015869, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6602, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.6031086444854736, + "learning_rate": 3.346872461413485e-06, + "loss": 1.6021, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.687382459640503, + "learning_rate": 3.344841592201463e-06, + "loss": 1.6207, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.6388001441955566, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.6194, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.603121042251587, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6532, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.8053836822509766, + "learning_rate": 3.338748984565394e-06, + "loss": 1.68, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.5472421646118164, + "learning_rate": 3.336718115353372e-06, + "loss": 1.7164, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.5142838954925537, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6441, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.8628681898117065, + "eval_runtime": 36.8298, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.6917803287506104, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.6235, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.6062285900115967, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6488, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.5503156185150146, + "learning_rate": 3.32859463850528e-06, + "loss": 1.639, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.433518648147583, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5951, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.600661516189575, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.6148, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.5558395385742188, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.6334, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.6363396644592285, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5858, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.819617748260498, + "learning_rate": 3.318440292445167e-06, + "loss": 1.6274, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 2.9121081829071045, + "learning_rate": 3.316409423233144e-06, + "loss": 1.6081, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.734858989715576, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6889, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8644160032272339, + "eval_runtime": 37.0389, + "eval_samples_per_second": 26.999, + "eval_steps_per_second": 1.134, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.6940014362335205, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6463, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.82663631439209, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6474, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.634434223175049, + "learning_rate": 3.308285946385053e-06, + "loss": 1.7172, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.6721079349517822, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5795, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.786130666732788, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6634, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7698960304260254, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.687, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.6305580139160156, + "learning_rate": 3.300162469536962e-06, + "loss": 1.6372, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.5221288204193115, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6746, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.583643674850464, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6538, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.5978968143463135, + "learning_rate": 3.294069861900894e-06, + "loss": 1.7177, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8665071725845337, + "eval_runtime": 37.1194, + "eval_samples_per_second": 26.94, + "eval_steps_per_second": 1.131, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.6446492671966553, + "learning_rate": 3.292038992688871e-06, + "loss": 1.657, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.646174192428589, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.682, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.671532392501831, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.7033, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.616450309753418, + "learning_rate": 3.285946385052803e-06, + "loss": 1.586, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.552800416946411, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6599, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.5700533390045166, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5593, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.675553560256958, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.7097, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.6610019207000732, + "learning_rate": 3.277822908204712e-06, + "loss": 1.6408, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.55435848236084, + "learning_rate": 3.275792038992689e-06, + "loss": 1.5909, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.811556816101074, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6552, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.8618569374084473, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.7060115337371826, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.6144, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.6817731857299805, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6481, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.8206212520599365, + "learning_rate": 3.267668562144598e-06, + "loss": 1.6294, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.834332227706909, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.6103, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.754049062728882, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6788, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.6089437007904053, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6469, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.6095848083496094, + "learning_rate": 3.259545085296507e-06, + "loss": 1.7122, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.606787919998169, + "learning_rate": 3.257514216084484e-06, + "loss": 1.5851, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.643245220184326, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.7042, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.8400895595550537, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6526, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8640165328979492, + "eval_runtime": 37.162, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.519524335861206, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.6268, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.565365791320801, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6562, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.6364638805389404, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.6143, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.6235382556915283, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.64, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.7358779907226562, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6509, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.6720035076141357, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5902, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5960681438446045, + "learning_rate": 3.23923639317628e-06, + "loss": 1.6378, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.6655092239379883, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.6102, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.7873294353485107, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6558, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.746189594268799, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.7116, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8605948686599731, + "eval_runtime": 37.1486, + "eval_samples_per_second": 26.919, + "eval_steps_per_second": 1.131, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5058770179748535, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.6156, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.8002450466156006, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6884, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.696186065673828, + "learning_rate": 3.227051177904143e-06, + "loss": 1.6403, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.5473484992980957, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.6375, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.6391875743865967, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6597, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.647244453430176, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5666, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849031448364258, + "learning_rate": 3.218927701056052e-06, + "loss": 1.6211, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.7565410137176514, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5686, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.938270092010498, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.7046, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.6462385654449463, + "learning_rate": 3.212835093419984e-06, + "loss": 1.602, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.8577098846435547, + "eval_runtime": 37.1448, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.679330587387085, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.636, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.6410343647003174, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6904, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6398510932922363, + "learning_rate": 3.206742485783916e-06, + "loss": 1.597, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.4636199474334717, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5823, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.646580457687378, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6591, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.8582923412323, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.6731, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.5474820137023926, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5656, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6609606742858887, + "learning_rate": 3.196588139723802e-06, + "loss": 1.6374, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.651947021484375, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6557, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.6957671642303467, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.6436, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.8583322763442993, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.5442283153533936, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6424, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.593792200088501, + "learning_rate": 3.188464662875711e-06, + "loss": 1.7012, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.5021846294403076, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.6268, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.7368624210357666, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5953, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.5100302696228027, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5866, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.5767033100128174, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6621, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.4694652557373047, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6554, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.606158971786499, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.6238, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.857424020767212, + "learning_rate": 3.174248578391552e-06, + "loss": 1.626, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.6120405197143555, + "learning_rate": 3.172217709179529e-06, + "loss": 1.6134, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8599685430526733, + "eval_runtime": 37.1373, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.61995530128479, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.6303, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6253163814544678, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5886, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.627302885055542, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6526, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.8007607460021973, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.6581, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.551227331161499, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5921, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.7288761138916016, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6478, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5698401927948, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6552, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.796701192855835, + "learning_rate": 3.155970755483347e-06, + "loss": 1.6391, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.6523337364196777, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6495, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.5528745651245117, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6396, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8551454544067383, + "eval_runtime": 37.1721, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.650125741958618, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6803, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.5522522926330566, + "learning_rate": 3.147847278635256e-06, + "loss": 1.726, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5302486419677734, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.6238, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.517793893814087, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5879, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.5749523639678955, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.6361, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6382429599761963, + "learning_rate": 3.139723801787165e-06, + "loss": 1.6143, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.7513139247894287, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6608, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.616007089614868, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5842, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.619389772415161, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6419, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.6845619678497314, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6726, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8519575595855713, + "eval_runtime": 36.823, + "eval_samples_per_second": 27.157, + "eval_steps_per_second": 1.141, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.9907562732696533, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6813, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5070035457611084, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6449, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.6286256313323975, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6614, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5735433101654053, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.6191, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.637507438659668, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6649, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.5027501583099365, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6779, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.8430395126342773, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.6428, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.6394317150115967, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.655, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.7121026515960693, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6798, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.6572396755218506, + "learning_rate": 3.111291632818847e-06, + "loss": 1.6937, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.857973575592041, + "eval_runtime": 37.0635, + "eval_samples_per_second": 26.981, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.660865306854248, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6454, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.6230008602142334, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6562, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.60079026222229, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.6311, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.6757969856262207, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6673, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.6204211711883545, + "learning_rate": 3.101137286758733e-06, + "loss": 1.665, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.7115979194641113, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6864, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.604799270629883, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.6364, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 2.7463159561157227, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.699, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.744046926498413, + "learning_rate": 3.093013809910642e-06, + "loss": 1.665, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8082664012908936, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.7027, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8549033403396606, + "eval_runtime": 37.1604, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.6146464347839355, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6635, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.724843978881836, + "learning_rate": 3.086921202274574e-06, + "loss": 1.64, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.699645757675171, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5988, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.769742727279663, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6478, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.5372848510742188, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.6285, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6461033821105957, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5669, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.8221795558929443, + "learning_rate": 3.07676685621446e-06, + "loss": 1.6455, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6035211086273193, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.6063, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.533945322036743, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.6234, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.627448797225952, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6381, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8547505140304565, + "eval_runtime": 37.226, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.6822423934936523, + "learning_rate": 3.068643379366369e-06, + "loss": 1.6436, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.647813558578491, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5901, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.619610548019409, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6559, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7284183502197266, + "learning_rate": 3.062550771730301e-06, + "loss": 1.6167, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.585660934448242, + "learning_rate": 3.060519902518278e-06, + "loss": 1.6176, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.689528465270996, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5651, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.484328508377075, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.6487, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.629913806915283, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.6099, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.711730480194092, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6702, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.590545415878296, + "learning_rate": 3.050365556458164e-06, + "loss": 1.6078, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8545923233032227, + "eval_runtime": 37.2534, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.5952389240264893, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.6084, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.739316940307617, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.6427, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.811095952987671, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6817, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.606715440750122, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6695, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.481234550476074, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5762, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.508068084716797, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.612, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.5694284439086914, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.6457, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.7029778957366943, + "learning_rate": 3.034118602761982e-06, + "loss": 1.6201, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.578904867172241, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5978, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.780586004257202, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6793, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.8522270917892456, + "eval_runtime": 37.2388, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.648653984069824, + "learning_rate": 3.028025995125914e-06, + "loss": 1.6626, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.631589651107788, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5822, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.5863327980041504, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.6322, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6556999683380127, + "learning_rate": 3.021933387489846e-06, + "loss": 1.5727, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.6391682624816895, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6334, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.6488680839538574, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6588, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.743858814239502, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5963, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.6499438285827637, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5997, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5450377464294434, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.6122, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.592026948928833, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.6362, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.84926176071167, + "eval_runtime": 37.2501, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.5699622631073, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.6055, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.779695987701416, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6503, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.6080546379089355, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.6121, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.70405650138855, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6774, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.69488525390625, + "learning_rate": 2.999593826157596e-06, + "loss": 1.6186, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.9266953468322754, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.6485, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.615251302719116, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.624, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.574878215789795, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6637, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6952786445617676, + "learning_rate": 2.991470349309505e-06, + "loss": 1.627, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.6076977252960205, + "learning_rate": 2.989439480097482e-06, + "loss": 1.6175, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.8475773334503174, + "eval_runtime": 37.2328, + "eval_samples_per_second": 26.858, + "eval_steps_per_second": 1.128, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.5707366466522217, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.6332, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.838609457015991, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6764, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.771082878112793, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6622, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6113975048065186, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6405, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.7537200450897217, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.6376, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.683002471923828, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6679, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.651897430419922, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.6091, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.6856229305267334, + "learning_rate": 2.9731925264013e-06, + "loss": 1.6223, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.646522045135498, + "learning_rate": 2.971161657189277e-06, + "loss": 1.6126, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.592247486114502, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.6088, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.8501410484313965, + "eval_runtime": 37.225, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.03291392326355, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.6051, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.903224468231201, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6753, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.5178730487823486, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5894, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.6278252601623535, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6687, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.8105742931365967, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6648, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.5897178649902344, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.6012, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.7172093391418457, + "learning_rate": 2.954914703493095e-06, + "loss": 1.6206, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.6845815181732178, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.6163, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.652451753616333, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.5916, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.740280866622925, + "learning_rate": 2.948822095857027e-06, + "loss": 1.627, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8489265441894531, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 1300 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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": 4.052502232170496e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1300/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/latest new file mode 100644 index 0000000000000000000000000000000000000000..c2a0b5b17b996514a9bea0debe9403a641844bce --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..eea9adc79dbfbd5d9382a38aed8e6aaaaf2fb618 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/trainer_state.json @@ -0,0 +1,10953 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.5342465753424657, + "eval_steps": 10, + "global_step": 1400, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6008102893829346, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7848, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.630127429962158, + "learning_rate": 4.362307067424858e-06, + "loss": 1.7135, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.720945358276367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7897, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.898836851119995, + "learning_rate": 4.358245329000813e-06, + "loss": 1.7434, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.6548235416412354, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7683, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.7218399047851562, + "learning_rate": 4.354183590576767e-06, + "loss": 1.7729, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.5972483158111572, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7517, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.6741726398468018, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7354, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.898418426513672, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7843, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.3964333534240723, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.808, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.9525212049484253, + "eval_runtime": 36.8297, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.9952592849731445, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7856, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.81740140914917, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7613, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6271541118621826, + "learning_rate": 4.339967506092607e-06, + "loss": 1.8409, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.540395498275757, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.7765, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.629131317138672, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7634, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.5796725749969482, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7414, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6980209350585938, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.8048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.658919095993042, + "learning_rate": 4.329813160032495e-06, + "loss": 1.7725, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5237927436828613, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7559, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.5822298526763916, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7129, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9492865800857544, + "eval_runtime": 37.0012, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.762810468673706, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7895, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.48427414894104, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7667, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.790052890777588, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7548, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.927957773208618, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.7489, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.4964418411254883, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7947, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.4845147132873535, + "learning_rate": 4.313566206336312e-06, + "loss": 1.7886, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.687361717224121, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7968, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.568202018737793, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7332, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.496429443359375, + "learning_rate": 4.307473598700244e-06, + "loss": 1.7583, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6063425540924072, + "learning_rate": 4.305442729488221e-06, + "loss": 1.779, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.9421359300613403, + "eval_runtime": 37.0712, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.5371274948120117, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7133, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.5601038932800293, + "learning_rate": 4.301380991064176e-06, + "loss": 1.8089, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.6244118213653564, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7983, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.4798192977905273, + "learning_rate": 4.29731925264013e-06, + "loss": 1.8221, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.483376979827881, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.8032, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.716745615005493, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7583, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.6685945987701416, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7365, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.465810537338257, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7682, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.7629129886627197, + "learning_rate": 4.287164906580017e-06, + "loss": 1.809, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5273563861846924, + "learning_rate": 4.285134037367994e-06, + "loss": 1.7464, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.9359965324401855, + "eval_runtime": 37.1317, + "eval_samples_per_second": 26.931, + "eval_steps_per_second": 1.131, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.514230966567993, + "learning_rate": 4.283103168155971e-06, + "loss": 1.7365, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.704042911529541, + "learning_rate": 4.281072298943948e-06, + "loss": 1.7268, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.4716262817382812, + "learning_rate": 4.279041429731926e-06, + "loss": 1.7212, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5695507526397705, + "learning_rate": 4.277010560519903e-06, + "loss": 1.7398, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.762845039367676, + "learning_rate": 4.27497969130788e-06, + "loss": 1.7158, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.571054220199585, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7685, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.908184289932251, + "learning_rate": 4.270917952883835e-06, + "loss": 1.7865, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.705479621887207, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7922, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.4418110847473145, + "learning_rate": 4.266856214459789e-06, + "loss": 1.7312, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.4619884490966797, + "learning_rate": 4.264825345247766e-06, + "loss": 1.8343, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.9352803230285645, + "eval_runtime": 37.0958, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5386202335357666, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6902, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9012045860290527, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7552, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.8028385639190674, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.6872, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.4505386352539062, + "learning_rate": 4.256701868399675e-06, + "loss": 1.75, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7551822662353516, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7612, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.596270799636841, + "learning_rate": 4.252640129975629e-06, + "loss": 1.7694, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.387662172317505, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.7776, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.7208738327026367, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7375, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.5317776203155518, + "learning_rate": 4.246547522339562e-06, + "loss": 1.6922, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.4803197383880615, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.741, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.9322715997695923, + "eval_runtime": 37.0962, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.5688276290893555, + "learning_rate": 4.242485783915517e-06, + "loss": 1.7518, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.4652810096740723, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6853, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.541903257369995, + "learning_rate": 4.23842404549147e-06, + "loss": 1.6841, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.377857208251953, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7312, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.418323040008545, + "learning_rate": 4.234362307067425e-06, + "loss": 1.6892, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5374410152435303, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7719, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7119228839874268, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.7819, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.5333240032196045, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7758, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.528696060180664, + "learning_rate": 4.226238830219334e-06, + "loss": 1.75, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.539039134979248, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7498, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.934200644493103, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.5384178161621094, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7615, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.601658821105957, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7645, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5359764099121094, + "learning_rate": 4.218115353371243e-06, + "loss": 1.7549, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.641118049621582, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7487, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.4146382808685303, + "learning_rate": 4.214053614947198e-06, + "loss": 1.7102, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4245245456695557, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7842, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.509824752807617, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6876, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.636444330215454, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7681, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.476858615875244, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7049, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.75732159614563, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7833, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.931411623954773, + "eval_runtime": 37.0994, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.5437941551208496, + "learning_rate": 4.201868399675061e-06, + "loss": 1.7751, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.5010673999786377, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7715, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.6352853775024414, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6798, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.588247776031494, + "learning_rate": 4.195775792038993e-06, + "loss": 1.7575, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.349921226501465, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7245, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.394432783126831, + "learning_rate": 4.191714053614948e-06, + "loss": 1.6778, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.883906602859497, + "learning_rate": 4.189683184402924e-06, + "loss": 1.7701, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.537799596786499, + "learning_rate": 4.187652315190902e-06, + "loss": 1.7286, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.575700283050537, + "learning_rate": 4.185621445978879e-06, + "loss": 1.7812, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.592252492904663, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7176, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.9259952306747437, + "eval_runtime": 37.144, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.511845111846924, + "learning_rate": 4.181559707554833e-06, + "loss": 1.7564, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.575446367263794, + "learning_rate": 4.179528838342811e-06, + "loss": 1.7943, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.379391670227051, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7098, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.701812982559204, + "learning_rate": 4.175467099918766e-06, + "loss": 1.7292, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.486640453338623, + "learning_rate": 4.173436230706742e-06, + "loss": 1.7963, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.561983823776245, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.7153, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.5709176063537598, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7412, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.3308839797973633, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7188, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.592470169067383, + "learning_rate": 4.165312753858652e-06, + "loss": 1.7599, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.3747010231018066, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7361, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.9229093790054321, + "eval_runtime": 37.1417, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.4062509536743164, + "learning_rate": 4.161251015434607e-06, + "loss": 1.7256, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.4951181411743164, + "learning_rate": 4.159220146222583e-06, + "loss": 1.7622, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.3453004360198975, + "learning_rate": 4.157189277010561e-06, + "loss": 1.6969, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.382671594619751, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7109, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4277336597442627, + "learning_rate": 4.153127538586516e-06, + "loss": 1.7471, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.4476678371429443, + "learning_rate": 4.151096669374492e-06, + "loss": 1.7483, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.451099395751953, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7366, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.618217706680298, + "learning_rate": 4.147034930950447e-06, + "loss": 1.7282, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.414400100708008, + "learning_rate": 4.145004061738425e-06, + "loss": 1.7409, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.535353422164917, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.7346, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.9240931272506714, + "eval_runtime": 36.844, + "eval_samples_per_second": 27.141, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.5075416564941406, + "learning_rate": 4.140942323314379e-06, + "loss": 1.6668, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.4136979579925537, + "learning_rate": 4.138911454102356e-06, + "loss": 1.696, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.4363577365875244, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7174, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.4578073024749756, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7621, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.370976686477661, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7443, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.363755702972412, + "learning_rate": 4.130787977254265e-06, + "loss": 1.7399, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.44704532623291, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7481, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.5463340282440186, + "learning_rate": 4.12672623883022e-06, + "loss": 1.7646, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7088072299957275, + "learning_rate": 4.124695369618197e-06, + "loss": 1.8108, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.463271141052246, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7311, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.918715238571167, + "eval_runtime": 37.0468, + "eval_samples_per_second": 26.993, + "eval_steps_per_second": 1.134, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6165266036987305, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.7295, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.3432133197784424, + "learning_rate": 4.118602761982129e-06, + "loss": 1.7512, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.552751302719116, + "learning_rate": 4.116571892770106e-06, + "loss": 1.6976, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7586560249328613, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6708, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.5326449871063232, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7824, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.7658956050872803, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7093, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.6868369579315186, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7475, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.543748140335083, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.7714, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.7235589027404785, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7593, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.697038173675537, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6296, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.91779363155365, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 2.6746180057525635, + "learning_rate": 4.100324939073924e-06, + "loss": 1.7971, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.6281094551086426, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7212, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.446025848388672, + "learning_rate": 4.096263200649879e-06, + "loss": 1.7324, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.5460824966430664, + "learning_rate": 4.094232331437856e-06, + "loss": 1.7435, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.5419771671295166, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6979, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.5545520782470703, + "learning_rate": 4.090170593013811e-06, + "loss": 1.7042, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.8353254795074463, + "learning_rate": 4.088139723801787e-06, + "loss": 1.7313, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.6521668434143066, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6901, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.6611177921295166, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7064, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.411482572555542, + "learning_rate": 4.08204711616572e-06, + "loss": 1.729, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.9176616668701172, + "eval_runtime": 37.1603, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.689110279083252, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7707, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.544018268585205, + "learning_rate": 4.077985377741674e-06, + "loss": 1.6724, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.55307936668396, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7677, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.3896875381469727, + "learning_rate": 4.073923639317629e-06, + "loss": 1.7729, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.381887197494507, + "learning_rate": 4.071892770105605e-06, + "loss": 1.7737, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.4873931407928467, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6928, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.4908432960510254, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6674, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.4346399307250977, + "learning_rate": 4.065800162469538e-06, + "loss": 1.6947, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6459081172943115, + "learning_rate": 4.063769293257514e-06, + "loss": 1.701, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.6092610359191895, + "learning_rate": 4.061738424045492e-06, + "loss": 1.7425, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.919799566268921, + "eval_runtime": 37.1632, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.4443421363830566, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7267, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.8252694606781006, + "learning_rate": 4.057676685621446e-06, + "loss": 1.7412, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.3391194343566895, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.7916, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.57979679107666, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7438, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.398383855819702, + "learning_rate": 4.051584077985378e-06, + "loss": 1.6606, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.394498348236084, + "learning_rate": 4.049553208773355e-06, + "loss": 1.736, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5070464611053467, + "learning_rate": 4.047522339561333e-06, + "loss": 1.7559, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.439722776412964, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7133, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.569152593612671, + "learning_rate": 4.043460601137287e-06, + "loss": 1.7123, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.637657642364502, + "learning_rate": 4.041429731925264e-06, + "loss": 1.8267, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.9148659706115723, + "eval_runtime": 37.2248, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.441269636154175, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7106, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.582172155380249, + "learning_rate": 4.037367993501219e-06, + "loss": 1.7167, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.5083518028259277, + "learning_rate": 4.035337124289196e-06, + "loss": 1.8031, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.504112958908081, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.694, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.5068695545196533, + "learning_rate": 4.031275385865151e-06, + "loss": 1.6798, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.600541114807129, + "learning_rate": 4.029244516653128e-06, + "loss": 1.7346, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.503020763397217, + "learning_rate": 4.027213647441105e-06, + "loss": 1.7507, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.7774016857147217, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6737, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.5771589279174805, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6943, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.3841609954833984, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7142, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.9146945476531982, + "eval_runtime": 37.1701, + "eval_samples_per_second": 26.903, + "eval_steps_per_second": 1.13, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.466207981109619, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.7192, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.72037935256958, + "learning_rate": 4.017059301380992e-06, + "loss": 1.8082, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.36811900138855, + "learning_rate": 4.015028432168969e-06, + "loss": 1.6964, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.4803504943847656, + "learning_rate": 4.012997562956945e-06, + "loss": 1.779, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.707174777984619, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6715, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.462543249130249, + "learning_rate": 4.0089358245329e-06, + "loss": 1.7117, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.481444835662842, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7007, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.4432222843170166, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.703, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.611387014389038, + "learning_rate": 4.002843216896833e-06, + "loss": 1.7722, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.456305742263794, + "learning_rate": 4.000812347684809e-06, + "loss": 1.7615, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.9089828729629517, + "eval_runtime": 37.1812, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4287772178649902, + "learning_rate": 3.998781478472787e-06, + "loss": 1.7098, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5659544467926025, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7143, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.467067003250122, + "learning_rate": 3.994719740048742e-06, + "loss": 1.6668, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.5660688877105713, + "learning_rate": 3.992688870836718e-06, + "loss": 1.8185, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.6185719966888428, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7356, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.5389695167541504, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6973, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.499234914779663, + "learning_rate": 3.98659626320065e-06, + "loss": 1.7539, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.6246273517608643, + "learning_rate": 3.984565393988627e-06, + "loss": 1.7055, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6127917766571045, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7177, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.4941787719726562, + "learning_rate": 3.980503655564582e-06, + "loss": 1.728, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.9063301086425781, + "eval_runtime": 37.3284, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.4694740772247314, + "learning_rate": 3.978472786352559e-06, + "loss": 1.7384, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.8288261890411377, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7409, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.518359661102295, + "learning_rate": 3.974411047928514e-06, + "loss": 1.714, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4874095916748047, + "learning_rate": 3.972380178716491e-06, + "loss": 1.781, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.523772716522217, + "learning_rate": 3.970349309504468e-06, + "loss": 1.7474, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.406132459640503, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.7486, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.388207197189331, + "learning_rate": 3.966287571080423e-06, + "loss": 1.67, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.5564754009246826, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6803, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.531890869140625, + "learning_rate": 3.962225832656377e-06, + "loss": 1.7368, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.404284954071045, + "learning_rate": 3.960194963444355e-06, + "loss": 1.7247, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.9008547067642212, + "eval_runtime": 37.1962, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.3903231620788574, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6465, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7126, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7174, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.6466639041900635, + "learning_rate": 3.954102355808286e-06, + "loss": 1.7166, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.530284881591797, + "learning_rate": 3.952071486596264e-06, + "loss": 1.7155, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.601516008377075, + "learning_rate": 3.950040617384241e-06, + "loss": 1.7157, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.6146240234375, + "learning_rate": 3.948009748172218e-06, + "loss": 1.6931, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.4904634952545166, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.7594, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.345689296722412, + "learning_rate": 3.943948009748173e-06, + "loss": 1.7175, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4469082355499268, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7161, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.9021154642105103, + "eval_runtime": 36.8772, + "eval_samples_per_second": 27.117, + "eval_steps_per_second": 1.139, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.473752737045288, + "learning_rate": 3.939886271324127e-06, + "loss": 1.7688, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.461033582687378, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6752, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.419179916381836, + "learning_rate": 3.935824532900082e-06, + "loss": 1.6685, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.543102502822876, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7348, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.4932174682617188, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.7572, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.4563698768615723, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6691, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.566558837890625, + "learning_rate": 3.927701056051991e-06, + "loss": 1.697, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.5550761222839355, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.7194, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.51362681388855, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6804, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.448714017868042, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6575, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.904768943786621, + "eval_runtime": 37.0644, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.582794666290283, + "learning_rate": 3.9195775792039e-06, + "loss": 1.717, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.613327980041504, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.6927, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.522038221359253, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7391, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.372134208679199, + "learning_rate": 3.913484971567831e-06, + "loss": 1.7448, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.246591329574585, + "learning_rate": 3.911454102355808e-06, + "loss": 1.6531, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.48276948928833, + "learning_rate": 3.909423233143786e-06, + "loss": 1.7268, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.7920989990234375, + "learning_rate": 3.907392363931763e-06, + "loss": 1.7155, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.5430562496185303, + "learning_rate": 3.90536149471974e-06, + "loss": 1.7866, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.5819969177246094, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.8024, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.726024866104126, + "learning_rate": 3.901299756295695e-06, + "loss": 1.8467, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.899052381515503, + "eval_runtime": 37.192, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.6913323402404785, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6976, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.5307703018188477, + "learning_rate": 3.897238017871649e-06, + "loss": 1.7072, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.4976935386657715, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.7183, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.435443878173828, + "learning_rate": 3.893176279447604e-06, + "loss": 1.6845, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.704554319381714, + "learning_rate": 3.891145410235581e-06, + "loss": 1.78, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.4770138263702393, + "learning_rate": 3.889114541023558e-06, + "loss": 1.7436, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.472780227661133, + "learning_rate": 3.887083671811536e-06, + "loss": 1.7007, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.4693336486816406, + "learning_rate": 3.885052802599513e-06, + "loss": 1.6487, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.458834648132324, + "learning_rate": 3.88302193338749e-06, + "loss": 1.7895, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.4430108070373535, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.72, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.9009006023406982, + "eval_runtime": 37.1906, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.6104867458343506, + "learning_rate": 3.878960194963445e-06, + "loss": 1.6938, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.5191142559051514, + "learning_rate": 3.876929325751422e-06, + "loss": 1.7528, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.3788414001464844, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6446, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.485631227493286, + "learning_rate": 3.872867587327377e-06, + "loss": 1.7087, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.3721091747283936, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6752, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4648489952087402, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6883, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.447054862976074, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.7109, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.70007061958313, + "learning_rate": 3.864744110479286e-06, + "loss": 1.7342, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.6336557865142822, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7308, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.807671070098877, + "learning_rate": 3.86068237205524e-06, + "loss": 1.7519, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8990603685379028, + "eval_runtime": 37.1643, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.7177038192749023, + "learning_rate": 3.858651502843217e-06, + "loss": 1.7065, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.570364236831665, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6529, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.590329170227051, + "learning_rate": 3.854589764419171e-06, + "loss": 1.6955, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.542722702026367, + "learning_rate": 3.852558895207149e-06, + "loss": 1.7803, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.6095776557922363, + "learning_rate": 3.850528025995126e-06, + "loss": 1.7267, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.550161838531494, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6767, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.5079615116119385, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6897, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.5113067626953125, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.7024, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.559811592102051, + "learning_rate": 3.842404549147035e-06, + "loss": 1.695, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.507901191711426, + "learning_rate": 3.840373679935013e-06, + "loss": 1.661, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8952044248580933, + "eval_runtime": 37.2303, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.288231134414673, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.7095, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.260040044784546, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6858, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.5524179935455322, + "learning_rate": 3.834281072298944e-06, + "loss": 1.7362, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.419013738632202, + "learning_rate": 3.832250203086921e-06, + "loss": 1.6803, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.5345773696899414, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7435, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.5771822929382324, + "learning_rate": 3.828188464662876e-06, + "loss": 1.7932, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.4880053997039795, + "learning_rate": 3.826157595450853e-06, + "loss": 1.711, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.596400022506714, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6569, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.620776891708374, + "learning_rate": 3.822095857026808e-06, + "loss": 1.7159, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.512397527694702, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6832, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8950546979904175, + "eval_runtime": 37.1526, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5925326347351074, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6977, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.3980817794799805, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.6997, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.4825663566589355, + "learning_rate": 3.813972380178717e-06, + "loss": 1.7063, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.535006284713745, + "learning_rate": 3.811941510966694e-06, + "loss": 1.7457, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.450657606124878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.6521, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.3942458629608154, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6194, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5781140327453613, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.7105, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.5196564197540283, + "learning_rate": 3.803818034118603e-06, + "loss": 1.6741, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4819812774658203, + "learning_rate": 3.80178716490658e-06, + "loss": 1.7559, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.6393206119537354, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.7421, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.8945845365524292, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5286684036254883, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6933, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.690138339996338, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.6804, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.7789907455444336, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7839, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.5196211338043213, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7564, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.550898790359497, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.758, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.617361545562744, + "learning_rate": 3.787571080422421e-06, + "loss": 1.7497, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.533339500427246, + "learning_rate": 3.785540211210398e-06, + "loss": 1.7336, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.475037097930908, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.712, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.485506534576416, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.6985, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.712912082672119, + "learning_rate": 3.77944760357433e-06, + "loss": 1.812, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8921257257461548, + "eval_runtime": 37.1309, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.6280291080474854, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.7504, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6162853240966797, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6793, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.665316581726074, + "learning_rate": 3.773354995938262e-06, + "loss": 1.7333, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.531297445297241, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7298, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.543412446975708, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.7677, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.3839635848999023, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.7039, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.3417305946350098, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6206, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.4434802532196045, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.713, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.5223472118377686, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6587, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.577479124069214, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.7032, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8891643285751343, + "eval_runtime": 37.1123, + "eval_samples_per_second": 26.945, + "eval_steps_per_second": 1.132, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.468871593475342, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.732, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.5907723903656006, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.6866, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.4454426765441895, + "learning_rate": 3.753046303818034e-06, + "loss": 1.746, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.4651715755462646, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7375, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6342053413391113, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.7693, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.454977512359619, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6796, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.440626859664917, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7075, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.4905781745910645, + "learning_rate": 3.742891957757921e-06, + "loss": 1.7422, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6843647956848145, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7107, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.5906362533569336, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.7545, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8889163732528687, + "eval_runtime": 36.8136, + "eval_samples_per_second": 27.164, + "eval_steps_per_second": 1.141, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.5185422897338867, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6857, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.4425783157348633, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.687, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.5200576782226562, + "learning_rate": 3.732737611697807e-06, + "loss": 1.643, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.4664227962493896, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.5965, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.495962381362915, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6663, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.5607051849365234, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.6388, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 2.9295718669891357, + "learning_rate": 3.724614134849716e-06, + "loss": 1.7541, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.5632426738739014, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6359, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.6059279441833496, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.7427, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.6244208812713623, + "learning_rate": 3.718521527213648e-06, + "loss": 1.7097, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8882938623428345, + "eval_runtime": 36.9853, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.434684991836548, + "learning_rate": 3.716490658001625e-06, + "loss": 1.6875, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.6481709480285645, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5726, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.5398213863372803, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.6881, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.684122085571289, + "learning_rate": 3.710398050365557e-06, + "loss": 1.691, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.712214946746826, + "learning_rate": 3.708367181153534e-06, + "loss": 1.653, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.8427517414093018, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.6853, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.676410436630249, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.645, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.5397610664367676, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6171, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.606250286102295, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6729, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.3617212772369385, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.6546, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8915373086929321, + "eval_runtime": 37.094, + "eval_samples_per_second": 26.959, + "eval_steps_per_second": 1.132, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.6635043621063232, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.628, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.724837303161621, + "learning_rate": 3.694151096669375e-06, + "loss": 1.7537, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.56447434425354, + "learning_rate": 3.692120227457352e-06, + "loss": 1.6638, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6624584197998047, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.6806, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.7147905826568604, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5887, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.683840751647949, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.6664, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.8141531944274902, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6744, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.6172802448272705, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6866, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.7853286266326904, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.7034, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.7121260166168213, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6573, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.883426308631897, + "eval_runtime": 37.1964, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.4775121212005615, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6651, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.5361502170562744, + "learning_rate": 3.673842404549147e-06, + "loss": 1.6524, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.7969970703125, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6598, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.88427996635437, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.6821, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.9321346282958984, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.7777, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.802886724472046, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6808, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.85810923576355, + "learning_rate": 3.663688058489034e-06, + "loss": 1.7375, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.6501998901367188, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.6662, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.550652265548706, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6084, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.5003185272216797, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.681, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8800761699676514, + "eval_runtime": 37.2014, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.4262983798980713, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6608, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6474270820617676, + "learning_rate": 3.65353371242892e-06, + "loss": 1.6972, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.5103557109832764, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5581, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.529756546020508, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.693, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.578362464904785, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6302, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8062336444854736, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6855, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.935694932937622, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6861, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5260093212127686, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6534, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.7627487182617188, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.6833, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.9553534984588623, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6887, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.8832038640975952, + "eval_runtime": 37.1828, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.5526084899902344, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.7162, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.761218786239624, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6823, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.672189950942993, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6489, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.7650046348571777, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.7201, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.6351473331451416, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.6164, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5959131717681885, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.603, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.525513172149658, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6671, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.5529232025146484, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6446, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6846, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6786, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.885197639465332, + "eval_runtime": 37.2374, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.50577449798584, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6892, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.57772159576416, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5863, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.5455222129821777, + "learning_rate": 3.612916328188465e-06, + "loss": 1.692, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.6154329776763916, + "learning_rate": 3.610885458976442e-06, + "loss": 1.7027, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.5776023864746094, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.6353, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.5480730533599854, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.6358, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.5939579010009766, + "learning_rate": 3.604792851340374e-06, + "loss": 1.7099, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.65395188331604, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6763, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.5302748680114746, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.613, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.669433832168579, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.6584, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8816337585449219, + "eval_runtime": 37.2174, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.859063148498535, + "learning_rate": 3.596669374492283e-06, + "loss": 1.7096, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.531301259994507, + "learning_rate": 3.59463850528026e-06, + "loss": 1.6168, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.5920488834381104, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6953, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.698284387588501, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6602, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.762387752532959, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.7073, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.554579019546509, + "learning_rate": 3.586515028432169e-06, + "loss": 1.7151, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.644576072692871, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.6088, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.6375787258148193, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6434, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9475791454315186, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6936, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.5753467082977295, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5915, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8799077272415161, + "eval_runtime": 37.2335, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8163156509399414, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6824, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.6741368770599365, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.7066, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.5235214233398438, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6779, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.620927095413208, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6392, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7678003311157227, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6768, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.5900607109069824, + "learning_rate": 3.566206336311942e-06, + "loss": 1.618, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.4440128803253174, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5981, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.686119318008423, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6649, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.6912872791290283, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.6426, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.7249274253845215, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.6639, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8750032186508179, + "eval_runtime": 37.2319, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.750321626663208, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.7014, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.593047618865967, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6557, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.744931936264038, + "learning_rate": 3.551990251827783e-06, + "loss": 1.6257, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.4940085411071777, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6546, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.6905856132507324, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.6395, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.6744513511657715, + "learning_rate": 3.545897644191714e-06, + "loss": 1.6159, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6766908168792725, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6754, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.743103504180908, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.6273, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.886472463607788, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6836, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.823476791381836, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6475, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8762013912200928, + "eval_runtime": 36.8594, + "eval_samples_per_second": 27.13, + "eval_steps_per_second": 1.139, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.741985559463501, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6765, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.6321418285369873, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6997, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7359395027160645, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6529, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.610210657119751, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.68, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.7770233154296875, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6836, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.7146646976470947, + "learning_rate": 3.525588952071487e-06, + "loss": 1.699, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.782172679901123, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6941, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.577692747116089, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.6303, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.721348285675049, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.6381, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.825709342956543, + "learning_rate": 3.517465475223396e-06, + "loss": 1.6144, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8759764432907104, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.8446385860443115, + "learning_rate": 3.515434606011373e-06, + "loss": 1.685, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.70577335357666, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.7276, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.6624722480773926, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6735, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.7956180572509766, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6413, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.792559862136841, + "learning_rate": 3.507311129163282e-06, + "loss": 1.6035, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.7113680839538574, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6461, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.656461715698242, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.662, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.8650670051574707, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5952, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.8043570518493652, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6738, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.607752799987793, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6828, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8729931116104126, + "eval_runtime": 37.2275, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.5819315910339355, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.6108, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6494336128234863, + "learning_rate": 3.493095044679123e-06, + "loss": 1.7348, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.552762508392334, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5534, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.6580469608306885, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6675, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.922295331954956, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5819, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.7051825523376465, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6573, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.738743782043457, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6939, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.51552677154541, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.665, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.578549385070801, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6642, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.7047011852264404, + "learning_rate": 3.476848090982941e-06, + "loss": 1.6298, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8726528882980347, + "eval_runtime": 37.25, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.722877025604248, + "learning_rate": 3.474817221770918e-06, + "loss": 1.7522, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.464240074157715, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6876, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.6737945079803467, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.6453, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.7960164546966553, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.607, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.6985723972320557, + "learning_rate": 3.466693744922827e-06, + "loss": 1.6385, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.7561540603637695, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6753, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.828885555267334, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6402, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.653183698654175, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6889, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.5803189277648926, + "learning_rate": 3.458570268074736e-06, + "loss": 1.654, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.5868985652923584, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6366, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8656586408615112, + "eval_runtime": 37.2805, + "eval_samples_per_second": 26.824, + "eval_steps_per_second": 1.127, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.648036241531372, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.6019, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.5482382774353027, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6597, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.700294017791748, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.6075, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.9346179962158203, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6558, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.893852710723877, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6395, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.789057970046997, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6833, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.745492935180664, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.6266, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.7930359840393066, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5781, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.770575761795044, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6518, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 2.7367143630981445, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6982, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8708553314208984, + "eval_runtime": 37.2843, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.89182448387146, + "learning_rate": 3.434199837530463e-06, + "loss": 1.6587, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.5572729110717773, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5974, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8172996044158936, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6941, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.6480889320373535, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6902, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.654735803604126, + "learning_rate": 3.426076360682372e-06, + "loss": 1.7083, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.4182465076446533, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6494, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.6808230876922607, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5782, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.648282527923584, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6695, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.5920112133026123, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.7234, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.690538167953491, + "learning_rate": 3.415922014622259e-06, + "loss": 1.686, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8717143535614014, + "eval_runtime": 37.2714, + "eval_samples_per_second": 26.83, + "eval_steps_per_second": 1.127, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.672515392303467, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6883, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.8012726306915283, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6762, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.7551519870758057, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6694, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.687445878982544, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.6267, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.674445390701294, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6944, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.797832489013672, + "learning_rate": 3.403736799350122e-06, + "loss": 1.6735, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7843174934387207, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.6183, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.678809404373169, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.7336, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.748838424682617, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6499, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.5785276889801025, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6589, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8695456981658936, + "eval_runtime": 37.2493, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.7307844161987305, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5989, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.83357572555542, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6568, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.6178860664367676, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6654, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.699328899383545, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6657, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.7418158054351807, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6351, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.5589001178741455, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.6351, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.492795944213867, + "learning_rate": 3.381397238017872e-06, + "loss": 1.6188, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 2.8586132526397705, + "learning_rate": 3.379366368805849e-06, + "loss": 1.7127, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7773218154907227, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6529, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.6956539154052734, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.6566, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8692638874053955, + "eval_runtime": 37.2317, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.7406537532806396, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6482, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.746194362640381, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.6343, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.767970323562622, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6577, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.730640172958374, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5941, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.5661444664001465, + "learning_rate": 3.36515028432169e-06, + "loss": 1.6362, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.5915091037750244, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6526, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.7135071754455566, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6999, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.591201066970825, + "learning_rate": 3.359057676685622e-06, + "loss": 1.684, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.564439296722412, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5815, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.4480631351470947, + "learning_rate": 3.354995938261576e-06, + "loss": 1.6284, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8628013134002686, + "eval_runtime": 37.23, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.5392649173736572, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5841, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6359004974365234, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.6278, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.803671360015869, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6602, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.6031086444854736, + "learning_rate": 3.346872461413485e-06, + "loss": 1.6021, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.687382459640503, + "learning_rate": 3.344841592201463e-06, + "loss": 1.6207, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.6388001441955566, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.6194, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.603121042251587, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6532, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.8053836822509766, + "learning_rate": 3.338748984565394e-06, + "loss": 1.68, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.5472421646118164, + "learning_rate": 3.336718115353372e-06, + "loss": 1.7164, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.5142838954925537, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6441, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.8628681898117065, + "eval_runtime": 36.8298, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.6917803287506104, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.6235, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.6062285900115967, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6488, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.5503156185150146, + "learning_rate": 3.32859463850528e-06, + "loss": 1.639, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.433518648147583, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5951, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.600661516189575, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.6148, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.5558395385742188, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.6334, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.6363396644592285, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5858, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.819617748260498, + "learning_rate": 3.318440292445167e-06, + "loss": 1.6274, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 2.9121081829071045, + "learning_rate": 3.316409423233144e-06, + "loss": 1.6081, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.734858989715576, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6889, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8644160032272339, + "eval_runtime": 37.0389, + "eval_samples_per_second": 26.999, + "eval_steps_per_second": 1.134, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.6940014362335205, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6463, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.82663631439209, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6474, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.634434223175049, + "learning_rate": 3.308285946385053e-06, + "loss": 1.7172, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.6721079349517822, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5795, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.786130666732788, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6634, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7698960304260254, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.687, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.6305580139160156, + "learning_rate": 3.300162469536962e-06, + "loss": 1.6372, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.5221288204193115, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6746, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.583643674850464, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6538, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.5978968143463135, + "learning_rate": 3.294069861900894e-06, + "loss": 1.7177, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8665071725845337, + "eval_runtime": 37.1194, + "eval_samples_per_second": 26.94, + "eval_steps_per_second": 1.131, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.6446492671966553, + "learning_rate": 3.292038992688871e-06, + "loss": 1.657, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.646174192428589, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.682, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.671532392501831, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.7033, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.616450309753418, + "learning_rate": 3.285946385052803e-06, + "loss": 1.586, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.552800416946411, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6599, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.5700533390045166, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5593, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.675553560256958, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.7097, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.6610019207000732, + "learning_rate": 3.277822908204712e-06, + "loss": 1.6408, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.55435848236084, + "learning_rate": 3.275792038992689e-06, + "loss": 1.5909, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.811556816101074, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6552, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.8618569374084473, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.7060115337371826, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.6144, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.6817731857299805, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6481, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.8206212520599365, + "learning_rate": 3.267668562144598e-06, + "loss": 1.6294, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.834332227706909, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.6103, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.754049062728882, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6788, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.6089437007904053, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6469, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.6095848083496094, + "learning_rate": 3.259545085296507e-06, + "loss": 1.7122, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.606787919998169, + "learning_rate": 3.257514216084484e-06, + "loss": 1.5851, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.643245220184326, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.7042, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.8400895595550537, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6526, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8640165328979492, + "eval_runtime": 37.162, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.519524335861206, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.6268, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.565365791320801, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6562, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.6364638805389404, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.6143, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.6235382556915283, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.64, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.7358779907226562, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6509, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.6720035076141357, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5902, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5960681438446045, + "learning_rate": 3.23923639317628e-06, + "loss": 1.6378, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.6655092239379883, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.6102, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.7873294353485107, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6558, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.746189594268799, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.7116, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8605948686599731, + "eval_runtime": 37.1486, + "eval_samples_per_second": 26.919, + "eval_steps_per_second": 1.131, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5058770179748535, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.6156, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.8002450466156006, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6884, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.696186065673828, + "learning_rate": 3.227051177904143e-06, + "loss": 1.6403, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.5473484992980957, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.6375, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.6391875743865967, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6597, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.647244453430176, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5666, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849031448364258, + "learning_rate": 3.218927701056052e-06, + "loss": 1.6211, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.7565410137176514, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5686, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.938270092010498, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.7046, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.6462385654449463, + "learning_rate": 3.212835093419984e-06, + "loss": 1.602, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.8577098846435547, + "eval_runtime": 37.1448, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.679330587387085, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.636, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.6410343647003174, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6904, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6398510932922363, + "learning_rate": 3.206742485783916e-06, + "loss": 1.597, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.4636199474334717, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5823, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.646580457687378, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6591, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.8582923412323, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.6731, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.5474820137023926, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5656, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6609606742858887, + "learning_rate": 3.196588139723802e-06, + "loss": 1.6374, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.651947021484375, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6557, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.6957671642303467, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.6436, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.8583322763442993, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.5442283153533936, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6424, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.593792200088501, + "learning_rate": 3.188464662875711e-06, + "loss": 1.7012, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.5021846294403076, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.6268, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.7368624210357666, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5953, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.5100302696228027, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5866, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.5767033100128174, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6621, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.4694652557373047, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6554, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.606158971786499, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.6238, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.857424020767212, + "learning_rate": 3.174248578391552e-06, + "loss": 1.626, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.6120405197143555, + "learning_rate": 3.172217709179529e-06, + "loss": 1.6134, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8599685430526733, + "eval_runtime": 37.1373, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.61995530128479, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.6303, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6253163814544678, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5886, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.627302885055542, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6526, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.8007607460021973, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.6581, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.551227331161499, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5921, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.7288761138916016, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6478, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5698401927948, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6552, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.796701192855835, + "learning_rate": 3.155970755483347e-06, + "loss": 1.6391, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.6523337364196777, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6495, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.5528745651245117, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6396, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8551454544067383, + "eval_runtime": 37.1721, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.650125741958618, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6803, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.5522522926330566, + "learning_rate": 3.147847278635256e-06, + "loss": 1.726, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5302486419677734, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.6238, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.517793893814087, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5879, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.5749523639678955, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.6361, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6382429599761963, + "learning_rate": 3.139723801787165e-06, + "loss": 1.6143, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.7513139247894287, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6608, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.616007089614868, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5842, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.619389772415161, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6419, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.6845619678497314, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6726, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8519575595855713, + "eval_runtime": 36.823, + "eval_samples_per_second": 27.157, + "eval_steps_per_second": 1.141, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.9907562732696533, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6813, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5070035457611084, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6449, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.6286256313323975, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6614, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5735433101654053, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.6191, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.637507438659668, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6649, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.5027501583099365, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6779, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.8430395126342773, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.6428, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.6394317150115967, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.655, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.7121026515960693, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6798, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.6572396755218506, + "learning_rate": 3.111291632818847e-06, + "loss": 1.6937, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.857973575592041, + "eval_runtime": 37.0635, + "eval_samples_per_second": 26.981, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.660865306854248, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6454, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.6230008602142334, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6562, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.60079026222229, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.6311, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.6757969856262207, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6673, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.6204211711883545, + "learning_rate": 3.101137286758733e-06, + "loss": 1.665, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.7115979194641113, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6864, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.604799270629883, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.6364, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 2.7463159561157227, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.699, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.744046926498413, + "learning_rate": 3.093013809910642e-06, + "loss": 1.665, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8082664012908936, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.7027, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8549033403396606, + "eval_runtime": 37.1604, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.6146464347839355, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6635, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.724843978881836, + "learning_rate": 3.086921202274574e-06, + "loss": 1.64, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.699645757675171, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5988, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.769742727279663, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6478, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.5372848510742188, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.6285, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6461033821105957, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5669, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.8221795558929443, + "learning_rate": 3.07676685621446e-06, + "loss": 1.6455, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6035211086273193, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.6063, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.533945322036743, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.6234, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.627448797225952, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6381, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8547505140304565, + "eval_runtime": 37.226, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.6822423934936523, + "learning_rate": 3.068643379366369e-06, + "loss": 1.6436, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.647813558578491, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5901, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.619610548019409, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6559, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7284183502197266, + "learning_rate": 3.062550771730301e-06, + "loss": 1.6167, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.585660934448242, + "learning_rate": 3.060519902518278e-06, + "loss": 1.6176, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.689528465270996, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5651, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.484328508377075, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.6487, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.629913806915283, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.6099, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.711730480194092, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6702, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.590545415878296, + "learning_rate": 3.050365556458164e-06, + "loss": 1.6078, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8545923233032227, + "eval_runtime": 37.2534, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.5952389240264893, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.6084, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.739316940307617, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.6427, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.811095952987671, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6817, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.606715440750122, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6695, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.481234550476074, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5762, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.508068084716797, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.612, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.5694284439086914, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.6457, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.7029778957366943, + "learning_rate": 3.034118602761982e-06, + "loss": 1.6201, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.578904867172241, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5978, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.780586004257202, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6793, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.8522270917892456, + "eval_runtime": 37.2388, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.648653984069824, + "learning_rate": 3.028025995125914e-06, + "loss": 1.6626, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.631589651107788, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5822, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.5863327980041504, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.6322, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6556999683380127, + "learning_rate": 3.021933387489846e-06, + "loss": 1.5727, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.6391682624816895, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6334, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.6488680839538574, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6588, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.743858814239502, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5963, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.6499438285827637, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5997, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5450377464294434, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.6122, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.592026948928833, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.6362, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.84926176071167, + "eval_runtime": 37.2501, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.5699622631073, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.6055, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.779695987701416, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6503, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.6080546379089355, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.6121, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.70405650138855, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6774, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.69488525390625, + "learning_rate": 2.999593826157596e-06, + "loss": 1.6186, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.9266953468322754, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.6485, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.615251302719116, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.624, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.574878215789795, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6637, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6952786445617676, + "learning_rate": 2.991470349309505e-06, + "loss": 1.627, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.6076977252960205, + "learning_rate": 2.989439480097482e-06, + "loss": 1.6175, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.8475773334503174, + "eval_runtime": 37.2328, + "eval_samples_per_second": 26.858, + "eval_steps_per_second": 1.128, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.5707366466522217, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.6332, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.838609457015991, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6764, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.771082878112793, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6622, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6113975048065186, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6405, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.7537200450897217, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.6376, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.683002471923828, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6679, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.651897430419922, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.6091, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.6856229305267334, + "learning_rate": 2.9731925264013e-06, + "loss": 1.6223, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.646522045135498, + "learning_rate": 2.971161657189277e-06, + "loss": 1.6126, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.592247486114502, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.6088, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.8501410484313965, + "eval_runtime": 37.225, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.03291392326355, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.6051, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.903224468231201, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6753, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.5178730487823486, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5894, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.6278252601623535, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6687, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.8105742931365967, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6648, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.5897178649902344, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.6012, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.7172093391418457, + "learning_rate": 2.954914703493095e-06, + "loss": 1.6206, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.6845815181732178, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.6163, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.652451753616333, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.5916, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.740280866622925, + "learning_rate": 2.948822095857027e-06, + "loss": 1.627, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8489265441894531, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.786797285079956, + "learning_rate": 2.946791226645004e-06, + "loss": 1.6005, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.6826601028442383, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.5887, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 2.881887912750244, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.6198, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.86600923538208, + "learning_rate": 2.940698619008936e-06, + "loss": 1.6252, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.7267069816589355, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.6314, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.768399477005005, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.6434, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.70622181892395, + "learning_rate": 2.934606011372868e-06, + "loss": 1.6412, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.803354501724243, + "learning_rate": 2.932575142160845e-06, + "loss": 1.6402, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8491063117980957, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.6244, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.5652339458465576, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.6315, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.8461908102035522, + "eval_runtime": 36.9233, + "eval_samples_per_second": 27.083, + "eval_steps_per_second": 1.137, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.7153713703155518, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.6714, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 2.779289484024048, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.6355, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.6597917079925537, + "learning_rate": 2.922420796100731e-06, + "loss": 1.5574, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.7135894298553467, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.6721, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.562045097351074, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.5745, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.6727097034454346, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.6324, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 2.688309907913208, + "learning_rate": 2.91429731925264e-06, + "loss": 1.6698, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.8429853916168213, + "learning_rate": 2.912266450040618e-06, + "loss": 1.6378, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.658046245574951, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.6237, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 2.6601390838623047, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.6163, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.843716025352478, + "eval_runtime": 37.2087, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.949215888977051, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.5883, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 2.823991298675537, + "learning_rate": 2.904142973192527e-06, + "loss": 1.6786, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.6177661418914795, + "learning_rate": 2.902112103980504e-06, + "loss": 1.6337, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.493638038635254, + "learning_rate": 2.900081234768481e-06, + "loss": 1.6328, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.657914161682129, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.6068, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.7800004482269287, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.5963, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.7934350967407227, + "learning_rate": 2.893988627132413e-06, + "loss": 1.611, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.8139140605926514, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.5621, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.5909969806671143, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.6195, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.631871461868286, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.6146, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.846065878868103, + "eval_runtime": 37.2421, + "eval_samples_per_second": 26.851, + "eval_steps_per_second": 1.128, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.667304754257202, + "learning_rate": 2.885865150284322e-06, + "loss": 1.6482, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.5922656059265137, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.6231, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.663452386856079, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.5772, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.6463005542755127, + "learning_rate": 2.879772542648254e-06, + "loss": 1.6333, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.5469250679016113, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.5465, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.7316017150878906, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.6856, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.7127692699432373, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.667, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 2.7006423473358154, + "learning_rate": 2.871649065800163e-06, + "loss": 1.677, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 2.788804054260254, + "learning_rate": 2.86961819658814e-06, + "loss": 1.6092, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.7734181880950928, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.6377, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.8444476127624512, + "eval_runtime": 37.2371, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.6005876064300537, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.6072, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.684293270111084, + "learning_rate": 2.863525588952072e-06, + "loss": 1.6669, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.6160528659820557, + "learning_rate": 2.861494719740049e-06, + "loss": 1.5439, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.68396258354187, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.5909, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.6292917728424072, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.6225, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.702333450317383, + "learning_rate": 2.855402112103981e-06, + "loss": 1.6224, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.850862503051758, + "learning_rate": 2.853371242891958e-06, + "loss": 1.5995, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.861351490020752, + "learning_rate": 2.851340373679935e-06, + "loss": 1.6254, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.7731313705444336, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.6555, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.669189691543579, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.6238, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.845217227935791, + "eval_runtime": 37.2809, + "eval_samples_per_second": 26.823, + "eval_steps_per_second": 1.127, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.64485239982605, + "learning_rate": 2.845247766043867e-06, + "loss": 1.5989, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.7558820247650146, + "learning_rate": 2.843216896831844e-06, + "loss": 1.5615, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.7818500995635986, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.5962, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.571547746658325, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.6434, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 2.8242406845092773, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.6537, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.5545597076416016, + "learning_rate": 2.835093419983753e-06, + "loss": 1.6158, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.7329461574554443, + "learning_rate": 2.833062550771731e-06, + "loss": 1.591, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.5414955615997314, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.5727, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.7625439167022705, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.6343, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.7839064598083496, + "learning_rate": 2.826969943135662e-06, + "loss": 1.6187, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.8442766666412354, + "eval_runtime": 37.2101, + "eval_samples_per_second": 26.874, + "eval_steps_per_second": 1.129, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.615706443786621, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.621, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.6295979022979736, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.5913, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.7428646087646484, + "learning_rate": 2.820877335499594e-06, + "loss": 1.6015, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.5353381633758545, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.5798, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.6209659576416016, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.5783, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.791372299194336, + "learning_rate": 2.814784727863526e-06, + "loss": 1.6082, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.781949520111084, + "learning_rate": 2.812753858651503e-06, + "loss": 1.6337, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 3.0437769889831543, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.5969, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 2.7201249599456787, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.5717, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 2.766634464263916, + "learning_rate": 2.806661251015435e-06, + "loss": 1.6654, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.8452297449111938, + "eval_runtime": 37.2452, + "eval_samples_per_second": 26.849, + "eval_steps_per_second": 1.128, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.726876735687256, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.6652, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 2.775012969970703, + "learning_rate": 2.802599512591389e-06, + "loss": 1.7293, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.7603225708007812, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.6226, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.723904609680176, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.5663, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 2.667384624481201, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.6209, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.7580318450927734, + "learning_rate": 2.794476035743298e-06, + "loss": 1.6705, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.9216315746307373, + "learning_rate": 2.792445166531276e-06, + "loss": 1.6361, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 2.6552748680114746, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.6498, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.764836549758911, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.6009, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.5660064220428467, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.576, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.8392945528030396, + "eval_runtime": 37.2515, + "eval_samples_per_second": 26.845, + "eval_steps_per_second": 1.127, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.5110044479370117, + "learning_rate": 2.784321689683185e-06, + "loss": 1.5744, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.7352850437164307, + "learning_rate": 2.782290820471162e-06, + "loss": 1.5852, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.850947618484497, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.6692, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.7107245922088623, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.6427, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.8236117362976074, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.6375, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.763416051864624, + "learning_rate": 2.774167343623071e-06, + "loss": 1.6864, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 2.7317910194396973, + "learning_rate": 2.772136474411048e-06, + "loss": 1.6015, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.7454333305358887, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.5379, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.6102535724639893, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.6282, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.658440113067627, + "learning_rate": 2.76604386677498e-06, + "loss": 1.6042, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.8391484022140503, + "eval_runtime": 37.1746, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 2.5731043815612793, + "learning_rate": 2.764012997562957e-06, + "loss": 1.6446, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.680983543395996, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.5663, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 2.7525506019592285, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.6628, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.6905839443206787, + "learning_rate": 2.757920389926889e-06, + "loss": 1.6021, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 2.850900888442993, + "learning_rate": 2.755889520714866e-06, + "loss": 1.5984, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.6612954139709473, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.5627, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.632106304168701, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.6304, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.660377025604248, + "learning_rate": 2.749796913078798e-06, + "loss": 1.6622, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.497938394546509, + "learning_rate": 2.747766043866775e-06, + "loss": 1.6152, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.655977725982666, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.6703, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.8361152410507202, + "eval_runtime": 37.2147, + "eval_samples_per_second": 26.871, + "eval_steps_per_second": 1.129, + "step": 1400 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/latest new file mode 100644 index 0000000000000000000000000000000000000000..c56ff7708f44fb7928fea2f70d6d7342ce0d5b67 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/latest @@ -0,0 +1 @@ +global_step1500 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..1734dab7fa0865b1506b69de6849294bfd031f70 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/trainer_state.json @@ -0,0 +1,11733 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.643835616438356, + "eval_steps": 10, + "global_step": 1500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6008102893829346, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7848, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.630127429962158, + "learning_rate": 4.362307067424858e-06, + "loss": 1.7135, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.720945358276367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7897, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.898836851119995, + "learning_rate": 4.358245329000813e-06, + "loss": 1.7434, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.6548235416412354, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7683, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.7218399047851562, + "learning_rate": 4.354183590576767e-06, + "loss": 1.7729, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.5972483158111572, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7517, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.6741726398468018, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7354, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.898418426513672, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7843, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.3964333534240723, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.808, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.9525212049484253, + "eval_runtime": 36.8297, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.9952592849731445, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7856, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.81740140914917, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7613, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6271541118621826, + "learning_rate": 4.339967506092607e-06, + "loss": 1.8409, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.540395498275757, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.7765, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.629131317138672, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7634, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.5796725749969482, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7414, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6980209350585938, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.8048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.658919095993042, + "learning_rate": 4.329813160032495e-06, + "loss": 1.7725, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5237927436828613, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7559, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.5822298526763916, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7129, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9492865800857544, + "eval_runtime": 37.0012, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.762810468673706, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7895, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.48427414894104, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7667, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.790052890777588, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7548, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.927957773208618, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.7489, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.4964418411254883, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7947, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.4845147132873535, + "learning_rate": 4.313566206336312e-06, + "loss": 1.7886, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.687361717224121, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7968, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.568202018737793, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7332, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.496429443359375, + "learning_rate": 4.307473598700244e-06, + "loss": 1.7583, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6063425540924072, + "learning_rate": 4.305442729488221e-06, + "loss": 1.779, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.9421359300613403, + "eval_runtime": 37.0712, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.5371274948120117, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7133, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.5601038932800293, + "learning_rate": 4.301380991064176e-06, + "loss": 1.8089, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.6244118213653564, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7983, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.4798192977905273, + "learning_rate": 4.29731925264013e-06, + "loss": 1.8221, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.483376979827881, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.8032, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.716745615005493, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7583, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.6685945987701416, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7365, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.465810537338257, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7682, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.7629129886627197, + "learning_rate": 4.287164906580017e-06, + "loss": 1.809, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5273563861846924, + "learning_rate": 4.285134037367994e-06, + "loss": 1.7464, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.9359965324401855, + "eval_runtime": 37.1317, + "eval_samples_per_second": 26.931, + "eval_steps_per_second": 1.131, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.514230966567993, + "learning_rate": 4.283103168155971e-06, + "loss": 1.7365, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.704042911529541, + "learning_rate": 4.281072298943948e-06, + "loss": 1.7268, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.4716262817382812, + "learning_rate": 4.279041429731926e-06, + "loss": 1.7212, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5695507526397705, + "learning_rate": 4.277010560519903e-06, + "loss": 1.7398, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.762845039367676, + "learning_rate": 4.27497969130788e-06, + "loss": 1.7158, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.571054220199585, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7685, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.908184289932251, + "learning_rate": 4.270917952883835e-06, + "loss": 1.7865, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.705479621887207, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7922, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.4418110847473145, + "learning_rate": 4.266856214459789e-06, + "loss": 1.7312, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.4619884490966797, + "learning_rate": 4.264825345247766e-06, + "loss": 1.8343, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.9352803230285645, + "eval_runtime": 37.0958, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5386202335357666, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6902, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9012045860290527, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7552, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.8028385639190674, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.6872, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.4505386352539062, + "learning_rate": 4.256701868399675e-06, + "loss": 1.75, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7551822662353516, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7612, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.596270799636841, + "learning_rate": 4.252640129975629e-06, + "loss": 1.7694, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.387662172317505, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.7776, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.7208738327026367, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7375, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.5317776203155518, + "learning_rate": 4.246547522339562e-06, + "loss": 1.6922, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.4803197383880615, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.741, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.9322715997695923, + "eval_runtime": 37.0962, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.5688276290893555, + "learning_rate": 4.242485783915517e-06, + "loss": 1.7518, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.4652810096740723, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6853, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.541903257369995, + "learning_rate": 4.23842404549147e-06, + "loss": 1.6841, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.377857208251953, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7312, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.418323040008545, + "learning_rate": 4.234362307067425e-06, + "loss": 1.6892, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5374410152435303, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7719, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7119228839874268, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.7819, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.5333240032196045, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7758, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.528696060180664, + "learning_rate": 4.226238830219334e-06, + "loss": 1.75, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.539039134979248, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7498, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.934200644493103, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.5384178161621094, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7615, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.601658821105957, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7645, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5359764099121094, + "learning_rate": 4.218115353371243e-06, + "loss": 1.7549, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.641118049621582, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7487, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.4146382808685303, + "learning_rate": 4.214053614947198e-06, + "loss": 1.7102, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4245245456695557, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7842, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.509824752807617, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6876, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.636444330215454, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7681, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.476858615875244, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7049, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.75732159614563, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7833, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.931411623954773, + "eval_runtime": 37.0994, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.5437941551208496, + "learning_rate": 4.201868399675061e-06, + "loss": 1.7751, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.5010673999786377, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7715, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.6352853775024414, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6798, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.588247776031494, + "learning_rate": 4.195775792038993e-06, + "loss": 1.7575, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.349921226501465, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7245, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.394432783126831, + "learning_rate": 4.191714053614948e-06, + "loss": 1.6778, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.883906602859497, + "learning_rate": 4.189683184402924e-06, + "loss": 1.7701, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.537799596786499, + "learning_rate": 4.187652315190902e-06, + "loss": 1.7286, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.575700283050537, + "learning_rate": 4.185621445978879e-06, + "loss": 1.7812, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.592252492904663, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7176, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.9259952306747437, + "eval_runtime": 37.144, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.511845111846924, + "learning_rate": 4.181559707554833e-06, + "loss": 1.7564, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.575446367263794, + "learning_rate": 4.179528838342811e-06, + "loss": 1.7943, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.379391670227051, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7098, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.701812982559204, + "learning_rate": 4.175467099918766e-06, + "loss": 1.7292, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.486640453338623, + "learning_rate": 4.173436230706742e-06, + "loss": 1.7963, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.561983823776245, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.7153, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.5709176063537598, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7412, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.3308839797973633, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7188, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.592470169067383, + "learning_rate": 4.165312753858652e-06, + "loss": 1.7599, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.3747010231018066, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7361, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.9229093790054321, + "eval_runtime": 37.1417, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.4062509536743164, + "learning_rate": 4.161251015434607e-06, + "loss": 1.7256, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.4951181411743164, + "learning_rate": 4.159220146222583e-06, + "loss": 1.7622, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.3453004360198975, + "learning_rate": 4.157189277010561e-06, + "loss": 1.6969, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.382671594619751, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7109, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4277336597442627, + "learning_rate": 4.153127538586516e-06, + "loss": 1.7471, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.4476678371429443, + "learning_rate": 4.151096669374492e-06, + "loss": 1.7483, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.451099395751953, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7366, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.618217706680298, + "learning_rate": 4.147034930950447e-06, + "loss": 1.7282, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.414400100708008, + "learning_rate": 4.145004061738425e-06, + "loss": 1.7409, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.535353422164917, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.7346, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.9240931272506714, + "eval_runtime": 36.844, + "eval_samples_per_second": 27.141, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.5075416564941406, + "learning_rate": 4.140942323314379e-06, + "loss": 1.6668, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.4136979579925537, + "learning_rate": 4.138911454102356e-06, + "loss": 1.696, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.4363577365875244, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7174, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.4578073024749756, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7621, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.370976686477661, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7443, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.363755702972412, + "learning_rate": 4.130787977254265e-06, + "loss": 1.7399, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.44704532623291, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7481, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.5463340282440186, + "learning_rate": 4.12672623883022e-06, + "loss": 1.7646, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7088072299957275, + "learning_rate": 4.124695369618197e-06, + "loss": 1.8108, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.463271141052246, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7311, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.918715238571167, + "eval_runtime": 37.0468, + "eval_samples_per_second": 26.993, + "eval_steps_per_second": 1.134, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6165266036987305, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.7295, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.3432133197784424, + "learning_rate": 4.118602761982129e-06, + "loss": 1.7512, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.552751302719116, + "learning_rate": 4.116571892770106e-06, + "loss": 1.6976, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7586560249328613, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6708, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.5326449871063232, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7824, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.7658956050872803, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7093, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.6868369579315186, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7475, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.543748140335083, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.7714, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.7235589027404785, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7593, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.697038173675537, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6296, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.91779363155365, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 2.6746180057525635, + "learning_rate": 4.100324939073924e-06, + "loss": 1.7971, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.6281094551086426, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7212, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.446025848388672, + "learning_rate": 4.096263200649879e-06, + "loss": 1.7324, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.5460824966430664, + "learning_rate": 4.094232331437856e-06, + "loss": 1.7435, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.5419771671295166, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6979, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.5545520782470703, + "learning_rate": 4.090170593013811e-06, + "loss": 1.7042, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.8353254795074463, + "learning_rate": 4.088139723801787e-06, + "loss": 1.7313, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.6521668434143066, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6901, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.6611177921295166, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7064, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.411482572555542, + "learning_rate": 4.08204711616572e-06, + "loss": 1.729, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.9176616668701172, + "eval_runtime": 37.1603, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.689110279083252, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7707, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.544018268585205, + "learning_rate": 4.077985377741674e-06, + "loss": 1.6724, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.55307936668396, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7677, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.3896875381469727, + "learning_rate": 4.073923639317629e-06, + "loss": 1.7729, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.381887197494507, + "learning_rate": 4.071892770105605e-06, + "loss": 1.7737, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.4873931407928467, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6928, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.4908432960510254, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6674, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.4346399307250977, + "learning_rate": 4.065800162469538e-06, + "loss": 1.6947, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6459081172943115, + "learning_rate": 4.063769293257514e-06, + "loss": 1.701, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.6092610359191895, + "learning_rate": 4.061738424045492e-06, + "loss": 1.7425, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.919799566268921, + "eval_runtime": 37.1632, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.4443421363830566, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7267, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.8252694606781006, + "learning_rate": 4.057676685621446e-06, + "loss": 1.7412, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.3391194343566895, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.7916, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.57979679107666, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7438, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.398383855819702, + "learning_rate": 4.051584077985378e-06, + "loss": 1.6606, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.394498348236084, + "learning_rate": 4.049553208773355e-06, + "loss": 1.736, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5070464611053467, + "learning_rate": 4.047522339561333e-06, + "loss": 1.7559, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.439722776412964, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7133, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.569152593612671, + "learning_rate": 4.043460601137287e-06, + "loss": 1.7123, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.637657642364502, + "learning_rate": 4.041429731925264e-06, + "loss": 1.8267, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.9148659706115723, + "eval_runtime": 37.2248, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.441269636154175, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7106, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.582172155380249, + "learning_rate": 4.037367993501219e-06, + "loss": 1.7167, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.5083518028259277, + "learning_rate": 4.035337124289196e-06, + "loss": 1.8031, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.504112958908081, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.694, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.5068695545196533, + "learning_rate": 4.031275385865151e-06, + "loss": 1.6798, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.600541114807129, + "learning_rate": 4.029244516653128e-06, + "loss": 1.7346, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.503020763397217, + "learning_rate": 4.027213647441105e-06, + "loss": 1.7507, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.7774016857147217, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6737, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.5771589279174805, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6943, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.3841609954833984, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7142, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.9146945476531982, + "eval_runtime": 37.1701, + "eval_samples_per_second": 26.903, + "eval_steps_per_second": 1.13, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.466207981109619, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.7192, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.72037935256958, + "learning_rate": 4.017059301380992e-06, + "loss": 1.8082, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.36811900138855, + "learning_rate": 4.015028432168969e-06, + "loss": 1.6964, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.4803504943847656, + "learning_rate": 4.012997562956945e-06, + "loss": 1.779, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.707174777984619, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6715, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.462543249130249, + "learning_rate": 4.0089358245329e-06, + "loss": 1.7117, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.481444835662842, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7007, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.4432222843170166, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.703, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.611387014389038, + "learning_rate": 4.002843216896833e-06, + "loss": 1.7722, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.456305742263794, + "learning_rate": 4.000812347684809e-06, + "loss": 1.7615, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.9089828729629517, + "eval_runtime": 37.1812, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4287772178649902, + "learning_rate": 3.998781478472787e-06, + "loss": 1.7098, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5659544467926025, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7143, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.467067003250122, + "learning_rate": 3.994719740048742e-06, + "loss": 1.6668, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.5660688877105713, + "learning_rate": 3.992688870836718e-06, + "loss": 1.8185, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.6185719966888428, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7356, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.5389695167541504, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6973, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.499234914779663, + "learning_rate": 3.98659626320065e-06, + "loss": 1.7539, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.6246273517608643, + "learning_rate": 3.984565393988627e-06, + "loss": 1.7055, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6127917766571045, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7177, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.4941787719726562, + "learning_rate": 3.980503655564582e-06, + "loss": 1.728, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.9063301086425781, + "eval_runtime": 37.3284, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.4694740772247314, + "learning_rate": 3.978472786352559e-06, + "loss": 1.7384, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.8288261890411377, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7409, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.518359661102295, + "learning_rate": 3.974411047928514e-06, + "loss": 1.714, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4874095916748047, + "learning_rate": 3.972380178716491e-06, + "loss": 1.781, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.523772716522217, + "learning_rate": 3.970349309504468e-06, + "loss": 1.7474, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.406132459640503, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.7486, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.388207197189331, + "learning_rate": 3.966287571080423e-06, + "loss": 1.67, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.5564754009246826, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6803, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.531890869140625, + "learning_rate": 3.962225832656377e-06, + "loss": 1.7368, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.404284954071045, + "learning_rate": 3.960194963444355e-06, + "loss": 1.7247, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.9008547067642212, + "eval_runtime": 37.1962, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.3903231620788574, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6465, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7126, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7174, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.6466639041900635, + "learning_rate": 3.954102355808286e-06, + "loss": 1.7166, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.530284881591797, + "learning_rate": 3.952071486596264e-06, + "loss": 1.7155, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.601516008377075, + "learning_rate": 3.950040617384241e-06, + "loss": 1.7157, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.6146240234375, + "learning_rate": 3.948009748172218e-06, + "loss": 1.6931, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.4904634952545166, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.7594, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.345689296722412, + "learning_rate": 3.943948009748173e-06, + "loss": 1.7175, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4469082355499268, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7161, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.9021154642105103, + "eval_runtime": 36.8772, + "eval_samples_per_second": 27.117, + "eval_steps_per_second": 1.139, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.473752737045288, + "learning_rate": 3.939886271324127e-06, + "loss": 1.7688, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.461033582687378, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6752, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.419179916381836, + "learning_rate": 3.935824532900082e-06, + "loss": 1.6685, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.543102502822876, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7348, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.4932174682617188, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.7572, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.4563698768615723, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6691, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.566558837890625, + "learning_rate": 3.927701056051991e-06, + "loss": 1.697, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.5550761222839355, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.7194, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.51362681388855, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6804, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.448714017868042, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6575, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.904768943786621, + "eval_runtime": 37.0644, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.582794666290283, + "learning_rate": 3.9195775792039e-06, + "loss": 1.717, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.613327980041504, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.6927, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.522038221359253, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7391, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.372134208679199, + "learning_rate": 3.913484971567831e-06, + "loss": 1.7448, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.246591329574585, + "learning_rate": 3.911454102355808e-06, + "loss": 1.6531, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.48276948928833, + "learning_rate": 3.909423233143786e-06, + "loss": 1.7268, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.7920989990234375, + "learning_rate": 3.907392363931763e-06, + "loss": 1.7155, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.5430562496185303, + "learning_rate": 3.90536149471974e-06, + "loss": 1.7866, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.5819969177246094, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.8024, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.726024866104126, + "learning_rate": 3.901299756295695e-06, + "loss": 1.8467, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.899052381515503, + "eval_runtime": 37.192, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.6913323402404785, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6976, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.5307703018188477, + "learning_rate": 3.897238017871649e-06, + "loss": 1.7072, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.4976935386657715, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.7183, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.435443878173828, + "learning_rate": 3.893176279447604e-06, + "loss": 1.6845, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.704554319381714, + "learning_rate": 3.891145410235581e-06, + "loss": 1.78, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.4770138263702393, + "learning_rate": 3.889114541023558e-06, + "loss": 1.7436, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.472780227661133, + "learning_rate": 3.887083671811536e-06, + "loss": 1.7007, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.4693336486816406, + "learning_rate": 3.885052802599513e-06, + "loss": 1.6487, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.458834648132324, + "learning_rate": 3.88302193338749e-06, + "loss": 1.7895, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.4430108070373535, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.72, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.9009006023406982, + "eval_runtime": 37.1906, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.6104867458343506, + "learning_rate": 3.878960194963445e-06, + "loss": 1.6938, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.5191142559051514, + "learning_rate": 3.876929325751422e-06, + "loss": 1.7528, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.3788414001464844, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6446, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.485631227493286, + "learning_rate": 3.872867587327377e-06, + "loss": 1.7087, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.3721091747283936, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6752, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4648489952087402, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6883, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.447054862976074, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.7109, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.70007061958313, + "learning_rate": 3.864744110479286e-06, + "loss": 1.7342, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.6336557865142822, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7308, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.807671070098877, + "learning_rate": 3.86068237205524e-06, + "loss": 1.7519, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8990603685379028, + "eval_runtime": 37.1643, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.7177038192749023, + "learning_rate": 3.858651502843217e-06, + "loss": 1.7065, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.570364236831665, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6529, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.590329170227051, + "learning_rate": 3.854589764419171e-06, + "loss": 1.6955, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.542722702026367, + "learning_rate": 3.852558895207149e-06, + "loss": 1.7803, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.6095776557922363, + "learning_rate": 3.850528025995126e-06, + "loss": 1.7267, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.550161838531494, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6767, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.5079615116119385, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6897, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.5113067626953125, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.7024, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.559811592102051, + "learning_rate": 3.842404549147035e-06, + "loss": 1.695, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.507901191711426, + "learning_rate": 3.840373679935013e-06, + "loss": 1.661, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8952044248580933, + "eval_runtime": 37.2303, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.288231134414673, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.7095, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.260040044784546, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6858, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.5524179935455322, + "learning_rate": 3.834281072298944e-06, + "loss": 1.7362, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.419013738632202, + "learning_rate": 3.832250203086921e-06, + "loss": 1.6803, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.5345773696899414, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7435, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.5771822929382324, + "learning_rate": 3.828188464662876e-06, + "loss": 1.7932, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.4880053997039795, + "learning_rate": 3.826157595450853e-06, + "loss": 1.711, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.596400022506714, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6569, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.620776891708374, + "learning_rate": 3.822095857026808e-06, + "loss": 1.7159, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.512397527694702, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6832, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8950546979904175, + "eval_runtime": 37.1526, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5925326347351074, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6977, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.3980817794799805, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.6997, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.4825663566589355, + "learning_rate": 3.813972380178717e-06, + "loss": 1.7063, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.535006284713745, + "learning_rate": 3.811941510966694e-06, + "loss": 1.7457, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.450657606124878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.6521, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.3942458629608154, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6194, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5781140327453613, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.7105, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.5196564197540283, + "learning_rate": 3.803818034118603e-06, + "loss": 1.6741, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4819812774658203, + "learning_rate": 3.80178716490658e-06, + "loss": 1.7559, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.6393206119537354, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.7421, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.8945845365524292, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5286684036254883, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6933, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.690138339996338, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.6804, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.7789907455444336, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7839, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.5196211338043213, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7564, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.550898790359497, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.758, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.617361545562744, + "learning_rate": 3.787571080422421e-06, + "loss": 1.7497, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.533339500427246, + "learning_rate": 3.785540211210398e-06, + "loss": 1.7336, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.475037097930908, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.712, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.485506534576416, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.6985, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.712912082672119, + "learning_rate": 3.77944760357433e-06, + "loss": 1.812, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8921257257461548, + "eval_runtime": 37.1309, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.6280291080474854, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.7504, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6162853240966797, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6793, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.665316581726074, + "learning_rate": 3.773354995938262e-06, + "loss": 1.7333, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.531297445297241, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7298, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.543412446975708, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.7677, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.3839635848999023, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.7039, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.3417305946350098, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6206, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.4434802532196045, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.713, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.5223472118377686, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6587, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.577479124069214, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.7032, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8891643285751343, + "eval_runtime": 37.1123, + "eval_samples_per_second": 26.945, + "eval_steps_per_second": 1.132, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.468871593475342, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.732, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.5907723903656006, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.6866, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.4454426765441895, + "learning_rate": 3.753046303818034e-06, + "loss": 1.746, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.4651715755462646, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7375, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6342053413391113, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.7693, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.454977512359619, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6796, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.440626859664917, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7075, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.4905781745910645, + "learning_rate": 3.742891957757921e-06, + "loss": 1.7422, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6843647956848145, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7107, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.5906362533569336, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.7545, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8889163732528687, + "eval_runtime": 36.8136, + "eval_samples_per_second": 27.164, + "eval_steps_per_second": 1.141, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.5185422897338867, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6857, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.4425783157348633, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.687, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.5200576782226562, + "learning_rate": 3.732737611697807e-06, + "loss": 1.643, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.4664227962493896, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.5965, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.495962381362915, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6663, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.5607051849365234, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.6388, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 2.9295718669891357, + "learning_rate": 3.724614134849716e-06, + "loss": 1.7541, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.5632426738739014, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6359, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.6059279441833496, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.7427, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.6244208812713623, + "learning_rate": 3.718521527213648e-06, + "loss": 1.7097, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8882938623428345, + "eval_runtime": 36.9853, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.434684991836548, + "learning_rate": 3.716490658001625e-06, + "loss": 1.6875, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.6481709480285645, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5726, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.5398213863372803, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.6881, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.684122085571289, + "learning_rate": 3.710398050365557e-06, + "loss": 1.691, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.712214946746826, + "learning_rate": 3.708367181153534e-06, + "loss": 1.653, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.8427517414093018, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.6853, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.676410436630249, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.645, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.5397610664367676, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6171, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.606250286102295, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6729, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.3617212772369385, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.6546, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8915373086929321, + "eval_runtime": 37.094, + "eval_samples_per_second": 26.959, + "eval_steps_per_second": 1.132, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.6635043621063232, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.628, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.724837303161621, + "learning_rate": 3.694151096669375e-06, + "loss": 1.7537, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.56447434425354, + "learning_rate": 3.692120227457352e-06, + "loss": 1.6638, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6624584197998047, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.6806, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.7147905826568604, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5887, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.683840751647949, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.6664, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.8141531944274902, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6744, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.6172802448272705, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6866, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.7853286266326904, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.7034, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.7121260166168213, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6573, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.883426308631897, + "eval_runtime": 37.1964, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.4775121212005615, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6651, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.5361502170562744, + "learning_rate": 3.673842404549147e-06, + "loss": 1.6524, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.7969970703125, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6598, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.88427996635437, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.6821, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.9321346282958984, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.7777, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.802886724472046, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6808, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.85810923576355, + "learning_rate": 3.663688058489034e-06, + "loss": 1.7375, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.6501998901367188, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.6662, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.550652265548706, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6084, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.5003185272216797, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.681, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8800761699676514, + "eval_runtime": 37.2014, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.4262983798980713, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6608, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6474270820617676, + "learning_rate": 3.65353371242892e-06, + "loss": 1.6972, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.5103557109832764, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5581, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.529756546020508, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.693, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.578362464904785, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6302, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8062336444854736, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6855, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.935694932937622, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6861, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5260093212127686, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6534, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.7627487182617188, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.6833, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.9553534984588623, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6887, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.8832038640975952, + "eval_runtime": 37.1828, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.5526084899902344, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.7162, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.761218786239624, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6823, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.672189950942993, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6489, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.7650046348571777, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.7201, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.6351473331451416, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.6164, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5959131717681885, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.603, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.525513172149658, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6671, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.5529232025146484, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6446, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6846, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6786, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.885197639465332, + "eval_runtime": 37.2374, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.50577449798584, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6892, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.57772159576416, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5863, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.5455222129821777, + "learning_rate": 3.612916328188465e-06, + "loss": 1.692, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.6154329776763916, + "learning_rate": 3.610885458976442e-06, + "loss": 1.7027, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.5776023864746094, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.6353, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.5480730533599854, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.6358, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.5939579010009766, + "learning_rate": 3.604792851340374e-06, + "loss": 1.7099, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.65395188331604, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6763, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.5302748680114746, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.613, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.669433832168579, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.6584, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8816337585449219, + "eval_runtime": 37.2174, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.859063148498535, + "learning_rate": 3.596669374492283e-06, + "loss": 1.7096, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.531301259994507, + "learning_rate": 3.59463850528026e-06, + "loss": 1.6168, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.5920488834381104, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6953, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.698284387588501, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6602, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.762387752532959, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.7073, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.554579019546509, + "learning_rate": 3.586515028432169e-06, + "loss": 1.7151, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.644576072692871, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.6088, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.6375787258148193, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6434, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9475791454315186, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6936, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.5753467082977295, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5915, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8799077272415161, + "eval_runtime": 37.2335, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8163156509399414, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6824, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.6741368770599365, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.7066, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.5235214233398438, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6779, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.620927095413208, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6392, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7678003311157227, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6768, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.5900607109069824, + "learning_rate": 3.566206336311942e-06, + "loss": 1.618, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.4440128803253174, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5981, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.686119318008423, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6649, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.6912872791290283, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.6426, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.7249274253845215, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.6639, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8750032186508179, + "eval_runtime": 37.2319, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.750321626663208, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.7014, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.593047618865967, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6557, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.744931936264038, + "learning_rate": 3.551990251827783e-06, + "loss": 1.6257, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.4940085411071777, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6546, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.6905856132507324, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.6395, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.6744513511657715, + "learning_rate": 3.545897644191714e-06, + "loss": 1.6159, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6766908168792725, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6754, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.743103504180908, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.6273, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.886472463607788, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6836, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.823476791381836, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6475, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8762013912200928, + "eval_runtime": 36.8594, + "eval_samples_per_second": 27.13, + "eval_steps_per_second": 1.139, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.741985559463501, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6765, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.6321418285369873, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6997, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7359395027160645, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6529, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.610210657119751, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.68, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.7770233154296875, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6836, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.7146646976470947, + "learning_rate": 3.525588952071487e-06, + "loss": 1.699, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.782172679901123, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6941, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.577692747116089, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.6303, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.721348285675049, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.6381, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.825709342956543, + "learning_rate": 3.517465475223396e-06, + "loss": 1.6144, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8759764432907104, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.8446385860443115, + "learning_rate": 3.515434606011373e-06, + "loss": 1.685, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.70577335357666, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.7276, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.6624722480773926, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6735, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.7956180572509766, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6413, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.792559862136841, + "learning_rate": 3.507311129163282e-06, + "loss": 1.6035, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.7113680839538574, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6461, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.656461715698242, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.662, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.8650670051574707, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5952, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.8043570518493652, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6738, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.607752799987793, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6828, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8729931116104126, + "eval_runtime": 37.2275, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.5819315910339355, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.6108, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6494336128234863, + "learning_rate": 3.493095044679123e-06, + "loss": 1.7348, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.552762508392334, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5534, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.6580469608306885, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6675, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.922295331954956, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5819, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.7051825523376465, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6573, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.738743782043457, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6939, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.51552677154541, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.665, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.578549385070801, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6642, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.7047011852264404, + "learning_rate": 3.476848090982941e-06, + "loss": 1.6298, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8726528882980347, + "eval_runtime": 37.25, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.722877025604248, + "learning_rate": 3.474817221770918e-06, + "loss": 1.7522, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.464240074157715, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6876, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.6737945079803467, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.6453, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.7960164546966553, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.607, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.6985723972320557, + "learning_rate": 3.466693744922827e-06, + "loss": 1.6385, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.7561540603637695, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6753, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.828885555267334, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6402, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.653183698654175, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6889, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.5803189277648926, + "learning_rate": 3.458570268074736e-06, + "loss": 1.654, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.5868985652923584, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6366, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8656586408615112, + "eval_runtime": 37.2805, + "eval_samples_per_second": 26.824, + "eval_steps_per_second": 1.127, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.648036241531372, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.6019, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.5482382774353027, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6597, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.700294017791748, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.6075, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.9346179962158203, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6558, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.893852710723877, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6395, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.789057970046997, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6833, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.745492935180664, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.6266, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.7930359840393066, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5781, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.770575761795044, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6518, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 2.7367143630981445, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6982, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8708553314208984, + "eval_runtime": 37.2843, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.89182448387146, + "learning_rate": 3.434199837530463e-06, + "loss": 1.6587, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.5572729110717773, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5974, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8172996044158936, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6941, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.6480889320373535, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6902, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.654735803604126, + "learning_rate": 3.426076360682372e-06, + "loss": 1.7083, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.4182465076446533, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6494, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.6808230876922607, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5782, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.648282527923584, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6695, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.5920112133026123, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.7234, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.690538167953491, + "learning_rate": 3.415922014622259e-06, + "loss": 1.686, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8717143535614014, + "eval_runtime": 37.2714, + "eval_samples_per_second": 26.83, + "eval_steps_per_second": 1.127, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.672515392303467, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6883, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.8012726306915283, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6762, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.7551519870758057, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6694, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.687445878982544, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.6267, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.674445390701294, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6944, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.797832489013672, + "learning_rate": 3.403736799350122e-06, + "loss": 1.6735, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7843174934387207, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.6183, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.678809404373169, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.7336, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.748838424682617, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6499, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.5785276889801025, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6589, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8695456981658936, + "eval_runtime": 37.2493, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.7307844161987305, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5989, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.83357572555542, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6568, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.6178860664367676, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6654, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.699328899383545, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6657, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.7418158054351807, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6351, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.5589001178741455, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.6351, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.492795944213867, + "learning_rate": 3.381397238017872e-06, + "loss": 1.6188, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 2.8586132526397705, + "learning_rate": 3.379366368805849e-06, + "loss": 1.7127, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7773218154907227, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6529, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.6956539154052734, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.6566, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8692638874053955, + "eval_runtime": 37.2317, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.7406537532806396, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6482, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.746194362640381, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.6343, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.767970323562622, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6577, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.730640172958374, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5941, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.5661444664001465, + "learning_rate": 3.36515028432169e-06, + "loss": 1.6362, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.5915091037750244, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6526, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.7135071754455566, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6999, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.591201066970825, + "learning_rate": 3.359057676685622e-06, + "loss": 1.684, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.564439296722412, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5815, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.4480631351470947, + "learning_rate": 3.354995938261576e-06, + "loss": 1.6284, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8628013134002686, + "eval_runtime": 37.23, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.5392649173736572, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5841, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6359004974365234, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.6278, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.803671360015869, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6602, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.6031086444854736, + "learning_rate": 3.346872461413485e-06, + "loss": 1.6021, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.687382459640503, + "learning_rate": 3.344841592201463e-06, + "loss": 1.6207, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.6388001441955566, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.6194, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.603121042251587, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6532, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.8053836822509766, + "learning_rate": 3.338748984565394e-06, + "loss": 1.68, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.5472421646118164, + "learning_rate": 3.336718115353372e-06, + "loss": 1.7164, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.5142838954925537, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6441, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.8628681898117065, + "eval_runtime": 36.8298, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.6917803287506104, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.6235, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.6062285900115967, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6488, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.5503156185150146, + "learning_rate": 3.32859463850528e-06, + "loss": 1.639, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.433518648147583, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5951, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.600661516189575, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.6148, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.5558395385742188, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.6334, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.6363396644592285, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5858, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.819617748260498, + "learning_rate": 3.318440292445167e-06, + "loss": 1.6274, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 2.9121081829071045, + "learning_rate": 3.316409423233144e-06, + "loss": 1.6081, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.734858989715576, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6889, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8644160032272339, + "eval_runtime": 37.0389, + "eval_samples_per_second": 26.999, + "eval_steps_per_second": 1.134, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.6940014362335205, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6463, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.82663631439209, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6474, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.634434223175049, + "learning_rate": 3.308285946385053e-06, + "loss": 1.7172, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.6721079349517822, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5795, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.786130666732788, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6634, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7698960304260254, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.687, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.6305580139160156, + "learning_rate": 3.300162469536962e-06, + "loss": 1.6372, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.5221288204193115, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6746, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.583643674850464, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6538, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.5978968143463135, + "learning_rate": 3.294069861900894e-06, + "loss": 1.7177, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8665071725845337, + "eval_runtime": 37.1194, + "eval_samples_per_second": 26.94, + "eval_steps_per_second": 1.131, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.6446492671966553, + "learning_rate": 3.292038992688871e-06, + "loss": 1.657, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.646174192428589, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.682, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.671532392501831, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.7033, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.616450309753418, + "learning_rate": 3.285946385052803e-06, + "loss": 1.586, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.552800416946411, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6599, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.5700533390045166, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5593, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.675553560256958, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.7097, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.6610019207000732, + "learning_rate": 3.277822908204712e-06, + "loss": 1.6408, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.55435848236084, + "learning_rate": 3.275792038992689e-06, + "loss": 1.5909, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.811556816101074, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6552, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.8618569374084473, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.7060115337371826, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.6144, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.6817731857299805, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6481, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.8206212520599365, + "learning_rate": 3.267668562144598e-06, + "loss": 1.6294, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.834332227706909, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.6103, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.754049062728882, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6788, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.6089437007904053, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6469, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.6095848083496094, + "learning_rate": 3.259545085296507e-06, + "loss": 1.7122, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.606787919998169, + "learning_rate": 3.257514216084484e-06, + "loss": 1.5851, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.643245220184326, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.7042, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.8400895595550537, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6526, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8640165328979492, + "eval_runtime": 37.162, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.519524335861206, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.6268, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.565365791320801, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6562, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.6364638805389404, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.6143, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.6235382556915283, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.64, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.7358779907226562, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6509, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.6720035076141357, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5902, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5960681438446045, + "learning_rate": 3.23923639317628e-06, + "loss": 1.6378, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.6655092239379883, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.6102, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.7873294353485107, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6558, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.746189594268799, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.7116, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8605948686599731, + "eval_runtime": 37.1486, + "eval_samples_per_second": 26.919, + "eval_steps_per_second": 1.131, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5058770179748535, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.6156, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.8002450466156006, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6884, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.696186065673828, + "learning_rate": 3.227051177904143e-06, + "loss": 1.6403, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.5473484992980957, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.6375, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.6391875743865967, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6597, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.647244453430176, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5666, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849031448364258, + "learning_rate": 3.218927701056052e-06, + "loss": 1.6211, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.7565410137176514, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5686, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.938270092010498, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.7046, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.6462385654449463, + "learning_rate": 3.212835093419984e-06, + "loss": 1.602, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.8577098846435547, + "eval_runtime": 37.1448, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.679330587387085, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.636, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.6410343647003174, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6904, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6398510932922363, + "learning_rate": 3.206742485783916e-06, + "loss": 1.597, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.4636199474334717, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5823, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.646580457687378, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6591, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.8582923412323, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.6731, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.5474820137023926, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5656, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6609606742858887, + "learning_rate": 3.196588139723802e-06, + "loss": 1.6374, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.651947021484375, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6557, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.6957671642303467, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.6436, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.8583322763442993, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.5442283153533936, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6424, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.593792200088501, + "learning_rate": 3.188464662875711e-06, + "loss": 1.7012, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.5021846294403076, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.6268, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.7368624210357666, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5953, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.5100302696228027, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5866, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.5767033100128174, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6621, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.4694652557373047, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6554, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.606158971786499, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.6238, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.857424020767212, + "learning_rate": 3.174248578391552e-06, + "loss": 1.626, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.6120405197143555, + "learning_rate": 3.172217709179529e-06, + "loss": 1.6134, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8599685430526733, + "eval_runtime": 37.1373, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.61995530128479, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.6303, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6253163814544678, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5886, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.627302885055542, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6526, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.8007607460021973, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.6581, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.551227331161499, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5921, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.7288761138916016, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6478, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5698401927948, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6552, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.796701192855835, + "learning_rate": 3.155970755483347e-06, + "loss": 1.6391, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.6523337364196777, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6495, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.5528745651245117, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6396, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8551454544067383, + "eval_runtime": 37.1721, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.650125741958618, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6803, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.5522522926330566, + "learning_rate": 3.147847278635256e-06, + "loss": 1.726, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5302486419677734, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.6238, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.517793893814087, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5879, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.5749523639678955, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.6361, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6382429599761963, + "learning_rate": 3.139723801787165e-06, + "loss": 1.6143, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.7513139247894287, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6608, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.616007089614868, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5842, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.619389772415161, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6419, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.6845619678497314, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6726, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8519575595855713, + "eval_runtime": 36.823, + "eval_samples_per_second": 27.157, + "eval_steps_per_second": 1.141, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.9907562732696533, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6813, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5070035457611084, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6449, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.6286256313323975, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6614, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5735433101654053, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.6191, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.637507438659668, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6649, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.5027501583099365, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6779, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.8430395126342773, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.6428, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.6394317150115967, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.655, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.7121026515960693, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6798, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.6572396755218506, + "learning_rate": 3.111291632818847e-06, + "loss": 1.6937, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.857973575592041, + "eval_runtime": 37.0635, + "eval_samples_per_second": 26.981, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.660865306854248, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6454, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.6230008602142334, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6562, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.60079026222229, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.6311, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.6757969856262207, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6673, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.6204211711883545, + "learning_rate": 3.101137286758733e-06, + "loss": 1.665, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.7115979194641113, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6864, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.604799270629883, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.6364, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 2.7463159561157227, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.699, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.744046926498413, + "learning_rate": 3.093013809910642e-06, + "loss": 1.665, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8082664012908936, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.7027, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8549033403396606, + "eval_runtime": 37.1604, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.6146464347839355, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6635, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.724843978881836, + "learning_rate": 3.086921202274574e-06, + "loss": 1.64, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.699645757675171, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5988, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.769742727279663, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6478, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.5372848510742188, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.6285, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6461033821105957, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5669, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.8221795558929443, + "learning_rate": 3.07676685621446e-06, + "loss": 1.6455, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6035211086273193, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.6063, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.533945322036743, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.6234, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.627448797225952, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6381, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8547505140304565, + "eval_runtime": 37.226, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.6822423934936523, + "learning_rate": 3.068643379366369e-06, + "loss": 1.6436, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.647813558578491, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5901, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.619610548019409, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6559, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7284183502197266, + "learning_rate": 3.062550771730301e-06, + "loss": 1.6167, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.585660934448242, + "learning_rate": 3.060519902518278e-06, + "loss": 1.6176, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.689528465270996, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5651, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.484328508377075, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.6487, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.629913806915283, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.6099, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.711730480194092, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6702, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.590545415878296, + "learning_rate": 3.050365556458164e-06, + "loss": 1.6078, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8545923233032227, + "eval_runtime": 37.2534, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.5952389240264893, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.6084, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.739316940307617, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.6427, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.811095952987671, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6817, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.606715440750122, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6695, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.481234550476074, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5762, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.508068084716797, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.612, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.5694284439086914, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.6457, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.7029778957366943, + "learning_rate": 3.034118602761982e-06, + "loss": 1.6201, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.578904867172241, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5978, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.780586004257202, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6793, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.8522270917892456, + "eval_runtime": 37.2388, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.648653984069824, + "learning_rate": 3.028025995125914e-06, + "loss": 1.6626, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.631589651107788, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5822, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.5863327980041504, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.6322, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6556999683380127, + "learning_rate": 3.021933387489846e-06, + "loss": 1.5727, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.6391682624816895, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6334, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.6488680839538574, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6588, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.743858814239502, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5963, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.6499438285827637, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5997, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5450377464294434, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.6122, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.592026948928833, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.6362, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.84926176071167, + "eval_runtime": 37.2501, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.5699622631073, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.6055, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.779695987701416, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6503, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.6080546379089355, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.6121, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.70405650138855, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6774, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.69488525390625, + "learning_rate": 2.999593826157596e-06, + "loss": 1.6186, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.9266953468322754, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.6485, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.615251302719116, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.624, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.574878215789795, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6637, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6952786445617676, + "learning_rate": 2.991470349309505e-06, + "loss": 1.627, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.6076977252960205, + "learning_rate": 2.989439480097482e-06, + "loss": 1.6175, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.8475773334503174, + "eval_runtime": 37.2328, + "eval_samples_per_second": 26.858, + "eval_steps_per_second": 1.128, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.5707366466522217, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.6332, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.838609457015991, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6764, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.771082878112793, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6622, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6113975048065186, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6405, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.7537200450897217, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.6376, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.683002471923828, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6679, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.651897430419922, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.6091, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.6856229305267334, + "learning_rate": 2.9731925264013e-06, + "loss": 1.6223, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.646522045135498, + "learning_rate": 2.971161657189277e-06, + "loss": 1.6126, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.592247486114502, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.6088, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.8501410484313965, + "eval_runtime": 37.225, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.03291392326355, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.6051, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.903224468231201, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6753, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.5178730487823486, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5894, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.6278252601623535, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6687, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.8105742931365967, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6648, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.5897178649902344, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.6012, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.7172093391418457, + "learning_rate": 2.954914703493095e-06, + "loss": 1.6206, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.6845815181732178, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.6163, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.652451753616333, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.5916, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.740280866622925, + "learning_rate": 2.948822095857027e-06, + "loss": 1.627, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8489265441894531, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.786797285079956, + "learning_rate": 2.946791226645004e-06, + "loss": 1.6005, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.6826601028442383, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.5887, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 2.881887912750244, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.6198, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.86600923538208, + "learning_rate": 2.940698619008936e-06, + "loss": 1.6252, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.7267069816589355, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.6314, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.768399477005005, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.6434, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.70622181892395, + "learning_rate": 2.934606011372868e-06, + "loss": 1.6412, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.803354501724243, + "learning_rate": 2.932575142160845e-06, + "loss": 1.6402, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8491063117980957, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.6244, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.5652339458465576, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.6315, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.8461908102035522, + "eval_runtime": 36.9233, + "eval_samples_per_second": 27.083, + "eval_steps_per_second": 1.137, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.7153713703155518, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.6714, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 2.779289484024048, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.6355, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.6597917079925537, + "learning_rate": 2.922420796100731e-06, + "loss": 1.5574, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.7135894298553467, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.6721, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.562045097351074, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.5745, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.6727097034454346, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.6324, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 2.688309907913208, + "learning_rate": 2.91429731925264e-06, + "loss": 1.6698, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.8429853916168213, + "learning_rate": 2.912266450040618e-06, + "loss": 1.6378, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.658046245574951, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.6237, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 2.6601390838623047, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.6163, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.843716025352478, + "eval_runtime": 37.2087, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.949215888977051, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.5883, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 2.823991298675537, + "learning_rate": 2.904142973192527e-06, + "loss": 1.6786, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.6177661418914795, + "learning_rate": 2.902112103980504e-06, + "loss": 1.6337, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.493638038635254, + "learning_rate": 2.900081234768481e-06, + "loss": 1.6328, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.657914161682129, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.6068, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.7800004482269287, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.5963, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.7934350967407227, + "learning_rate": 2.893988627132413e-06, + "loss": 1.611, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.8139140605926514, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.5621, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.5909969806671143, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.6195, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.631871461868286, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.6146, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.846065878868103, + "eval_runtime": 37.2421, + "eval_samples_per_second": 26.851, + "eval_steps_per_second": 1.128, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.667304754257202, + "learning_rate": 2.885865150284322e-06, + "loss": 1.6482, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.5922656059265137, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.6231, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.663452386856079, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.5772, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.6463005542755127, + "learning_rate": 2.879772542648254e-06, + "loss": 1.6333, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.5469250679016113, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.5465, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.7316017150878906, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.6856, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.7127692699432373, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.667, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 2.7006423473358154, + "learning_rate": 2.871649065800163e-06, + "loss": 1.677, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 2.788804054260254, + "learning_rate": 2.86961819658814e-06, + "loss": 1.6092, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.7734181880950928, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.6377, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.8444476127624512, + "eval_runtime": 37.2371, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.6005876064300537, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.6072, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.684293270111084, + "learning_rate": 2.863525588952072e-06, + "loss": 1.6669, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.6160528659820557, + "learning_rate": 2.861494719740049e-06, + "loss": 1.5439, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.68396258354187, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.5909, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.6292917728424072, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.6225, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.702333450317383, + "learning_rate": 2.855402112103981e-06, + "loss": 1.6224, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.850862503051758, + "learning_rate": 2.853371242891958e-06, + "loss": 1.5995, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.861351490020752, + "learning_rate": 2.851340373679935e-06, + "loss": 1.6254, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.7731313705444336, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.6555, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.669189691543579, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.6238, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.845217227935791, + "eval_runtime": 37.2809, + "eval_samples_per_second": 26.823, + "eval_steps_per_second": 1.127, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.64485239982605, + "learning_rate": 2.845247766043867e-06, + "loss": 1.5989, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.7558820247650146, + "learning_rate": 2.843216896831844e-06, + "loss": 1.5615, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.7818500995635986, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.5962, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.571547746658325, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.6434, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 2.8242406845092773, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.6537, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.5545597076416016, + "learning_rate": 2.835093419983753e-06, + "loss": 1.6158, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.7329461574554443, + "learning_rate": 2.833062550771731e-06, + "loss": 1.591, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.5414955615997314, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.5727, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.7625439167022705, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.6343, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.7839064598083496, + "learning_rate": 2.826969943135662e-06, + "loss": 1.6187, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.8442766666412354, + "eval_runtime": 37.2101, + "eval_samples_per_second": 26.874, + "eval_steps_per_second": 1.129, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.615706443786621, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.621, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.6295979022979736, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.5913, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.7428646087646484, + "learning_rate": 2.820877335499594e-06, + "loss": 1.6015, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.5353381633758545, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.5798, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.6209659576416016, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.5783, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.791372299194336, + "learning_rate": 2.814784727863526e-06, + "loss": 1.6082, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.781949520111084, + "learning_rate": 2.812753858651503e-06, + "loss": 1.6337, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 3.0437769889831543, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.5969, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 2.7201249599456787, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.5717, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 2.766634464263916, + "learning_rate": 2.806661251015435e-06, + "loss": 1.6654, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.8452297449111938, + "eval_runtime": 37.2452, + "eval_samples_per_second": 26.849, + "eval_steps_per_second": 1.128, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.726876735687256, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.6652, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 2.775012969970703, + "learning_rate": 2.802599512591389e-06, + "loss": 1.7293, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.7603225708007812, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.6226, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.723904609680176, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.5663, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 2.667384624481201, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.6209, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.7580318450927734, + "learning_rate": 2.794476035743298e-06, + "loss": 1.6705, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.9216315746307373, + "learning_rate": 2.792445166531276e-06, + "loss": 1.6361, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 2.6552748680114746, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.6498, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.764836549758911, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.6009, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.5660064220428467, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.576, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.8392945528030396, + "eval_runtime": 37.2515, + "eval_samples_per_second": 26.845, + "eval_steps_per_second": 1.127, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.5110044479370117, + "learning_rate": 2.784321689683185e-06, + "loss": 1.5744, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.7352850437164307, + "learning_rate": 2.782290820471162e-06, + "loss": 1.5852, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.850947618484497, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.6692, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.7107245922088623, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.6427, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.8236117362976074, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.6375, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.763416051864624, + "learning_rate": 2.774167343623071e-06, + "loss": 1.6864, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 2.7317910194396973, + "learning_rate": 2.772136474411048e-06, + "loss": 1.6015, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.7454333305358887, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.5379, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.6102535724639893, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.6282, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.658440113067627, + "learning_rate": 2.76604386677498e-06, + "loss": 1.6042, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.8391484022140503, + "eval_runtime": 37.1746, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 2.5731043815612793, + "learning_rate": 2.764012997562957e-06, + "loss": 1.6446, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.680983543395996, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.5663, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 2.7525506019592285, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.6628, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.6905839443206787, + "learning_rate": 2.757920389926889e-06, + "loss": 1.6021, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 2.850900888442993, + "learning_rate": 2.755889520714866e-06, + "loss": 1.5984, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.6612954139709473, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.5627, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.632106304168701, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.6304, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.660377025604248, + "learning_rate": 2.749796913078798e-06, + "loss": 1.6622, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.497938394546509, + "learning_rate": 2.747766043866775e-06, + "loss": 1.6152, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.655977725982666, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.6703, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.8361152410507202, + "eval_runtime": 37.2147, + "eval_samples_per_second": 26.871, + "eval_steps_per_second": 1.129, + "step": 1400 + }, + { + "epoch": 1.5353424657534247, + "grad_norm": 2.664874792098999, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.7109, + "step": 1401 + }, + { + "epoch": 1.5364383561643835, + "grad_norm": 2.5773956775665283, + "learning_rate": 2.741673436230707e-06, + "loss": 1.6176, + "step": 1402 + }, + { + "epoch": 1.5375342465753423, + "grad_norm": 2.7094814777374268, + "learning_rate": 2.739642567018684e-06, + "loss": 1.6075, + "step": 1403 + }, + { + "epoch": 1.5386301369863014, + "grad_norm": 2.652325391769409, + "learning_rate": 2.7376116978066615e-06, + "loss": 1.6189, + "step": 1404 + }, + { + "epoch": 1.5397260273972604, + "grad_norm": 2.6892807483673096, + "learning_rate": 2.7355808285946388e-06, + "loss": 1.6122, + "step": 1405 + }, + { + "epoch": 1.5408219178082192, + "grad_norm": 2.6621243953704834, + "learning_rate": 2.733549959382616e-06, + "loss": 1.6008, + "step": 1406 + }, + { + "epoch": 1.541917808219178, + "grad_norm": 2.773747444152832, + "learning_rate": 2.7315190901705933e-06, + "loss": 1.6723, + "step": 1407 + }, + { + "epoch": 1.543013698630137, + "grad_norm": 2.853525400161743, + "learning_rate": 2.7294882209585706e-06, + "loss": 1.641, + "step": 1408 + }, + { + "epoch": 1.5441095890410959, + "grad_norm": 2.6333844661712646, + "learning_rate": 2.727457351746548e-06, + "loss": 1.587, + "step": 1409 + }, + { + "epoch": 1.5452054794520547, + "grad_norm": 2.6299996376037598, + "learning_rate": 2.725426482534525e-06, + "loss": 1.5807, + "step": 1410 + }, + { + "epoch": 1.5452054794520547, + "eval_loss": 1.8377478122711182, + "eval_runtime": 36.8958, + "eval_samples_per_second": 27.103, + "eval_steps_per_second": 1.138, + "step": 1410 + }, + { + "epoch": 1.5463013698630137, + "grad_norm": 2.6222734451293945, + "learning_rate": 2.723395613322502e-06, + "loss": 1.6225, + "step": 1411 + }, + { + "epoch": 1.5473972602739727, + "grad_norm": 2.6702005863189697, + "learning_rate": 2.7213647441104797e-06, + "loss": 1.6031, + "step": 1412 + }, + { + "epoch": 1.5484931506849315, + "grad_norm": 2.6451900005340576, + "learning_rate": 2.7193338748984565e-06, + "loss": 1.5998, + "step": 1413 + }, + { + "epoch": 1.5495890410958904, + "grad_norm": 2.769784927368164, + "learning_rate": 2.7173030056864342e-06, + "loss": 1.7579, + "step": 1414 + }, + { + "epoch": 1.5506849315068494, + "grad_norm": 2.629871368408203, + "learning_rate": 2.715272136474411e-06, + "loss": 1.564, + "step": 1415 + }, + { + "epoch": 1.5517808219178082, + "grad_norm": 2.811937093734741, + "learning_rate": 2.713241267262389e-06, + "loss": 1.6236, + "step": 1416 + }, + { + "epoch": 1.552876712328767, + "grad_norm": 2.883277654647827, + "learning_rate": 2.7112103980503656e-06, + "loss": 1.5996, + "step": 1417 + }, + { + "epoch": 1.553972602739726, + "grad_norm": 2.610964059829712, + "learning_rate": 2.7091795288383433e-06, + "loss": 1.6391, + "step": 1418 + }, + { + "epoch": 1.555068493150685, + "grad_norm": 2.6637415885925293, + "learning_rate": 2.70714865962632e-06, + "loss": 1.5983, + "step": 1419 + }, + { + "epoch": 1.5561643835616439, + "grad_norm": 2.5372607707977295, + "learning_rate": 2.705117790414298e-06, + "loss": 1.6258, + "step": 1420 + }, + { + "epoch": 1.5561643835616439, + "eval_loss": 1.838557243347168, + "eval_runtime": 36.973, + "eval_samples_per_second": 27.047, + "eval_steps_per_second": 1.136, + "step": 1420 + }, + { + "epoch": 1.5572602739726027, + "grad_norm": 2.900594711303711, + "learning_rate": 2.7030869212022747e-06, + "loss": 1.67, + "step": 1421 + }, + { + "epoch": 1.5583561643835617, + "grad_norm": 2.4924063682556152, + "learning_rate": 2.7010560519902516e-06, + "loss": 1.5879, + "step": 1422 + }, + { + "epoch": 1.5594520547945205, + "grad_norm": 2.5761115550994873, + "learning_rate": 2.6990251827782293e-06, + "loss": 1.6287, + "step": 1423 + }, + { + "epoch": 1.5605479452054793, + "grad_norm": 2.645463466644287, + "learning_rate": 2.696994313566206e-06, + "loss": 1.5925, + "step": 1424 + }, + { + "epoch": 1.5616438356164384, + "grad_norm": 2.682466506958008, + "learning_rate": 2.694963444354184e-06, + "loss": 1.6723, + "step": 1425 + }, + { + "epoch": 1.5627397260273974, + "grad_norm": 2.499749183654785, + "learning_rate": 2.6929325751421607e-06, + "loss": 1.5998, + "step": 1426 + }, + { + "epoch": 1.563835616438356, + "grad_norm": 2.659956455230713, + "learning_rate": 2.6909017059301384e-06, + "loss": 1.6311, + "step": 1427 + }, + { + "epoch": 1.564931506849315, + "grad_norm": 2.6684396266937256, + "learning_rate": 2.6888708367181153e-06, + "loss": 1.5648, + "step": 1428 + }, + { + "epoch": 1.566027397260274, + "grad_norm": 2.6868791580200195, + "learning_rate": 2.686839967506093e-06, + "loss": 1.6419, + "step": 1429 + }, + { + "epoch": 1.5671232876712329, + "grad_norm": 2.6999104022979736, + "learning_rate": 2.68480909829407e-06, + "loss": 1.55, + "step": 1430 + }, + { + "epoch": 1.5671232876712329, + "eval_loss": 1.83835768699646, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1430 + }, + { + "epoch": 1.5682191780821917, + "grad_norm": 2.7338085174560547, + "learning_rate": 2.6827782290820475e-06, + "loss": 1.5507, + "step": 1431 + }, + { + "epoch": 1.5693150684931507, + "grad_norm": 2.7358195781707764, + "learning_rate": 2.6807473598700244e-06, + "loss": 1.6659, + "step": 1432 + }, + { + "epoch": 1.5704109589041098, + "grad_norm": 2.671983242034912, + "learning_rate": 2.678716490658002e-06, + "loss": 1.6293, + "step": 1433 + }, + { + "epoch": 1.5715068493150683, + "grad_norm": 2.551203489303589, + "learning_rate": 2.676685621445979e-06, + "loss": 1.623, + "step": 1434 + }, + { + "epoch": 1.5726027397260274, + "grad_norm": 2.6485984325408936, + "learning_rate": 2.674654752233956e-06, + "loss": 1.6127, + "step": 1435 + }, + { + "epoch": 1.5736986301369864, + "grad_norm": 2.643105983734131, + "learning_rate": 2.6726238830219335e-06, + "loss": 1.6155, + "step": 1436 + }, + { + "epoch": 1.5747945205479452, + "grad_norm": 2.769458055496216, + "learning_rate": 2.6705930138099107e-06, + "loss": 1.6576, + "step": 1437 + }, + { + "epoch": 1.575890410958904, + "grad_norm": 2.3996448516845703, + "learning_rate": 2.668562144597888e-06, + "loss": 1.5276, + "step": 1438 + }, + { + "epoch": 1.576986301369863, + "grad_norm": 2.5827524662017822, + "learning_rate": 2.6665312753858653e-06, + "loss": 1.6321, + "step": 1439 + }, + { + "epoch": 1.5780821917808219, + "grad_norm": 2.728794813156128, + "learning_rate": 2.6645004061738426e-06, + "loss": 1.6177, + "step": 1440 + }, + { + "epoch": 1.5780821917808219, + "eval_loss": 1.8374123573303223, + "eval_runtime": 37.1748, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1440 + }, + { + "epoch": 1.5791780821917807, + "grad_norm": 2.7500123977661133, + "learning_rate": 2.66246953696182e-06, + "loss": 1.5679, + "step": 1441 + }, + { + "epoch": 1.5802739726027397, + "grad_norm": 2.6778831481933594, + "learning_rate": 2.660438667749797e-06, + "loss": 1.5744, + "step": 1442 + }, + { + "epoch": 1.5813698630136988, + "grad_norm": 2.933138847351074, + "learning_rate": 2.6584077985377744e-06, + "loss": 1.6238, + "step": 1443 + }, + { + "epoch": 1.5824657534246576, + "grad_norm": 2.642246961593628, + "learning_rate": 2.6563769293257517e-06, + "loss": 1.6186, + "step": 1444 + }, + { + "epoch": 1.5835616438356164, + "grad_norm": 2.5519111156463623, + "learning_rate": 2.654346060113729e-06, + "loss": 1.5798, + "step": 1445 + }, + { + "epoch": 1.5846575342465754, + "grad_norm": 2.569063186645508, + "learning_rate": 2.6523151909017062e-06, + "loss": 1.5881, + "step": 1446 + }, + { + "epoch": 1.5857534246575342, + "grad_norm": 2.501680850982666, + "learning_rate": 2.6502843216896835e-06, + "loss": 1.5768, + "step": 1447 + }, + { + "epoch": 1.586849315068493, + "grad_norm": 2.7483339309692383, + "learning_rate": 2.6482534524776603e-06, + "loss": 1.6889, + "step": 1448 + }, + { + "epoch": 1.587945205479452, + "grad_norm": 3.004084587097168, + "learning_rate": 2.646222583265638e-06, + "loss": 1.6565, + "step": 1449 + }, + { + "epoch": 1.589041095890411, + "grad_norm": 2.793058156967163, + "learning_rate": 2.644191714053615e-06, + "loss": 1.6246, + "step": 1450 + }, + { + "epoch": 1.589041095890411, + "eval_loss": 1.83695387840271, + "eval_runtime": 37.1748, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1450 + }, + { + "epoch": 1.59013698630137, + "grad_norm": 2.803190231323242, + "learning_rate": 2.6421608448415926e-06, + "loss": 1.6051, + "step": 1451 + }, + { + "epoch": 1.5912328767123287, + "grad_norm": 2.7721307277679443, + "learning_rate": 2.6401299756295694e-06, + "loss": 1.5712, + "step": 1452 + }, + { + "epoch": 1.5923287671232877, + "grad_norm": 2.9235639572143555, + "learning_rate": 2.638099106417547e-06, + "loss": 1.6197, + "step": 1453 + }, + { + "epoch": 1.5934246575342466, + "grad_norm": 2.609719753265381, + "learning_rate": 2.636068237205524e-06, + "loss": 1.5976, + "step": 1454 + }, + { + "epoch": 1.5945205479452054, + "grad_norm": 2.7927262783050537, + "learning_rate": 2.6340373679935017e-06, + "loss": 1.6844, + "step": 1455 + }, + { + "epoch": 1.5956164383561644, + "grad_norm": 2.8688440322875977, + "learning_rate": 2.6320064987814785e-06, + "loss": 1.5285, + "step": 1456 + }, + { + "epoch": 1.5967123287671234, + "grad_norm": 2.819079637527466, + "learning_rate": 2.6299756295694562e-06, + "loss": 1.6171, + "step": 1457 + }, + { + "epoch": 1.5978082191780822, + "grad_norm": 2.825784206390381, + "learning_rate": 2.627944760357433e-06, + "loss": 1.6656, + "step": 1458 + }, + { + "epoch": 1.598904109589041, + "grad_norm": 2.6807522773742676, + "learning_rate": 2.6259138911454104e-06, + "loss": 1.6839, + "step": 1459 + }, + { + "epoch": 1.6, + "grad_norm": 2.9442570209503174, + "learning_rate": 2.6238830219333877e-06, + "loss": 1.6547, + "step": 1460 + }, + { + "epoch": 1.6, + "eval_loss": 1.8358429670333862, + "eval_runtime": 37.1499, + "eval_samples_per_second": 26.918, + "eval_steps_per_second": 1.131, + "step": 1460 + }, + { + "epoch": 1.601095890410959, + "grad_norm": 3.0025434494018555, + "learning_rate": 2.621852152721365e-06, + "loss": 1.5688, + "step": 1461 + }, + { + "epoch": 1.6021917808219177, + "grad_norm": 3.0999062061309814, + "learning_rate": 2.619821283509342e-06, + "loss": 1.639, + "step": 1462 + }, + { + "epoch": 1.6032876712328767, + "grad_norm": 2.7106330394744873, + "learning_rate": 2.6177904142973195e-06, + "loss": 1.5682, + "step": 1463 + }, + { + "epoch": 1.6043835616438358, + "grad_norm": 2.968259811401367, + "learning_rate": 2.6157595450852968e-06, + "loss": 1.6421, + "step": 1464 + }, + { + "epoch": 1.6054794520547944, + "grad_norm": 2.964728593826294, + "learning_rate": 2.613728675873274e-06, + "loss": 1.6626, + "step": 1465 + }, + { + "epoch": 1.6065753424657534, + "grad_norm": 2.8934402465820312, + "learning_rate": 2.6116978066612513e-06, + "loss": 1.6742, + "step": 1466 + }, + { + "epoch": 1.6076712328767124, + "grad_norm": 2.7678956985473633, + "learning_rate": 2.6096669374492286e-06, + "loss": 1.6139, + "step": 1467 + }, + { + "epoch": 1.6087671232876712, + "grad_norm": 2.74017333984375, + "learning_rate": 2.607636068237206e-06, + "loss": 1.5726, + "step": 1468 + }, + { + "epoch": 1.60986301369863, + "grad_norm": 2.8686470985412598, + "learning_rate": 2.605605199025183e-06, + "loss": 1.638, + "step": 1469 + }, + { + "epoch": 1.610958904109589, + "grad_norm": 3.008751392364502, + "learning_rate": 2.6035743298131604e-06, + "loss": 1.6246, + "step": 1470 + }, + { + "epoch": 1.610958904109589, + "eval_loss": 1.8375154733657837, + "eval_runtime": 37.1808, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 1470 + }, + { + "epoch": 1.612054794520548, + "grad_norm": 3.030925989151001, + "learning_rate": 2.6015434606011377e-06, + "loss": 1.6558, + "step": 1471 + }, + { + "epoch": 1.6131506849315067, + "grad_norm": 3.1855311393737793, + "learning_rate": 2.5995125913891145e-06, + "loss": 1.6488, + "step": 1472 + }, + { + "epoch": 1.6142465753424657, + "grad_norm": 2.943204879760742, + "learning_rate": 2.5974817221770922e-06, + "loss": 1.6211, + "step": 1473 + }, + { + "epoch": 1.6153424657534248, + "grad_norm": 2.681720733642578, + "learning_rate": 2.595450852965069e-06, + "loss": 1.655, + "step": 1474 + }, + { + "epoch": 1.6164383561643836, + "grad_norm": 2.750523567199707, + "learning_rate": 2.593419983753047e-06, + "loss": 1.6624, + "step": 1475 + }, + { + "epoch": 1.6175342465753424, + "grad_norm": 2.8923532962799072, + "learning_rate": 2.5913891145410236e-06, + "loss": 1.6046, + "step": 1476 + }, + { + "epoch": 1.6186301369863014, + "grad_norm": 2.852468490600586, + "learning_rate": 2.5893582453290013e-06, + "loss": 1.6557, + "step": 1477 + }, + { + "epoch": 1.6197260273972602, + "grad_norm": 2.7146172523498535, + "learning_rate": 2.587327376116978e-06, + "loss": 1.5888, + "step": 1478 + }, + { + "epoch": 1.620821917808219, + "grad_norm": 2.7109804153442383, + "learning_rate": 2.585296506904956e-06, + "loss": 1.6049, + "step": 1479 + }, + { + "epoch": 1.621917808219178, + "grad_norm": 2.9429404735565186, + "learning_rate": 2.5832656376929327e-06, + "loss": 1.6046, + "step": 1480 + }, + { + "epoch": 1.621917808219178, + "eval_loss": 1.837095856666565, + "eval_runtime": 37.19, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 1480 + }, + { + "epoch": 1.6230136986301371, + "grad_norm": 2.89047908782959, + "learning_rate": 2.5812347684809104e-06, + "loss": 1.5305, + "step": 1481 + }, + { + "epoch": 1.624109589041096, + "grad_norm": 3.0587005615234375, + "learning_rate": 2.5792038992688873e-06, + "loss": 1.7029, + "step": 1482 + }, + { + "epoch": 1.6252054794520547, + "grad_norm": 2.7209622859954834, + "learning_rate": 2.577173030056865e-06, + "loss": 1.5424, + "step": 1483 + }, + { + "epoch": 1.6263013698630138, + "grad_norm": 2.8400819301605225, + "learning_rate": 2.575142160844842e-06, + "loss": 1.6411, + "step": 1484 + }, + { + "epoch": 1.6273972602739726, + "grad_norm": 2.7427124977111816, + "learning_rate": 2.5731112916328187e-06, + "loss": 1.6421, + "step": 1485 + }, + { + "epoch": 1.6284931506849314, + "grad_norm": 2.6996700763702393, + "learning_rate": 2.5710804224207964e-06, + "loss": 1.5845, + "step": 1486 + }, + { + "epoch": 1.6295890410958904, + "grad_norm": 2.540188789367676, + "learning_rate": 2.5690495532087733e-06, + "loss": 1.5472, + "step": 1487 + }, + { + "epoch": 1.6306849315068495, + "grad_norm": 2.7682793140411377, + "learning_rate": 2.567018683996751e-06, + "loss": 1.6347, + "step": 1488 + }, + { + "epoch": 1.6317808219178083, + "grad_norm": 2.760413646697998, + "learning_rate": 2.564987814784728e-06, + "loss": 1.5826, + "step": 1489 + }, + { + "epoch": 1.632876712328767, + "grad_norm": 2.9961049556732178, + "learning_rate": 2.5629569455727055e-06, + "loss": 1.6037, + "step": 1490 + }, + { + "epoch": 1.632876712328767, + "eval_loss": 1.8322792053222656, + "eval_runtime": 37.2227, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 1490 + }, + { + "epoch": 1.633972602739726, + "grad_norm": 2.6983907222747803, + "learning_rate": 2.5609260763606824e-06, + "loss": 1.6142, + "step": 1491 + }, + { + "epoch": 1.635068493150685, + "grad_norm": 2.867401599884033, + "learning_rate": 2.55889520714866e-06, + "loss": 1.7227, + "step": 1492 + }, + { + "epoch": 1.6361643835616437, + "grad_norm": 2.7735934257507324, + "learning_rate": 2.556864337936637e-06, + "loss": 1.5856, + "step": 1493 + }, + { + "epoch": 1.6372602739726028, + "grad_norm": 2.7661404609680176, + "learning_rate": 2.5548334687246146e-06, + "loss": 1.6084, + "step": 1494 + }, + { + "epoch": 1.6383561643835618, + "grad_norm": 2.6881582736968994, + "learning_rate": 2.5528025995125915e-06, + "loss": 1.659, + "step": 1495 + }, + { + "epoch": 1.6394520547945204, + "grad_norm": 2.5745224952697754, + "learning_rate": 2.550771730300569e-06, + "loss": 1.6131, + "step": 1496 + }, + { + "epoch": 1.6405479452054794, + "grad_norm": 2.736159563064575, + "learning_rate": 2.548740861088546e-06, + "loss": 1.6469, + "step": 1497 + }, + { + "epoch": 1.6416438356164385, + "grad_norm": 2.6255311965942383, + "learning_rate": 2.5467099918765233e-06, + "loss": 1.5452, + "step": 1498 + }, + { + "epoch": 1.6427397260273973, + "grad_norm": 2.6919801235198975, + "learning_rate": 2.5446791226645006e-06, + "loss": 1.5282, + "step": 1499 + }, + { + "epoch": 1.643835616438356, + "grad_norm": 2.641533136367798, + "learning_rate": 2.542648253452478e-06, + "loss": 1.6524, + "step": 1500 + }, + { + "epoch": 1.643835616438356, + "eval_loss": 1.8287466764450073, + "eval_runtime": 37.2305, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1500 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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": 4.67596411404288e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1500/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/latest new file mode 100644 index 0000000000000000000000000000000000000000..10b09fac99bc80ff931649e8b3378aab683b28be --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/latest @@ -0,0 +1 @@ +global_step1600 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..b809eb215958233fedf5bed6b96a6e7e9b9f2318 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/trainer_state.json @@ -0,0 +1,12513 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.7534246575342465, + "eval_steps": 10, + "global_step": 1600, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6008102893829346, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7848, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.630127429962158, + "learning_rate": 4.362307067424858e-06, + "loss": 1.7135, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.720945358276367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7897, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.898836851119995, + "learning_rate": 4.358245329000813e-06, + "loss": 1.7434, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.6548235416412354, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7683, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.7218399047851562, + "learning_rate": 4.354183590576767e-06, + "loss": 1.7729, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.5972483158111572, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7517, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.6741726398468018, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7354, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.898418426513672, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7843, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.3964333534240723, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.808, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.9525212049484253, + "eval_runtime": 36.8297, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.9952592849731445, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7856, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.81740140914917, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7613, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6271541118621826, + "learning_rate": 4.339967506092607e-06, + "loss": 1.8409, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.540395498275757, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.7765, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.629131317138672, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7634, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.5796725749969482, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7414, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6980209350585938, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.8048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.658919095993042, + "learning_rate": 4.329813160032495e-06, + "loss": 1.7725, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5237927436828613, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7559, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.5822298526763916, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7129, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9492865800857544, + "eval_runtime": 37.0012, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.762810468673706, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7895, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.48427414894104, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7667, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.790052890777588, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7548, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.927957773208618, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.7489, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.4964418411254883, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7947, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.4845147132873535, + "learning_rate": 4.313566206336312e-06, + "loss": 1.7886, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.687361717224121, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7968, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.568202018737793, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7332, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.496429443359375, + "learning_rate": 4.307473598700244e-06, + "loss": 1.7583, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6063425540924072, + "learning_rate": 4.305442729488221e-06, + "loss": 1.779, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.9421359300613403, + "eval_runtime": 37.0712, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.5371274948120117, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7133, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.5601038932800293, + "learning_rate": 4.301380991064176e-06, + "loss": 1.8089, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.6244118213653564, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7983, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.4798192977905273, + "learning_rate": 4.29731925264013e-06, + "loss": 1.8221, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.483376979827881, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.8032, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.716745615005493, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7583, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.6685945987701416, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7365, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.465810537338257, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7682, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.7629129886627197, + "learning_rate": 4.287164906580017e-06, + "loss": 1.809, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5273563861846924, + "learning_rate": 4.285134037367994e-06, + "loss": 1.7464, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.9359965324401855, + "eval_runtime": 37.1317, + "eval_samples_per_second": 26.931, + "eval_steps_per_second": 1.131, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.514230966567993, + "learning_rate": 4.283103168155971e-06, + "loss": 1.7365, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.704042911529541, + "learning_rate": 4.281072298943948e-06, + "loss": 1.7268, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.4716262817382812, + "learning_rate": 4.279041429731926e-06, + "loss": 1.7212, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5695507526397705, + "learning_rate": 4.277010560519903e-06, + "loss": 1.7398, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.762845039367676, + "learning_rate": 4.27497969130788e-06, + "loss": 1.7158, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.571054220199585, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7685, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.908184289932251, + "learning_rate": 4.270917952883835e-06, + "loss": 1.7865, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.705479621887207, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7922, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.4418110847473145, + "learning_rate": 4.266856214459789e-06, + "loss": 1.7312, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.4619884490966797, + "learning_rate": 4.264825345247766e-06, + "loss": 1.8343, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.9352803230285645, + "eval_runtime": 37.0958, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5386202335357666, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6902, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9012045860290527, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7552, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.8028385639190674, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.6872, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.4505386352539062, + "learning_rate": 4.256701868399675e-06, + "loss": 1.75, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7551822662353516, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7612, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.596270799636841, + "learning_rate": 4.252640129975629e-06, + "loss": 1.7694, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.387662172317505, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.7776, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.7208738327026367, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7375, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.5317776203155518, + "learning_rate": 4.246547522339562e-06, + "loss": 1.6922, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.4803197383880615, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.741, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.9322715997695923, + "eval_runtime": 37.0962, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.5688276290893555, + "learning_rate": 4.242485783915517e-06, + "loss": 1.7518, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.4652810096740723, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6853, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.541903257369995, + "learning_rate": 4.23842404549147e-06, + "loss": 1.6841, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.377857208251953, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7312, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.418323040008545, + "learning_rate": 4.234362307067425e-06, + "loss": 1.6892, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5374410152435303, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7719, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7119228839874268, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.7819, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.5333240032196045, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7758, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.528696060180664, + "learning_rate": 4.226238830219334e-06, + "loss": 1.75, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.539039134979248, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7498, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.934200644493103, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.5384178161621094, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7615, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.601658821105957, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7645, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5359764099121094, + "learning_rate": 4.218115353371243e-06, + "loss": 1.7549, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.641118049621582, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7487, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.4146382808685303, + "learning_rate": 4.214053614947198e-06, + "loss": 1.7102, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4245245456695557, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7842, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.509824752807617, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6876, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.636444330215454, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7681, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.476858615875244, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7049, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.75732159614563, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7833, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.931411623954773, + "eval_runtime": 37.0994, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.5437941551208496, + "learning_rate": 4.201868399675061e-06, + "loss": 1.7751, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.5010673999786377, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7715, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.6352853775024414, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6798, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.588247776031494, + "learning_rate": 4.195775792038993e-06, + "loss": 1.7575, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.349921226501465, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7245, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.394432783126831, + "learning_rate": 4.191714053614948e-06, + "loss": 1.6778, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.883906602859497, + "learning_rate": 4.189683184402924e-06, + "loss": 1.7701, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.537799596786499, + "learning_rate": 4.187652315190902e-06, + "loss": 1.7286, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.575700283050537, + "learning_rate": 4.185621445978879e-06, + "loss": 1.7812, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.592252492904663, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7176, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.9259952306747437, + "eval_runtime": 37.144, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.511845111846924, + "learning_rate": 4.181559707554833e-06, + "loss": 1.7564, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.575446367263794, + "learning_rate": 4.179528838342811e-06, + "loss": 1.7943, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.379391670227051, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7098, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.701812982559204, + "learning_rate": 4.175467099918766e-06, + "loss": 1.7292, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.486640453338623, + "learning_rate": 4.173436230706742e-06, + "loss": 1.7963, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.561983823776245, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.7153, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.5709176063537598, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7412, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.3308839797973633, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7188, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.592470169067383, + "learning_rate": 4.165312753858652e-06, + "loss": 1.7599, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.3747010231018066, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7361, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.9229093790054321, + "eval_runtime": 37.1417, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.4062509536743164, + "learning_rate": 4.161251015434607e-06, + "loss": 1.7256, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.4951181411743164, + "learning_rate": 4.159220146222583e-06, + "loss": 1.7622, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.3453004360198975, + "learning_rate": 4.157189277010561e-06, + "loss": 1.6969, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.382671594619751, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7109, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4277336597442627, + "learning_rate": 4.153127538586516e-06, + "loss": 1.7471, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.4476678371429443, + "learning_rate": 4.151096669374492e-06, + "loss": 1.7483, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.451099395751953, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7366, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.618217706680298, + "learning_rate": 4.147034930950447e-06, + "loss": 1.7282, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.414400100708008, + "learning_rate": 4.145004061738425e-06, + "loss": 1.7409, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.535353422164917, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.7346, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.9240931272506714, + "eval_runtime": 36.844, + "eval_samples_per_second": 27.141, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.5075416564941406, + "learning_rate": 4.140942323314379e-06, + "loss": 1.6668, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.4136979579925537, + "learning_rate": 4.138911454102356e-06, + "loss": 1.696, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.4363577365875244, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7174, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.4578073024749756, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7621, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.370976686477661, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7443, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.363755702972412, + "learning_rate": 4.130787977254265e-06, + "loss": 1.7399, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.44704532623291, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7481, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.5463340282440186, + "learning_rate": 4.12672623883022e-06, + "loss": 1.7646, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7088072299957275, + "learning_rate": 4.124695369618197e-06, + "loss": 1.8108, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.463271141052246, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7311, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.918715238571167, + "eval_runtime": 37.0468, + "eval_samples_per_second": 26.993, + "eval_steps_per_second": 1.134, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6165266036987305, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.7295, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.3432133197784424, + "learning_rate": 4.118602761982129e-06, + "loss": 1.7512, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.552751302719116, + "learning_rate": 4.116571892770106e-06, + "loss": 1.6976, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7586560249328613, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6708, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.5326449871063232, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7824, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.7658956050872803, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7093, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.6868369579315186, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7475, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.543748140335083, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.7714, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.7235589027404785, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7593, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.697038173675537, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6296, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.91779363155365, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 2.6746180057525635, + "learning_rate": 4.100324939073924e-06, + "loss": 1.7971, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.6281094551086426, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7212, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.446025848388672, + "learning_rate": 4.096263200649879e-06, + "loss": 1.7324, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.5460824966430664, + "learning_rate": 4.094232331437856e-06, + "loss": 1.7435, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.5419771671295166, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6979, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.5545520782470703, + "learning_rate": 4.090170593013811e-06, + "loss": 1.7042, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.8353254795074463, + "learning_rate": 4.088139723801787e-06, + "loss": 1.7313, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.6521668434143066, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6901, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.6611177921295166, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7064, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.411482572555542, + "learning_rate": 4.08204711616572e-06, + "loss": 1.729, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.9176616668701172, + "eval_runtime": 37.1603, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.689110279083252, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7707, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.544018268585205, + "learning_rate": 4.077985377741674e-06, + "loss": 1.6724, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.55307936668396, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7677, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.3896875381469727, + "learning_rate": 4.073923639317629e-06, + "loss": 1.7729, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.381887197494507, + "learning_rate": 4.071892770105605e-06, + "loss": 1.7737, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.4873931407928467, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6928, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.4908432960510254, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6674, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.4346399307250977, + "learning_rate": 4.065800162469538e-06, + "loss": 1.6947, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6459081172943115, + "learning_rate": 4.063769293257514e-06, + "loss": 1.701, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.6092610359191895, + "learning_rate": 4.061738424045492e-06, + "loss": 1.7425, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.919799566268921, + "eval_runtime": 37.1632, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.4443421363830566, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7267, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.8252694606781006, + "learning_rate": 4.057676685621446e-06, + "loss": 1.7412, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.3391194343566895, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.7916, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.57979679107666, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7438, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.398383855819702, + "learning_rate": 4.051584077985378e-06, + "loss": 1.6606, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.394498348236084, + "learning_rate": 4.049553208773355e-06, + "loss": 1.736, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5070464611053467, + "learning_rate": 4.047522339561333e-06, + "loss": 1.7559, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.439722776412964, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7133, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.569152593612671, + "learning_rate": 4.043460601137287e-06, + "loss": 1.7123, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.637657642364502, + "learning_rate": 4.041429731925264e-06, + "loss": 1.8267, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.9148659706115723, + "eval_runtime": 37.2248, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.441269636154175, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7106, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.582172155380249, + "learning_rate": 4.037367993501219e-06, + "loss": 1.7167, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.5083518028259277, + "learning_rate": 4.035337124289196e-06, + "loss": 1.8031, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.504112958908081, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.694, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.5068695545196533, + "learning_rate": 4.031275385865151e-06, + "loss": 1.6798, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.600541114807129, + "learning_rate": 4.029244516653128e-06, + "loss": 1.7346, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.503020763397217, + "learning_rate": 4.027213647441105e-06, + "loss": 1.7507, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.7774016857147217, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6737, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.5771589279174805, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6943, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.3841609954833984, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7142, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.9146945476531982, + "eval_runtime": 37.1701, + "eval_samples_per_second": 26.903, + "eval_steps_per_second": 1.13, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.466207981109619, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.7192, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.72037935256958, + "learning_rate": 4.017059301380992e-06, + "loss": 1.8082, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.36811900138855, + "learning_rate": 4.015028432168969e-06, + "loss": 1.6964, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.4803504943847656, + "learning_rate": 4.012997562956945e-06, + "loss": 1.779, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.707174777984619, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6715, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.462543249130249, + "learning_rate": 4.0089358245329e-06, + "loss": 1.7117, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.481444835662842, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7007, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.4432222843170166, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.703, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.611387014389038, + "learning_rate": 4.002843216896833e-06, + "loss": 1.7722, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.456305742263794, + "learning_rate": 4.000812347684809e-06, + "loss": 1.7615, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.9089828729629517, + "eval_runtime": 37.1812, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4287772178649902, + "learning_rate": 3.998781478472787e-06, + "loss": 1.7098, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5659544467926025, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7143, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.467067003250122, + "learning_rate": 3.994719740048742e-06, + "loss": 1.6668, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.5660688877105713, + "learning_rate": 3.992688870836718e-06, + "loss": 1.8185, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.6185719966888428, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7356, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.5389695167541504, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6973, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.499234914779663, + "learning_rate": 3.98659626320065e-06, + "loss": 1.7539, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.6246273517608643, + "learning_rate": 3.984565393988627e-06, + "loss": 1.7055, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6127917766571045, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7177, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.4941787719726562, + "learning_rate": 3.980503655564582e-06, + "loss": 1.728, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.9063301086425781, + "eval_runtime": 37.3284, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.4694740772247314, + "learning_rate": 3.978472786352559e-06, + "loss": 1.7384, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.8288261890411377, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7409, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.518359661102295, + "learning_rate": 3.974411047928514e-06, + "loss": 1.714, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4874095916748047, + "learning_rate": 3.972380178716491e-06, + "loss": 1.781, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.523772716522217, + "learning_rate": 3.970349309504468e-06, + "loss": 1.7474, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.406132459640503, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.7486, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.388207197189331, + "learning_rate": 3.966287571080423e-06, + "loss": 1.67, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.5564754009246826, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6803, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.531890869140625, + "learning_rate": 3.962225832656377e-06, + "loss": 1.7368, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.404284954071045, + "learning_rate": 3.960194963444355e-06, + "loss": 1.7247, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.9008547067642212, + "eval_runtime": 37.1962, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.3903231620788574, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6465, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7126, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7174, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.6466639041900635, + "learning_rate": 3.954102355808286e-06, + "loss": 1.7166, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.530284881591797, + "learning_rate": 3.952071486596264e-06, + "loss": 1.7155, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.601516008377075, + "learning_rate": 3.950040617384241e-06, + "loss": 1.7157, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.6146240234375, + "learning_rate": 3.948009748172218e-06, + "loss": 1.6931, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.4904634952545166, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.7594, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.345689296722412, + "learning_rate": 3.943948009748173e-06, + "loss": 1.7175, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4469082355499268, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7161, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.9021154642105103, + "eval_runtime": 36.8772, + "eval_samples_per_second": 27.117, + "eval_steps_per_second": 1.139, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.473752737045288, + "learning_rate": 3.939886271324127e-06, + "loss": 1.7688, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.461033582687378, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6752, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.419179916381836, + "learning_rate": 3.935824532900082e-06, + "loss": 1.6685, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.543102502822876, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7348, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.4932174682617188, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.7572, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.4563698768615723, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6691, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.566558837890625, + "learning_rate": 3.927701056051991e-06, + "loss": 1.697, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.5550761222839355, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.7194, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.51362681388855, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6804, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.448714017868042, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6575, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.904768943786621, + "eval_runtime": 37.0644, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.582794666290283, + "learning_rate": 3.9195775792039e-06, + "loss": 1.717, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.613327980041504, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.6927, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.522038221359253, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7391, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.372134208679199, + "learning_rate": 3.913484971567831e-06, + "loss": 1.7448, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.246591329574585, + "learning_rate": 3.911454102355808e-06, + "loss": 1.6531, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.48276948928833, + "learning_rate": 3.909423233143786e-06, + "loss": 1.7268, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.7920989990234375, + "learning_rate": 3.907392363931763e-06, + "loss": 1.7155, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.5430562496185303, + "learning_rate": 3.90536149471974e-06, + "loss": 1.7866, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.5819969177246094, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.8024, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.726024866104126, + "learning_rate": 3.901299756295695e-06, + "loss": 1.8467, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.899052381515503, + "eval_runtime": 37.192, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.6913323402404785, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6976, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.5307703018188477, + "learning_rate": 3.897238017871649e-06, + "loss": 1.7072, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.4976935386657715, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.7183, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.435443878173828, + "learning_rate": 3.893176279447604e-06, + "loss": 1.6845, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.704554319381714, + "learning_rate": 3.891145410235581e-06, + "loss": 1.78, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.4770138263702393, + "learning_rate": 3.889114541023558e-06, + "loss": 1.7436, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.472780227661133, + "learning_rate": 3.887083671811536e-06, + "loss": 1.7007, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.4693336486816406, + "learning_rate": 3.885052802599513e-06, + "loss": 1.6487, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.458834648132324, + "learning_rate": 3.88302193338749e-06, + "loss": 1.7895, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.4430108070373535, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.72, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.9009006023406982, + "eval_runtime": 37.1906, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.6104867458343506, + "learning_rate": 3.878960194963445e-06, + "loss": 1.6938, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.5191142559051514, + "learning_rate": 3.876929325751422e-06, + "loss": 1.7528, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.3788414001464844, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6446, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.485631227493286, + "learning_rate": 3.872867587327377e-06, + "loss": 1.7087, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.3721091747283936, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6752, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4648489952087402, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6883, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.447054862976074, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.7109, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.70007061958313, + "learning_rate": 3.864744110479286e-06, + "loss": 1.7342, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.6336557865142822, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7308, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.807671070098877, + "learning_rate": 3.86068237205524e-06, + "loss": 1.7519, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8990603685379028, + "eval_runtime": 37.1643, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.7177038192749023, + "learning_rate": 3.858651502843217e-06, + "loss": 1.7065, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.570364236831665, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6529, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.590329170227051, + "learning_rate": 3.854589764419171e-06, + "loss": 1.6955, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.542722702026367, + "learning_rate": 3.852558895207149e-06, + "loss": 1.7803, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.6095776557922363, + "learning_rate": 3.850528025995126e-06, + "loss": 1.7267, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.550161838531494, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6767, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.5079615116119385, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6897, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.5113067626953125, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.7024, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.559811592102051, + "learning_rate": 3.842404549147035e-06, + "loss": 1.695, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.507901191711426, + "learning_rate": 3.840373679935013e-06, + "loss": 1.661, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8952044248580933, + "eval_runtime": 37.2303, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.288231134414673, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.7095, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.260040044784546, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6858, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.5524179935455322, + "learning_rate": 3.834281072298944e-06, + "loss": 1.7362, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.419013738632202, + "learning_rate": 3.832250203086921e-06, + "loss": 1.6803, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.5345773696899414, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7435, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.5771822929382324, + "learning_rate": 3.828188464662876e-06, + "loss": 1.7932, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.4880053997039795, + "learning_rate": 3.826157595450853e-06, + "loss": 1.711, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.596400022506714, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6569, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.620776891708374, + "learning_rate": 3.822095857026808e-06, + "loss": 1.7159, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.512397527694702, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6832, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8950546979904175, + "eval_runtime": 37.1526, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5925326347351074, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6977, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.3980817794799805, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.6997, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.4825663566589355, + "learning_rate": 3.813972380178717e-06, + "loss": 1.7063, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.535006284713745, + "learning_rate": 3.811941510966694e-06, + "loss": 1.7457, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.450657606124878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.6521, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.3942458629608154, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6194, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5781140327453613, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.7105, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.5196564197540283, + "learning_rate": 3.803818034118603e-06, + "loss": 1.6741, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4819812774658203, + "learning_rate": 3.80178716490658e-06, + "loss": 1.7559, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.6393206119537354, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.7421, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.8945845365524292, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5286684036254883, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6933, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.690138339996338, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.6804, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.7789907455444336, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7839, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.5196211338043213, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7564, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.550898790359497, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.758, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.617361545562744, + "learning_rate": 3.787571080422421e-06, + "loss": 1.7497, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.533339500427246, + "learning_rate": 3.785540211210398e-06, + "loss": 1.7336, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.475037097930908, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.712, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.485506534576416, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.6985, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.712912082672119, + "learning_rate": 3.77944760357433e-06, + "loss": 1.812, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8921257257461548, + "eval_runtime": 37.1309, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.6280291080474854, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.7504, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6162853240966797, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6793, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.665316581726074, + "learning_rate": 3.773354995938262e-06, + "loss": 1.7333, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.531297445297241, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7298, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.543412446975708, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.7677, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.3839635848999023, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.7039, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.3417305946350098, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6206, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.4434802532196045, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.713, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.5223472118377686, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6587, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.577479124069214, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.7032, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8891643285751343, + "eval_runtime": 37.1123, + "eval_samples_per_second": 26.945, + "eval_steps_per_second": 1.132, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.468871593475342, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.732, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.5907723903656006, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.6866, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.4454426765441895, + "learning_rate": 3.753046303818034e-06, + "loss": 1.746, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.4651715755462646, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7375, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6342053413391113, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.7693, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.454977512359619, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6796, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.440626859664917, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7075, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.4905781745910645, + "learning_rate": 3.742891957757921e-06, + "loss": 1.7422, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6843647956848145, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7107, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.5906362533569336, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.7545, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8889163732528687, + "eval_runtime": 36.8136, + "eval_samples_per_second": 27.164, + "eval_steps_per_second": 1.141, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.5185422897338867, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6857, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.4425783157348633, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.687, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.5200576782226562, + "learning_rate": 3.732737611697807e-06, + "loss": 1.643, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.4664227962493896, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.5965, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.495962381362915, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6663, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.5607051849365234, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.6388, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 2.9295718669891357, + "learning_rate": 3.724614134849716e-06, + "loss": 1.7541, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.5632426738739014, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6359, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.6059279441833496, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.7427, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.6244208812713623, + "learning_rate": 3.718521527213648e-06, + "loss": 1.7097, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8882938623428345, + "eval_runtime": 36.9853, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.434684991836548, + "learning_rate": 3.716490658001625e-06, + "loss": 1.6875, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.6481709480285645, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5726, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.5398213863372803, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.6881, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.684122085571289, + "learning_rate": 3.710398050365557e-06, + "loss": 1.691, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.712214946746826, + "learning_rate": 3.708367181153534e-06, + "loss": 1.653, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.8427517414093018, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.6853, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.676410436630249, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.645, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.5397610664367676, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6171, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.606250286102295, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6729, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.3617212772369385, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.6546, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8915373086929321, + "eval_runtime": 37.094, + "eval_samples_per_second": 26.959, + "eval_steps_per_second": 1.132, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.6635043621063232, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.628, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.724837303161621, + "learning_rate": 3.694151096669375e-06, + "loss": 1.7537, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.56447434425354, + "learning_rate": 3.692120227457352e-06, + "loss": 1.6638, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6624584197998047, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.6806, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.7147905826568604, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5887, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.683840751647949, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.6664, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.8141531944274902, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6744, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.6172802448272705, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6866, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.7853286266326904, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.7034, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.7121260166168213, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6573, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.883426308631897, + "eval_runtime": 37.1964, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.4775121212005615, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6651, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.5361502170562744, + "learning_rate": 3.673842404549147e-06, + "loss": 1.6524, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.7969970703125, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6598, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.88427996635437, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.6821, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.9321346282958984, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.7777, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.802886724472046, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6808, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.85810923576355, + "learning_rate": 3.663688058489034e-06, + "loss": 1.7375, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.6501998901367188, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.6662, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.550652265548706, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6084, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.5003185272216797, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.681, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8800761699676514, + "eval_runtime": 37.2014, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.4262983798980713, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6608, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6474270820617676, + "learning_rate": 3.65353371242892e-06, + "loss": 1.6972, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.5103557109832764, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5581, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.529756546020508, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.693, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.578362464904785, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6302, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8062336444854736, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6855, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.935694932937622, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6861, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5260093212127686, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6534, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.7627487182617188, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.6833, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.9553534984588623, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6887, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.8832038640975952, + "eval_runtime": 37.1828, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.5526084899902344, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.7162, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.761218786239624, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6823, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.672189950942993, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6489, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.7650046348571777, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.7201, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.6351473331451416, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.6164, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5959131717681885, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.603, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.525513172149658, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6671, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.5529232025146484, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6446, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6846, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6786, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.885197639465332, + "eval_runtime": 37.2374, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.50577449798584, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6892, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.57772159576416, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5863, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.5455222129821777, + "learning_rate": 3.612916328188465e-06, + "loss": 1.692, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.6154329776763916, + "learning_rate": 3.610885458976442e-06, + "loss": 1.7027, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.5776023864746094, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.6353, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.5480730533599854, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.6358, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.5939579010009766, + "learning_rate": 3.604792851340374e-06, + "loss": 1.7099, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.65395188331604, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6763, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.5302748680114746, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.613, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.669433832168579, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.6584, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8816337585449219, + "eval_runtime": 37.2174, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.859063148498535, + "learning_rate": 3.596669374492283e-06, + "loss": 1.7096, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.531301259994507, + "learning_rate": 3.59463850528026e-06, + "loss": 1.6168, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.5920488834381104, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6953, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.698284387588501, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6602, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.762387752532959, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.7073, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.554579019546509, + "learning_rate": 3.586515028432169e-06, + "loss": 1.7151, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.644576072692871, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.6088, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.6375787258148193, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6434, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9475791454315186, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6936, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.5753467082977295, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5915, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8799077272415161, + "eval_runtime": 37.2335, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8163156509399414, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6824, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.6741368770599365, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.7066, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.5235214233398438, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6779, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.620927095413208, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6392, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7678003311157227, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6768, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.5900607109069824, + "learning_rate": 3.566206336311942e-06, + "loss": 1.618, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.4440128803253174, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5981, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.686119318008423, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6649, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.6912872791290283, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.6426, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.7249274253845215, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.6639, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8750032186508179, + "eval_runtime": 37.2319, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.750321626663208, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.7014, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.593047618865967, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6557, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.744931936264038, + "learning_rate": 3.551990251827783e-06, + "loss": 1.6257, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.4940085411071777, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6546, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.6905856132507324, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.6395, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.6744513511657715, + "learning_rate": 3.545897644191714e-06, + "loss": 1.6159, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6766908168792725, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6754, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.743103504180908, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.6273, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.886472463607788, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6836, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.823476791381836, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6475, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8762013912200928, + "eval_runtime": 36.8594, + "eval_samples_per_second": 27.13, + "eval_steps_per_second": 1.139, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.741985559463501, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6765, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.6321418285369873, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6997, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7359395027160645, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6529, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.610210657119751, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.68, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.7770233154296875, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6836, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.7146646976470947, + "learning_rate": 3.525588952071487e-06, + "loss": 1.699, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.782172679901123, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6941, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.577692747116089, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.6303, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.721348285675049, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.6381, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.825709342956543, + "learning_rate": 3.517465475223396e-06, + "loss": 1.6144, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8759764432907104, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.8446385860443115, + "learning_rate": 3.515434606011373e-06, + "loss": 1.685, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.70577335357666, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.7276, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.6624722480773926, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6735, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.7956180572509766, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6413, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.792559862136841, + "learning_rate": 3.507311129163282e-06, + "loss": 1.6035, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.7113680839538574, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6461, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.656461715698242, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.662, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.8650670051574707, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5952, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.8043570518493652, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6738, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.607752799987793, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6828, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8729931116104126, + "eval_runtime": 37.2275, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.5819315910339355, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.6108, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6494336128234863, + "learning_rate": 3.493095044679123e-06, + "loss": 1.7348, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.552762508392334, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5534, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.6580469608306885, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6675, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.922295331954956, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5819, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.7051825523376465, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6573, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.738743782043457, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6939, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.51552677154541, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.665, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.578549385070801, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6642, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.7047011852264404, + "learning_rate": 3.476848090982941e-06, + "loss": 1.6298, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8726528882980347, + "eval_runtime": 37.25, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.722877025604248, + "learning_rate": 3.474817221770918e-06, + "loss": 1.7522, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.464240074157715, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6876, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.6737945079803467, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.6453, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.7960164546966553, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.607, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.6985723972320557, + "learning_rate": 3.466693744922827e-06, + "loss": 1.6385, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.7561540603637695, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6753, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.828885555267334, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6402, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.653183698654175, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6889, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.5803189277648926, + "learning_rate": 3.458570268074736e-06, + "loss": 1.654, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.5868985652923584, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6366, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8656586408615112, + "eval_runtime": 37.2805, + "eval_samples_per_second": 26.824, + "eval_steps_per_second": 1.127, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.648036241531372, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.6019, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.5482382774353027, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6597, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.700294017791748, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.6075, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.9346179962158203, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6558, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.893852710723877, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6395, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.789057970046997, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6833, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.745492935180664, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.6266, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.7930359840393066, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5781, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.770575761795044, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6518, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 2.7367143630981445, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6982, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8708553314208984, + "eval_runtime": 37.2843, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.89182448387146, + "learning_rate": 3.434199837530463e-06, + "loss": 1.6587, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.5572729110717773, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5974, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8172996044158936, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6941, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.6480889320373535, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6902, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.654735803604126, + "learning_rate": 3.426076360682372e-06, + "loss": 1.7083, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.4182465076446533, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6494, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.6808230876922607, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5782, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.648282527923584, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6695, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.5920112133026123, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.7234, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.690538167953491, + "learning_rate": 3.415922014622259e-06, + "loss": 1.686, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8717143535614014, + "eval_runtime": 37.2714, + "eval_samples_per_second": 26.83, + "eval_steps_per_second": 1.127, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.672515392303467, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6883, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.8012726306915283, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6762, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.7551519870758057, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6694, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.687445878982544, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.6267, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.674445390701294, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6944, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.797832489013672, + "learning_rate": 3.403736799350122e-06, + "loss": 1.6735, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7843174934387207, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.6183, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.678809404373169, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.7336, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.748838424682617, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6499, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.5785276889801025, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6589, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8695456981658936, + "eval_runtime": 37.2493, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.7307844161987305, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5989, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.83357572555542, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6568, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.6178860664367676, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6654, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.699328899383545, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6657, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.7418158054351807, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6351, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.5589001178741455, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.6351, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.492795944213867, + "learning_rate": 3.381397238017872e-06, + "loss": 1.6188, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 2.8586132526397705, + "learning_rate": 3.379366368805849e-06, + "loss": 1.7127, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7773218154907227, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6529, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.6956539154052734, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.6566, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8692638874053955, + "eval_runtime": 37.2317, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.7406537532806396, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6482, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.746194362640381, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.6343, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.767970323562622, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6577, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.730640172958374, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5941, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.5661444664001465, + "learning_rate": 3.36515028432169e-06, + "loss": 1.6362, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.5915091037750244, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6526, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.7135071754455566, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6999, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.591201066970825, + "learning_rate": 3.359057676685622e-06, + "loss": 1.684, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.564439296722412, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5815, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.4480631351470947, + "learning_rate": 3.354995938261576e-06, + "loss": 1.6284, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8628013134002686, + "eval_runtime": 37.23, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.5392649173736572, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5841, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6359004974365234, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.6278, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.803671360015869, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6602, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.6031086444854736, + "learning_rate": 3.346872461413485e-06, + "loss": 1.6021, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.687382459640503, + "learning_rate": 3.344841592201463e-06, + "loss": 1.6207, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.6388001441955566, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.6194, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.603121042251587, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6532, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.8053836822509766, + "learning_rate": 3.338748984565394e-06, + "loss": 1.68, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.5472421646118164, + "learning_rate": 3.336718115353372e-06, + "loss": 1.7164, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.5142838954925537, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6441, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.8628681898117065, + "eval_runtime": 36.8298, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.6917803287506104, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.6235, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.6062285900115967, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6488, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.5503156185150146, + "learning_rate": 3.32859463850528e-06, + "loss": 1.639, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.433518648147583, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5951, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.600661516189575, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.6148, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.5558395385742188, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.6334, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.6363396644592285, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5858, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.819617748260498, + "learning_rate": 3.318440292445167e-06, + "loss": 1.6274, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 2.9121081829071045, + "learning_rate": 3.316409423233144e-06, + "loss": 1.6081, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.734858989715576, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6889, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8644160032272339, + "eval_runtime": 37.0389, + "eval_samples_per_second": 26.999, + "eval_steps_per_second": 1.134, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.6940014362335205, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6463, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.82663631439209, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6474, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.634434223175049, + "learning_rate": 3.308285946385053e-06, + "loss": 1.7172, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.6721079349517822, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5795, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.786130666732788, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6634, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7698960304260254, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.687, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.6305580139160156, + "learning_rate": 3.300162469536962e-06, + "loss": 1.6372, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.5221288204193115, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6746, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.583643674850464, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6538, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.5978968143463135, + "learning_rate": 3.294069861900894e-06, + "loss": 1.7177, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8665071725845337, + "eval_runtime": 37.1194, + "eval_samples_per_second": 26.94, + "eval_steps_per_second": 1.131, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.6446492671966553, + "learning_rate": 3.292038992688871e-06, + "loss": 1.657, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.646174192428589, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.682, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.671532392501831, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.7033, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.616450309753418, + "learning_rate": 3.285946385052803e-06, + "loss": 1.586, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.552800416946411, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6599, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.5700533390045166, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5593, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.675553560256958, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.7097, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.6610019207000732, + "learning_rate": 3.277822908204712e-06, + "loss": 1.6408, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.55435848236084, + "learning_rate": 3.275792038992689e-06, + "loss": 1.5909, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.811556816101074, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6552, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.8618569374084473, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.7060115337371826, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.6144, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.6817731857299805, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6481, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.8206212520599365, + "learning_rate": 3.267668562144598e-06, + "loss": 1.6294, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.834332227706909, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.6103, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.754049062728882, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6788, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.6089437007904053, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6469, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.6095848083496094, + "learning_rate": 3.259545085296507e-06, + "loss": 1.7122, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.606787919998169, + "learning_rate": 3.257514216084484e-06, + "loss": 1.5851, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.643245220184326, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.7042, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.8400895595550537, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6526, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8640165328979492, + "eval_runtime": 37.162, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.519524335861206, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.6268, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.565365791320801, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6562, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.6364638805389404, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.6143, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.6235382556915283, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.64, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.7358779907226562, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6509, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.6720035076141357, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5902, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5960681438446045, + "learning_rate": 3.23923639317628e-06, + "loss": 1.6378, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.6655092239379883, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.6102, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.7873294353485107, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6558, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.746189594268799, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.7116, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8605948686599731, + "eval_runtime": 37.1486, + "eval_samples_per_second": 26.919, + "eval_steps_per_second": 1.131, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5058770179748535, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.6156, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.8002450466156006, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6884, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.696186065673828, + "learning_rate": 3.227051177904143e-06, + "loss": 1.6403, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.5473484992980957, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.6375, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.6391875743865967, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6597, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.647244453430176, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5666, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849031448364258, + "learning_rate": 3.218927701056052e-06, + "loss": 1.6211, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.7565410137176514, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5686, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.938270092010498, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.7046, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.6462385654449463, + "learning_rate": 3.212835093419984e-06, + "loss": 1.602, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.8577098846435547, + "eval_runtime": 37.1448, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.679330587387085, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.636, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.6410343647003174, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6904, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6398510932922363, + "learning_rate": 3.206742485783916e-06, + "loss": 1.597, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.4636199474334717, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5823, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.646580457687378, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6591, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.8582923412323, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.6731, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.5474820137023926, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5656, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6609606742858887, + "learning_rate": 3.196588139723802e-06, + "loss": 1.6374, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.651947021484375, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6557, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.6957671642303467, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.6436, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.8583322763442993, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.5442283153533936, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6424, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.593792200088501, + "learning_rate": 3.188464662875711e-06, + "loss": 1.7012, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.5021846294403076, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.6268, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.7368624210357666, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5953, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.5100302696228027, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5866, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.5767033100128174, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6621, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.4694652557373047, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6554, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.606158971786499, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.6238, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.857424020767212, + "learning_rate": 3.174248578391552e-06, + "loss": 1.626, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.6120405197143555, + "learning_rate": 3.172217709179529e-06, + "loss": 1.6134, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8599685430526733, + "eval_runtime": 37.1373, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.61995530128479, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.6303, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6253163814544678, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5886, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.627302885055542, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6526, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.8007607460021973, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.6581, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.551227331161499, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5921, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.7288761138916016, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6478, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5698401927948, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6552, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.796701192855835, + "learning_rate": 3.155970755483347e-06, + "loss": 1.6391, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.6523337364196777, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6495, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.5528745651245117, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6396, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8551454544067383, + "eval_runtime": 37.1721, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.650125741958618, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6803, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.5522522926330566, + "learning_rate": 3.147847278635256e-06, + "loss": 1.726, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5302486419677734, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.6238, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.517793893814087, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5879, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.5749523639678955, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.6361, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6382429599761963, + "learning_rate": 3.139723801787165e-06, + "loss": 1.6143, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.7513139247894287, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6608, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.616007089614868, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5842, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.619389772415161, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6419, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.6845619678497314, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6726, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8519575595855713, + "eval_runtime": 36.823, + "eval_samples_per_second": 27.157, + "eval_steps_per_second": 1.141, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.9907562732696533, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6813, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5070035457611084, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6449, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.6286256313323975, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6614, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5735433101654053, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.6191, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.637507438659668, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6649, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.5027501583099365, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6779, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.8430395126342773, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.6428, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.6394317150115967, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.655, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.7121026515960693, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6798, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.6572396755218506, + "learning_rate": 3.111291632818847e-06, + "loss": 1.6937, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.857973575592041, + "eval_runtime": 37.0635, + "eval_samples_per_second": 26.981, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.660865306854248, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6454, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.6230008602142334, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6562, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.60079026222229, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.6311, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.6757969856262207, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6673, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.6204211711883545, + "learning_rate": 3.101137286758733e-06, + "loss": 1.665, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.7115979194641113, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6864, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.604799270629883, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.6364, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 2.7463159561157227, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.699, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.744046926498413, + "learning_rate": 3.093013809910642e-06, + "loss": 1.665, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8082664012908936, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.7027, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8549033403396606, + "eval_runtime": 37.1604, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.6146464347839355, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6635, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.724843978881836, + "learning_rate": 3.086921202274574e-06, + "loss": 1.64, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.699645757675171, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5988, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.769742727279663, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6478, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.5372848510742188, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.6285, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6461033821105957, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5669, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.8221795558929443, + "learning_rate": 3.07676685621446e-06, + "loss": 1.6455, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6035211086273193, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.6063, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.533945322036743, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.6234, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.627448797225952, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6381, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8547505140304565, + "eval_runtime": 37.226, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.6822423934936523, + "learning_rate": 3.068643379366369e-06, + "loss": 1.6436, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.647813558578491, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5901, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.619610548019409, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6559, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7284183502197266, + "learning_rate": 3.062550771730301e-06, + "loss": 1.6167, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.585660934448242, + "learning_rate": 3.060519902518278e-06, + "loss": 1.6176, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.689528465270996, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5651, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.484328508377075, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.6487, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.629913806915283, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.6099, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.711730480194092, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6702, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.590545415878296, + "learning_rate": 3.050365556458164e-06, + "loss": 1.6078, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8545923233032227, + "eval_runtime": 37.2534, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.5952389240264893, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.6084, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.739316940307617, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.6427, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.811095952987671, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6817, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.606715440750122, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6695, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.481234550476074, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5762, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.508068084716797, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.612, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.5694284439086914, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.6457, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.7029778957366943, + "learning_rate": 3.034118602761982e-06, + "loss": 1.6201, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.578904867172241, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5978, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.780586004257202, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6793, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.8522270917892456, + "eval_runtime": 37.2388, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.648653984069824, + "learning_rate": 3.028025995125914e-06, + "loss": 1.6626, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.631589651107788, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5822, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.5863327980041504, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.6322, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6556999683380127, + "learning_rate": 3.021933387489846e-06, + "loss": 1.5727, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.6391682624816895, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6334, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.6488680839538574, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6588, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.743858814239502, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5963, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.6499438285827637, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5997, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5450377464294434, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.6122, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.592026948928833, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.6362, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.84926176071167, + "eval_runtime": 37.2501, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.5699622631073, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.6055, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.779695987701416, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6503, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.6080546379089355, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.6121, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.70405650138855, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6774, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.69488525390625, + "learning_rate": 2.999593826157596e-06, + "loss": 1.6186, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.9266953468322754, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.6485, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.615251302719116, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.624, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.574878215789795, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6637, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6952786445617676, + "learning_rate": 2.991470349309505e-06, + "loss": 1.627, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.6076977252960205, + "learning_rate": 2.989439480097482e-06, + "loss": 1.6175, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.8475773334503174, + "eval_runtime": 37.2328, + "eval_samples_per_second": 26.858, + "eval_steps_per_second": 1.128, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.5707366466522217, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.6332, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.838609457015991, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6764, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.771082878112793, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6622, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6113975048065186, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6405, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.7537200450897217, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.6376, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.683002471923828, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6679, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.651897430419922, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.6091, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.6856229305267334, + "learning_rate": 2.9731925264013e-06, + "loss": 1.6223, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.646522045135498, + "learning_rate": 2.971161657189277e-06, + "loss": 1.6126, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.592247486114502, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.6088, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.8501410484313965, + "eval_runtime": 37.225, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.03291392326355, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.6051, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.903224468231201, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6753, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.5178730487823486, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5894, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.6278252601623535, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6687, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.8105742931365967, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6648, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.5897178649902344, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.6012, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.7172093391418457, + "learning_rate": 2.954914703493095e-06, + "loss": 1.6206, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.6845815181732178, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.6163, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.652451753616333, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.5916, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.740280866622925, + "learning_rate": 2.948822095857027e-06, + "loss": 1.627, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8489265441894531, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.786797285079956, + "learning_rate": 2.946791226645004e-06, + "loss": 1.6005, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.6826601028442383, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.5887, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 2.881887912750244, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.6198, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.86600923538208, + "learning_rate": 2.940698619008936e-06, + "loss": 1.6252, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.7267069816589355, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.6314, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.768399477005005, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.6434, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.70622181892395, + "learning_rate": 2.934606011372868e-06, + "loss": 1.6412, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.803354501724243, + "learning_rate": 2.932575142160845e-06, + "loss": 1.6402, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8491063117980957, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.6244, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.5652339458465576, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.6315, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.8461908102035522, + "eval_runtime": 36.9233, + "eval_samples_per_second": 27.083, + "eval_steps_per_second": 1.137, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.7153713703155518, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.6714, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 2.779289484024048, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.6355, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.6597917079925537, + "learning_rate": 2.922420796100731e-06, + "loss": 1.5574, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.7135894298553467, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.6721, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.562045097351074, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.5745, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.6727097034454346, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.6324, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 2.688309907913208, + "learning_rate": 2.91429731925264e-06, + "loss": 1.6698, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.8429853916168213, + "learning_rate": 2.912266450040618e-06, + "loss": 1.6378, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.658046245574951, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.6237, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 2.6601390838623047, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.6163, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.843716025352478, + "eval_runtime": 37.2087, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.949215888977051, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.5883, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 2.823991298675537, + "learning_rate": 2.904142973192527e-06, + "loss": 1.6786, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.6177661418914795, + "learning_rate": 2.902112103980504e-06, + "loss": 1.6337, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.493638038635254, + "learning_rate": 2.900081234768481e-06, + "loss": 1.6328, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.657914161682129, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.6068, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.7800004482269287, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.5963, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.7934350967407227, + "learning_rate": 2.893988627132413e-06, + "loss": 1.611, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.8139140605926514, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.5621, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.5909969806671143, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.6195, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.631871461868286, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.6146, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.846065878868103, + "eval_runtime": 37.2421, + "eval_samples_per_second": 26.851, + "eval_steps_per_second": 1.128, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.667304754257202, + "learning_rate": 2.885865150284322e-06, + "loss": 1.6482, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.5922656059265137, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.6231, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.663452386856079, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.5772, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.6463005542755127, + "learning_rate": 2.879772542648254e-06, + "loss": 1.6333, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.5469250679016113, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.5465, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.7316017150878906, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.6856, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.7127692699432373, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.667, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 2.7006423473358154, + "learning_rate": 2.871649065800163e-06, + "loss": 1.677, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 2.788804054260254, + "learning_rate": 2.86961819658814e-06, + "loss": 1.6092, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.7734181880950928, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.6377, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.8444476127624512, + "eval_runtime": 37.2371, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.6005876064300537, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.6072, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.684293270111084, + "learning_rate": 2.863525588952072e-06, + "loss": 1.6669, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.6160528659820557, + "learning_rate": 2.861494719740049e-06, + "loss": 1.5439, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.68396258354187, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.5909, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.6292917728424072, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.6225, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.702333450317383, + "learning_rate": 2.855402112103981e-06, + "loss": 1.6224, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.850862503051758, + "learning_rate": 2.853371242891958e-06, + "loss": 1.5995, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.861351490020752, + "learning_rate": 2.851340373679935e-06, + "loss": 1.6254, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.7731313705444336, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.6555, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.669189691543579, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.6238, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.845217227935791, + "eval_runtime": 37.2809, + "eval_samples_per_second": 26.823, + "eval_steps_per_second": 1.127, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.64485239982605, + "learning_rate": 2.845247766043867e-06, + "loss": 1.5989, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.7558820247650146, + "learning_rate": 2.843216896831844e-06, + "loss": 1.5615, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.7818500995635986, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.5962, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.571547746658325, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.6434, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 2.8242406845092773, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.6537, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.5545597076416016, + "learning_rate": 2.835093419983753e-06, + "loss": 1.6158, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.7329461574554443, + "learning_rate": 2.833062550771731e-06, + "loss": 1.591, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.5414955615997314, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.5727, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.7625439167022705, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.6343, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.7839064598083496, + "learning_rate": 2.826969943135662e-06, + "loss": 1.6187, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.8442766666412354, + "eval_runtime": 37.2101, + "eval_samples_per_second": 26.874, + "eval_steps_per_second": 1.129, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.615706443786621, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.621, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.6295979022979736, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.5913, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.7428646087646484, + "learning_rate": 2.820877335499594e-06, + "loss": 1.6015, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.5353381633758545, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.5798, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.6209659576416016, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.5783, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.791372299194336, + "learning_rate": 2.814784727863526e-06, + "loss": 1.6082, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.781949520111084, + "learning_rate": 2.812753858651503e-06, + "loss": 1.6337, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 3.0437769889831543, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.5969, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 2.7201249599456787, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.5717, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 2.766634464263916, + "learning_rate": 2.806661251015435e-06, + "loss": 1.6654, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.8452297449111938, + "eval_runtime": 37.2452, + "eval_samples_per_second": 26.849, + "eval_steps_per_second": 1.128, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.726876735687256, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.6652, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 2.775012969970703, + "learning_rate": 2.802599512591389e-06, + "loss": 1.7293, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.7603225708007812, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.6226, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.723904609680176, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.5663, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 2.667384624481201, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.6209, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.7580318450927734, + "learning_rate": 2.794476035743298e-06, + "loss": 1.6705, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.9216315746307373, + "learning_rate": 2.792445166531276e-06, + "loss": 1.6361, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 2.6552748680114746, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.6498, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.764836549758911, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.6009, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.5660064220428467, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.576, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.8392945528030396, + "eval_runtime": 37.2515, + "eval_samples_per_second": 26.845, + "eval_steps_per_second": 1.127, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.5110044479370117, + "learning_rate": 2.784321689683185e-06, + "loss": 1.5744, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.7352850437164307, + "learning_rate": 2.782290820471162e-06, + "loss": 1.5852, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.850947618484497, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.6692, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.7107245922088623, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.6427, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.8236117362976074, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.6375, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.763416051864624, + "learning_rate": 2.774167343623071e-06, + "loss": 1.6864, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 2.7317910194396973, + "learning_rate": 2.772136474411048e-06, + "loss": 1.6015, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.7454333305358887, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.5379, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.6102535724639893, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.6282, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.658440113067627, + "learning_rate": 2.76604386677498e-06, + "loss": 1.6042, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.8391484022140503, + "eval_runtime": 37.1746, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 2.5731043815612793, + "learning_rate": 2.764012997562957e-06, + "loss": 1.6446, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.680983543395996, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.5663, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 2.7525506019592285, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.6628, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.6905839443206787, + "learning_rate": 2.757920389926889e-06, + "loss": 1.6021, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 2.850900888442993, + "learning_rate": 2.755889520714866e-06, + "loss": 1.5984, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.6612954139709473, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.5627, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.632106304168701, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.6304, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.660377025604248, + "learning_rate": 2.749796913078798e-06, + "loss": 1.6622, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.497938394546509, + "learning_rate": 2.747766043866775e-06, + "loss": 1.6152, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.655977725982666, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.6703, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.8361152410507202, + "eval_runtime": 37.2147, + "eval_samples_per_second": 26.871, + "eval_steps_per_second": 1.129, + "step": 1400 + }, + { + "epoch": 1.5353424657534247, + "grad_norm": 2.664874792098999, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.7109, + "step": 1401 + }, + { + "epoch": 1.5364383561643835, + "grad_norm": 2.5773956775665283, + "learning_rate": 2.741673436230707e-06, + "loss": 1.6176, + "step": 1402 + }, + { + "epoch": 1.5375342465753423, + "grad_norm": 2.7094814777374268, + "learning_rate": 2.739642567018684e-06, + "loss": 1.6075, + "step": 1403 + }, + { + "epoch": 1.5386301369863014, + "grad_norm": 2.652325391769409, + "learning_rate": 2.7376116978066615e-06, + "loss": 1.6189, + "step": 1404 + }, + { + "epoch": 1.5397260273972604, + "grad_norm": 2.6892807483673096, + "learning_rate": 2.7355808285946388e-06, + "loss": 1.6122, + "step": 1405 + }, + { + "epoch": 1.5408219178082192, + "grad_norm": 2.6621243953704834, + "learning_rate": 2.733549959382616e-06, + "loss": 1.6008, + "step": 1406 + }, + { + "epoch": 1.541917808219178, + "grad_norm": 2.773747444152832, + "learning_rate": 2.7315190901705933e-06, + "loss": 1.6723, + "step": 1407 + }, + { + "epoch": 1.543013698630137, + "grad_norm": 2.853525400161743, + "learning_rate": 2.7294882209585706e-06, + "loss": 1.641, + "step": 1408 + }, + { + "epoch": 1.5441095890410959, + "grad_norm": 2.6333844661712646, + "learning_rate": 2.727457351746548e-06, + "loss": 1.587, + "step": 1409 + }, + { + "epoch": 1.5452054794520547, + "grad_norm": 2.6299996376037598, + "learning_rate": 2.725426482534525e-06, + "loss": 1.5807, + "step": 1410 + }, + { + "epoch": 1.5452054794520547, + "eval_loss": 1.8377478122711182, + "eval_runtime": 36.8958, + "eval_samples_per_second": 27.103, + "eval_steps_per_second": 1.138, + "step": 1410 + }, + { + "epoch": 1.5463013698630137, + "grad_norm": 2.6222734451293945, + "learning_rate": 2.723395613322502e-06, + "loss": 1.6225, + "step": 1411 + }, + { + "epoch": 1.5473972602739727, + "grad_norm": 2.6702005863189697, + "learning_rate": 2.7213647441104797e-06, + "loss": 1.6031, + "step": 1412 + }, + { + "epoch": 1.5484931506849315, + "grad_norm": 2.6451900005340576, + "learning_rate": 2.7193338748984565e-06, + "loss": 1.5998, + "step": 1413 + }, + { + "epoch": 1.5495890410958904, + "grad_norm": 2.769784927368164, + "learning_rate": 2.7173030056864342e-06, + "loss": 1.7579, + "step": 1414 + }, + { + "epoch": 1.5506849315068494, + "grad_norm": 2.629871368408203, + "learning_rate": 2.715272136474411e-06, + "loss": 1.564, + "step": 1415 + }, + { + "epoch": 1.5517808219178082, + "grad_norm": 2.811937093734741, + "learning_rate": 2.713241267262389e-06, + "loss": 1.6236, + "step": 1416 + }, + { + "epoch": 1.552876712328767, + "grad_norm": 2.883277654647827, + "learning_rate": 2.7112103980503656e-06, + "loss": 1.5996, + "step": 1417 + }, + { + "epoch": 1.553972602739726, + "grad_norm": 2.610964059829712, + "learning_rate": 2.7091795288383433e-06, + "loss": 1.6391, + "step": 1418 + }, + { + "epoch": 1.555068493150685, + "grad_norm": 2.6637415885925293, + "learning_rate": 2.70714865962632e-06, + "loss": 1.5983, + "step": 1419 + }, + { + "epoch": 1.5561643835616439, + "grad_norm": 2.5372607707977295, + "learning_rate": 2.705117790414298e-06, + "loss": 1.6258, + "step": 1420 + }, + { + "epoch": 1.5561643835616439, + "eval_loss": 1.838557243347168, + "eval_runtime": 36.973, + "eval_samples_per_second": 27.047, + "eval_steps_per_second": 1.136, + "step": 1420 + }, + { + "epoch": 1.5572602739726027, + "grad_norm": 2.900594711303711, + "learning_rate": 2.7030869212022747e-06, + "loss": 1.67, + "step": 1421 + }, + { + "epoch": 1.5583561643835617, + "grad_norm": 2.4924063682556152, + "learning_rate": 2.7010560519902516e-06, + "loss": 1.5879, + "step": 1422 + }, + { + "epoch": 1.5594520547945205, + "grad_norm": 2.5761115550994873, + "learning_rate": 2.6990251827782293e-06, + "loss": 1.6287, + "step": 1423 + }, + { + "epoch": 1.5605479452054793, + "grad_norm": 2.645463466644287, + "learning_rate": 2.696994313566206e-06, + "loss": 1.5925, + "step": 1424 + }, + { + "epoch": 1.5616438356164384, + "grad_norm": 2.682466506958008, + "learning_rate": 2.694963444354184e-06, + "loss": 1.6723, + "step": 1425 + }, + { + "epoch": 1.5627397260273974, + "grad_norm": 2.499749183654785, + "learning_rate": 2.6929325751421607e-06, + "loss": 1.5998, + "step": 1426 + }, + { + "epoch": 1.563835616438356, + "grad_norm": 2.659956455230713, + "learning_rate": 2.6909017059301384e-06, + "loss": 1.6311, + "step": 1427 + }, + { + "epoch": 1.564931506849315, + "grad_norm": 2.6684396266937256, + "learning_rate": 2.6888708367181153e-06, + "loss": 1.5648, + "step": 1428 + }, + { + "epoch": 1.566027397260274, + "grad_norm": 2.6868791580200195, + "learning_rate": 2.686839967506093e-06, + "loss": 1.6419, + "step": 1429 + }, + { + "epoch": 1.5671232876712329, + "grad_norm": 2.6999104022979736, + "learning_rate": 2.68480909829407e-06, + "loss": 1.55, + "step": 1430 + }, + { + "epoch": 1.5671232876712329, + "eval_loss": 1.83835768699646, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1430 + }, + { + "epoch": 1.5682191780821917, + "grad_norm": 2.7338085174560547, + "learning_rate": 2.6827782290820475e-06, + "loss": 1.5507, + "step": 1431 + }, + { + "epoch": 1.5693150684931507, + "grad_norm": 2.7358195781707764, + "learning_rate": 2.6807473598700244e-06, + "loss": 1.6659, + "step": 1432 + }, + { + "epoch": 1.5704109589041098, + "grad_norm": 2.671983242034912, + "learning_rate": 2.678716490658002e-06, + "loss": 1.6293, + "step": 1433 + }, + { + "epoch": 1.5715068493150683, + "grad_norm": 2.551203489303589, + "learning_rate": 2.676685621445979e-06, + "loss": 1.623, + "step": 1434 + }, + { + "epoch": 1.5726027397260274, + "grad_norm": 2.6485984325408936, + "learning_rate": 2.674654752233956e-06, + "loss": 1.6127, + "step": 1435 + }, + { + "epoch": 1.5736986301369864, + "grad_norm": 2.643105983734131, + "learning_rate": 2.6726238830219335e-06, + "loss": 1.6155, + "step": 1436 + }, + { + "epoch": 1.5747945205479452, + "grad_norm": 2.769458055496216, + "learning_rate": 2.6705930138099107e-06, + "loss": 1.6576, + "step": 1437 + }, + { + "epoch": 1.575890410958904, + "grad_norm": 2.3996448516845703, + "learning_rate": 2.668562144597888e-06, + "loss": 1.5276, + "step": 1438 + }, + { + "epoch": 1.576986301369863, + "grad_norm": 2.5827524662017822, + "learning_rate": 2.6665312753858653e-06, + "loss": 1.6321, + "step": 1439 + }, + { + "epoch": 1.5780821917808219, + "grad_norm": 2.728794813156128, + "learning_rate": 2.6645004061738426e-06, + "loss": 1.6177, + "step": 1440 + }, + { + "epoch": 1.5780821917808219, + "eval_loss": 1.8374123573303223, + "eval_runtime": 37.1748, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1440 + }, + { + "epoch": 1.5791780821917807, + "grad_norm": 2.7500123977661133, + "learning_rate": 2.66246953696182e-06, + "loss": 1.5679, + "step": 1441 + }, + { + "epoch": 1.5802739726027397, + "grad_norm": 2.6778831481933594, + "learning_rate": 2.660438667749797e-06, + "loss": 1.5744, + "step": 1442 + }, + { + "epoch": 1.5813698630136988, + "grad_norm": 2.933138847351074, + "learning_rate": 2.6584077985377744e-06, + "loss": 1.6238, + "step": 1443 + }, + { + "epoch": 1.5824657534246576, + "grad_norm": 2.642246961593628, + "learning_rate": 2.6563769293257517e-06, + "loss": 1.6186, + "step": 1444 + }, + { + "epoch": 1.5835616438356164, + "grad_norm": 2.5519111156463623, + "learning_rate": 2.654346060113729e-06, + "loss": 1.5798, + "step": 1445 + }, + { + "epoch": 1.5846575342465754, + "grad_norm": 2.569063186645508, + "learning_rate": 2.6523151909017062e-06, + "loss": 1.5881, + "step": 1446 + }, + { + "epoch": 1.5857534246575342, + "grad_norm": 2.501680850982666, + "learning_rate": 2.6502843216896835e-06, + "loss": 1.5768, + "step": 1447 + }, + { + "epoch": 1.586849315068493, + "grad_norm": 2.7483339309692383, + "learning_rate": 2.6482534524776603e-06, + "loss": 1.6889, + "step": 1448 + }, + { + "epoch": 1.587945205479452, + "grad_norm": 3.004084587097168, + "learning_rate": 2.646222583265638e-06, + "loss": 1.6565, + "step": 1449 + }, + { + "epoch": 1.589041095890411, + "grad_norm": 2.793058156967163, + "learning_rate": 2.644191714053615e-06, + "loss": 1.6246, + "step": 1450 + }, + { + "epoch": 1.589041095890411, + "eval_loss": 1.83695387840271, + "eval_runtime": 37.1748, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1450 + }, + { + "epoch": 1.59013698630137, + "grad_norm": 2.803190231323242, + "learning_rate": 2.6421608448415926e-06, + "loss": 1.6051, + "step": 1451 + }, + { + "epoch": 1.5912328767123287, + "grad_norm": 2.7721307277679443, + "learning_rate": 2.6401299756295694e-06, + "loss": 1.5712, + "step": 1452 + }, + { + "epoch": 1.5923287671232877, + "grad_norm": 2.9235639572143555, + "learning_rate": 2.638099106417547e-06, + "loss": 1.6197, + "step": 1453 + }, + { + "epoch": 1.5934246575342466, + "grad_norm": 2.609719753265381, + "learning_rate": 2.636068237205524e-06, + "loss": 1.5976, + "step": 1454 + }, + { + "epoch": 1.5945205479452054, + "grad_norm": 2.7927262783050537, + "learning_rate": 2.6340373679935017e-06, + "loss": 1.6844, + "step": 1455 + }, + { + "epoch": 1.5956164383561644, + "grad_norm": 2.8688440322875977, + "learning_rate": 2.6320064987814785e-06, + "loss": 1.5285, + "step": 1456 + }, + { + "epoch": 1.5967123287671234, + "grad_norm": 2.819079637527466, + "learning_rate": 2.6299756295694562e-06, + "loss": 1.6171, + "step": 1457 + }, + { + "epoch": 1.5978082191780822, + "grad_norm": 2.825784206390381, + "learning_rate": 2.627944760357433e-06, + "loss": 1.6656, + "step": 1458 + }, + { + "epoch": 1.598904109589041, + "grad_norm": 2.6807522773742676, + "learning_rate": 2.6259138911454104e-06, + "loss": 1.6839, + "step": 1459 + }, + { + "epoch": 1.6, + "grad_norm": 2.9442570209503174, + "learning_rate": 2.6238830219333877e-06, + "loss": 1.6547, + "step": 1460 + }, + { + "epoch": 1.6, + "eval_loss": 1.8358429670333862, + "eval_runtime": 37.1499, + "eval_samples_per_second": 26.918, + "eval_steps_per_second": 1.131, + "step": 1460 + }, + { + "epoch": 1.601095890410959, + "grad_norm": 3.0025434494018555, + "learning_rate": 2.621852152721365e-06, + "loss": 1.5688, + "step": 1461 + }, + { + "epoch": 1.6021917808219177, + "grad_norm": 3.0999062061309814, + "learning_rate": 2.619821283509342e-06, + "loss": 1.639, + "step": 1462 + }, + { + "epoch": 1.6032876712328767, + "grad_norm": 2.7106330394744873, + "learning_rate": 2.6177904142973195e-06, + "loss": 1.5682, + "step": 1463 + }, + { + "epoch": 1.6043835616438358, + "grad_norm": 2.968259811401367, + "learning_rate": 2.6157595450852968e-06, + "loss": 1.6421, + "step": 1464 + }, + { + "epoch": 1.6054794520547944, + "grad_norm": 2.964728593826294, + "learning_rate": 2.613728675873274e-06, + "loss": 1.6626, + "step": 1465 + }, + { + "epoch": 1.6065753424657534, + "grad_norm": 2.8934402465820312, + "learning_rate": 2.6116978066612513e-06, + "loss": 1.6742, + "step": 1466 + }, + { + "epoch": 1.6076712328767124, + "grad_norm": 2.7678956985473633, + "learning_rate": 2.6096669374492286e-06, + "loss": 1.6139, + "step": 1467 + }, + { + "epoch": 1.6087671232876712, + "grad_norm": 2.74017333984375, + "learning_rate": 2.607636068237206e-06, + "loss": 1.5726, + "step": 1468 + }, + { + "epoch": 1.60986301369863, + "grad_norm": 2.8686470985412598, + "learning_rate": 2.605605199025183e-06, + "loss": 1.638, + "step": 1469 + }, + { + "epoch": 1.610958904109589, + "grad_norm": 3.008751392364502, + "learning_rate": 2.6035743298131604e-06, + "loss": 1.6246, + "step": 1470 + }, + { + "epoch": 1.610958904109589, + "eval_loss": 1.8375154733657837, + "eval_runtime": 37.1808, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 1470 + }, + { + "epoch": 1.612054794520548, + "grad_norm": 3.030925989151001, + "learning_rate": 2.6015434606011377e-06, + "loss": 1.6558, + "step": 1471 + }, + { + "epoch": 1.6131506849315067, + "grad_norm": 3.1855311393737793, + "learning_rate": 2.5995125913891145e-06, + "loss": 1.6488, + "step": 1472 + }, + { + "epoch": 1.6142465753424657, + "grad_norm": 2.943204879760742, + "learning_rate": 2.5974817221770922e-06, + "loss": 1.6211, + "step": 1473 + }, + { + "epoch": 1.6153424657534248, + "grad_norm": 2.681720733642578, + "learning_rate": 2.595450852965069e-06, + "loss": 1.655, + "step": 1474 + }, + { + "epoch": 1.6164383561643836, + "grad_norm": 2.750523567199707, + "learning_rate": 2.593419983753047e-06, + "loss": 1.6624, + "step": 1475 + }, + { + "epoch": 1.6175342465753424, + "grad_norm": 2.8923532962799072, + "learning_rate": 2.5913891145410236e-06, + "loss": 1.6046, + "step": 1476 + }, + { + "epoch": 1.6186301369863014, + "grad_norm": 2.852468490600586, + "learning_rate": 2.5893582453290013e-06, + "loss": 1.6557, + "step": 1477 + }, + { + "epoch": 1.6197260273972602, + "grad_norm": 2.7146172523498535, + "learning_rate": 2.587327376116978e-06, + "loss": 1.5888, + "step": 1478 + }, + { + "epoch": 1.620821917808219, + "grad_norm": 2.7109804153442383, + "learning_rate": 2.585296506904956e-06, + "loss": 1.6049, + "step": 1479 + }, + { + "epoch": 1.621917808219178, + "grad_norm": 2.9429404735565186, + "learning_rate": 2.5832656376929327e-06, + "loss": 1.6046, + "step": 1480 + }, + { + "epoch": 1.621917808219178, + "eval_loss": 1.837095856666565, + "eval_runtime": 37.19, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 1480 + }, + { + "epoch": 1.6230136986301371, + "grad_norm": 2.89047908782959, + "learning_rate": 2.5812347684809104e-06, + "loss": 1.5305, + "step": 1481 + }, + { + "epoch": 1.624109589041096, + "grad_norm": 3.0587005615234375, + "learning_rate": 2.5792038992688873e-06, + "loss": 1.7029, + "step": 1482 + }, + { + "epoch": 1.6252054794520547, + "grad_norm": 2.7209622859954834, + "learning_rate": 2.577173030056865e-06, + "loss": 1.5424, + "step": 1483 + }, + { + "epoch": 1.6263013698630138, + "grad_norm": 2.8400819301605225, + "learning_rate": 2.575142160844842e-06, + "loss": 1.6411, + "step": 1484 + }, + { + "epoch": 1.6273972602739726, + "grad_norm": 2.7427124977111816, + "learning_rate": 2.5731112916328187e-06, + "loss": 1.6421, + "step": 1485 + }, + { + "epoch": 1.6284931506849314, + "grad_norm": 2.6996700763702393, + "learning_rate": 2.5710804224207964e-06, + "loss": 1.5845, + "step": 1486 + }, + { + "epoch": 1.6295890410958904, + "grad_norm": 2.540188789367676, + "learning_rate": 2.5690495532087733e-06, + "loss": 1.5472, + "step": 1487 + }, + { + "epoch": 1.6306849315068495, + "grad_norm": 2.7682793140411377, + "learning_rate": 2.567018683996751e-06, + "loss": 1.6347, + "step": 1488 + }, + { + "epoch": 1.6317808219178083, + "grad_norm": 2.760413646697998, + "learning_rate": 2.564987814784728e-06, + "loss": 1.5826, + "step": 1489 + }, + { + "epoch": 1.632876712328767, + "grad_norm": 2.9961049556732178, + "learning_rate": 2.5629569455727055e-06, + "loss": 1.6037, + "step": 1490 + }, + { + "epoch": 1.632876712328767, + "eval_loss": 1.8322792053222656, + "eval_runtime": 37.2227, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 1490 + }, + { + "epoch": 1.633972602739726, + "grad_norm": 2.6983907222747803, + "learning_rate": 2.5609260763606824e-06, + "loss": 1.6142, + "step": 1491 + }, + { + "epoch": 1.635068493150685, + "grad_norm": 2.867401599884033, + "learning_rate": 2.55889520714866e-06, + "loss": 1.7227, + "step": 1492 + }, + { + "epoch": 1.6361643835616437, + "grad_norm": 2.7735934257507324, + "learning_rate": 2.556864337936637e-06, + "loss": 1.5856, + "step": 1493 + }, + { + "epoch": 1.6372602739726028, + "grad_norm": 2.7661404609680176, + "learning_rate": 2.5548334687246146e-06, + "loss": 1.6084, + "step": 1494 + }, + { + "epoch": 1.6383561643835618, + "grad_norm": 2.6881582736968994, + "learning_rate": 2.5528025995125915e-06, + "loss": 1.659, + "step": 1495 + }, + { + "epoch": 1.6394520547945204, + "grad_norm": 2.5745224952697754, + "learning_rate": 2.550771730300569e-06, + "loss": 1.6131, + "step": 1496 + }, + { + "epoch": 1.6405479452054794, + "grad_norm": 2.736159563064575, + "learning_rate": 2.548740861088546e-06, + "loss": 1.6469, + "step": 1497 + }, + { + "epoch": 1.6416438356164385, + "grad_norm": 2.6255311965942383, + "learning_rate": 2.5467099918765233e-06, + "loss": 1.5452, + "step": 1498 + }, + { + "epoch": 1.6427397260273973, + "grad_norm": 2.6919801235198975, + "learning_rate": 2.5446791226645006e-06, + "loss": 1.5282, + "step": 1499 + }, + { + "epoch": 1.643835616438356, + "grad_norm": 2.641533136367798, + "learning_rate": 2.542648253452478e-06, + "loss": 1.6524, + "step": 1500 + }, + { + "epoch": 1.643835616438356, + "eval_loss": 1.8287466764450073, + "eval_runtime": 37.2305, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1500 + }, + { + "epoch": 1.644931506849315, + "grad_norm": 2.7633602619171143, + "learning_rate": 2.540617384240455e-06, + "loss": 1.6199, + "step": 1501 + }, + { + "epoch": 1.6460273972602741, + "grad_norm": 2.618852376937866, + "learning_rate": 2.5385865150284324e-06, + "loss": 1.5803, + "step": 1502 + }, + { + "epoch": 1.6471232876712327, + "grad_norm": 2.857733964920044, + "learning_rate": 2.5365556458164097e-06, + "loss": 1.5968, + "step": 1503 + }, + { + "epoch": 1.6482191780821918, + "grad_norm": 2.8334357738494873, + "learning_rate": 2.534524776604387e-06, + "loss": 1.611, + "step": 1504 + }, + { + "epoch": 1.6493150684931508, + "grad_norm": 2.638340950012207, + "learning_rate": 2.5324939073923642e-06, + "loss": 1.5048, + "step": 1505 + }, + { + "epoch": 1.6504109589041096, + "grad_norm": 2.6521124839782715, + "learning_rate": 2.5304630381803415e-06, + "loss": 1.5967, + "step": 1506 + }, + { + "epoch": 1.6515068493150684, + "grad_norm": 2.639162540435791, + "learning_rate": 2.5284321689683188e-06, + "loss": 1.6084, + "step": 1507 + }, + { + "epoch": 1.6526027397260274, + "grad_norm": 2.6480295658111572, + "learning_rate": 2.526401299756296e-06, + "loss": 1.6115, + "step": 1508 + }, + { + "epoch": 1.6536986301369863, + "grad_norm": 2.7809157371520996, + "learning_rate": 2.524370430544273e-06, + "loss": 1.5909, + "step": 1509 + }, + { + "epoch": 1.654794520547945, + "grad_norm": 2.6947197914123535, + "learning_rate": 2.5223395613322506e-06, + "loss": 1.575, + "step": 1510 + }, + { + "epoch": 1.654794520547945, + "eval_loss": 1.8312958478927612, + "eval_runtime": 36.8551, + "eval_samples_per_second": 27.133, + "eval_steps_per_second": 1.14, + "step": 1510 + }, + { + "epoch": 1.655890410958904, + "grad_norm": 2.7290775775909424, + "learning_rate": 2.5203086921202274e-06, + "loss": 1.6169, + "step": 1511 + }, + { + "epoch": 1.6569863013698631, + "grad_norm": 2.7147257328033447, + "learning_rate": 2.518277822908205e-06, + "loss": 1.589, + "step": 1512 + }, + { + "epoch": 1.658082191780822, + "grad_norm": 2.648577928543091, + "learning_rate": 2.516246953696182e-06, + "loss": 1.6113, + "step": 1513 + }, + { + "epoch": 1.6591780821917808, + "grad_norm": 2.8041188716888428, + "learning_rate": 2.5142160844841597e-06, + "loss": 1.6128, + "step": 1514 + }, + { + "epoch": 1.6602739726027398, + "grad_norm": 2.7862393856048584, + "learning_rate": 2.5121852152721365e-06, + "loss": 1.6227, + "step": 1515 + }, + { + "epoch": 1.6613698630136986, + "grad_norm": 2.698709487915039, + "learning_rate": 2.5101543460601142e-06, + "loss": 1.6169, + "step": 1516 + }, + { + "epoch": 1.6624657534246574, + "grad_norm": 2.6966660022735596, + "learning_rate": 2.508123476848091e-06, + "loss": 1.5522, + "step": 1517 + }, + { + "epoch": 1.6635616438356164, + "grad_norm": 2.86776065826416, + "learning_rate": 2.506092607636069e-06, + "loss": 1.636, + "step": 1518 + }, + { + "epoch": 1.6646575342465755, + "grad_norm": 2.835958242416382, + "learning_rate": 2.5040617384240457e-06, + "loss": 1.6021, + "step": 1519 + }, + { + "epoch": 1.6657534246575343, + "grad_norm": 2.645120143890381, + "learning_rate": 2.5020308692120234e-06, + "loss": 1.5842, + "step": 1520 + }, + { + "epoch": 1.6657534246575343, + "eval_loss": 1.8320025205612183, + "eval_runtime": 37.0745, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 1520 + }, + { + "epoch": 1.666849315068493, + "grad_norm": 2.6796860694885254, + "learning_rate": 2.5e-06, + "loss": 1.6579, + "step": 1521 + }, + { + "epoch": 1.6679452054794521, + "grad_norm": 2.824993133544922, + "learning_rate": 2.4979691307879775e-06, + "loss": 1.5842, + "step": 1522 + }, + { + "epoch": 1.669041095890411, + "grad_norm": 2.554858922958374, + "learning_rate": 2.4959382615759548e-06, + "loss": 1.5956, + "step": 1523 + }, + { + "epoch": 1.6701369863013698, + "grad_norm": 2.621208429336548, + "learning_rate": 2.493907392363932e-06, + "loss": 1.6622, + "step": 1524 + }, + { + "epoch": 1.6712328767123288, + "grad_norm": 2.711733818054199, + "learning_rate": 2.4918765231519093e-06, + "loss": 1.6059, + "step": 1525 + }, + { + "epoch": 1.6723287671232878, + "grad_norm": 2.6141200065612793, + "learning_rate": 2.4898456539398866e-06, + "loss": 1.6105, + "step": 1526 + }, + { + "epoch": 1.6734246575342466, + "grad_norm": 2.693882942199707, + "learning_rate": 2.487814784727864e-06, + "loss": 1.6675, + "step": 1527 + }, + { + "epoch": 1.6745205479452054, + "grad_norm": 2.71944260597229, + "learning_rate": 2.4857839155158407e-06, + "loss": 1.5964, + "step": 1528 + }, + { + "epoch": 1.6756164383561645, + "grad_norm": 2.6505935192108154, + "learning_rate": 2.483753046303818e-06, + "loss": 1.5966, + "step": 1529 + }, + { + "epoch": 1.6767123287671233, + "grad_norm": 2.710515022277832, + "learning_rate": 2.4817221770917953e-06, + "loss": 1.6025, + "step": 1530 + }, + { + "epoch": 1.6767123287671233, + "eval_loss": 1.8318896293640137, + "eval_runtime": 37.2071, + "eval_samples_per_second": 26.877, + "eval_steps_per_second": 1.129, + "step": 1530 + }, + { + "epoch": 1.677808219178082, + "grad_norm": 2.8350000381469727, + "learning_rate": 2.4796913078797725e-06, + "loss": 1.6357, + "step": 1531 + }, + { + "epoch": 1.6789041095890411, + "grad_norm": 2.641923427581787, + "learning_rate": 2.47766043866775e-06, + "loss": 1.6338, + "step": 1532 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 2.6234304904937744, + "learning_rate": 2.475629569455727e-06, + "loss": 1.6422, + "step": 1533 + }, + { + "epoch": 1.6810958904109587, + "grad_norm": 2.7153754234313965, + "learning_rate": 2.4735987002437044e-06, + "loss": 1.5758, + "step": 1534 + }, + { + "epoch": 1.6821917808219178, + "grad_norm": 2.6731090545654297, + "learning_rate": 2.4715678310316816e-06, + "loss": 1.5684, + "step": 1535 + }, + { + "epoch": 1.6832876712328768, + "grad_norm": 2.64077091217041, + "learning_rate": 2.469536961819659e-06, + "loss": 1.6039, + "step": 1536 + }, + { + "epoch": 1.6843835616438356, + "grad_norm": 2.522536516189575, + "learning_rate": 2.467506092607636e-06, + "loss": 1.6054, + "step": 1537 + }, + { + "epoch": 1.6854794520547944, + "grad_norm": 2.5827460289001465, + "learning_rate": 2.4654752233956135e-06, + "loss": 1.5705, + "step": 1538 + }, + { + "epoch": 1.6865753424657535, + "grad_norm": 2.7648563385009766, + "learning_rate": 2.4634443541835907e-06, + "loss": 1.6657, + "step": 1539 + }, + { + "epoch": 1.6876712328767123, + "grad_norm": 2.6262803077697754, + "learning_rate": 2.461413484971568e-06, + "loss": 1.6021, + "step": 1540 + }, + { + "epoch": 1.6876712328767123, + "eval_loss": 1.8337559700012207, + "eval_runtime": 37.2695, + "eval_samples_per_second": 26.832, + "eval_steps_per_second": 1.127, + "step": 1540 + }, + { + "epoch": 1.688767123287671, + "grad_norm": 2.735790491104126, + "learning_rate": 2.4593826157595453e-06, + "loss": 1.6836, + "step": 1541 + }, + { + "epoch": 1.6898630136986301, + "grad_norm": 2.601675271987915, + "learning_rate": 2.4573517465475226e-06, + "loss": 1.5497, + "step": 1542 + }, + { + "epoch": 1.6909589041095892, + "grad_norm": 2.7680232524871826, + "learning_rate": 2.4553208773355e-06, + "loss": 1.5773, + "step": 1543 + }, + { + "epoch": 1.692054794520548, + "grad_norm": 2.7184901237487793, + "learning_rate": 2.453290008123477e-06, + "loss": 1.61, + "step": 1544 + }, + { + "epoch": 1.6931506849315068, + "grad_norm": 2.743377685546875, + "learning_rate": 2.4512591389114544e-06, + "loss": 1.6277, + "step": 1545 + }, + { + "epoch": 1.6942465753424658, + "grad_norm": 2.686190605163574, + "learning_rate": 2.4492282696994317e-06, + "loss": 1.6108, + "step": 1546 + }, + { + "epoch": 1.6953424657534246, + "grad_norm": 2.6933157444000244, + "learning_rate": 2.447197400487409e-06, + "loss": 1.5984, + "step": 1547 + }, + { + "epoch": 1.6964383561643834, + "grad_norm": 2.764714479446411, + "learning_rate": 2.4451665312753862e-06, + "loss": 1.6154, + "step": 1548 + }, + { + "epoch": 1.6975342465753425, + "grad_norm": 2.6571526527404785, + "learning_rate": 2.4431356620633635e-06, + "loss": 1.6502, + "step": 1549 + }, + { + "epoch": 1.6986301369863015, + "grad_norm": 2.739225149154663, + "learning_rate": 2.4411047928513408e-06, + "loss": 1.6142, + "step": 1550 + }, + { + "epoch": 1.6986301369863015, + "eval_loss": 1.8304396867752075, + "eval_runtime": 37.271, + "eval_samples_per_second": 26.831, + "eval_steps_per_second": 1.127, + "step": 1550 + }, + { + "epoch": 1.6997260273972603, + "grad_norm": 2.866999864578247, + "learning_rate": 2.439073923639318e-06, + "loss": 1.6155, + "step": 1551 + }, + { + "epoch": 1.7008219178082191, + "grad_norm": 2.670107364654541, + "learning_rate": 2.4370430544272953e-06, + "loss": 1.5526, + "step": 1552 + }, + { + "epoch": 1.7019178082191782, + "grad_norm": 2.5540761947631836, + "learning_rate": 2.435012185215272e-06, + "loss": 1.5501, + "step": 1553 + }, + { + "epoch": 1.703013698630137, + "grad_norm": 2.8047935962677, + "learning_rate": 2.4329813160032495e-06, + "loss": 1.5782, + "step": 1554 + }, + { + "epoch": 1.7041095890410958, + "grad_norm": 2.717963218688965, + "learning_rate": 2.4309504467912267e-06, + "loss": 1.6043, + "step": 1555 + }, + { + "epoch": 1.7052054794520548, + "grad_norm": 2.7823145389556885, + "learning_rate": 2.428919577579204e-06, + "loss": 1.596, + "step": 1556 + }, + { + "epoch": 1.7063013698630138, + "grad_norm": 2.8534443378448486, + "learning_rate": 2.4268887083671813e-06, + "loss": 1.6807, + "step": 1557 + }, + { + "epoch": 1.7073972602739727, + "grad_norm": 2.6131157875061035, + "learning_rate": 2.4248578391551586e-06, + "loss": 1.5833, + "step": 1558 + }, + { + "epoch": 1.7084931506849315, + "grad_norm": 2.761531352996826, + "learning_rate": 2.422826969943136e-06, + "loss": 1.5799, + "step": 1559 + }, + { + "epoch": 1.7095890410958905, + "grad_norm": 2.550750970840454, + "learning_rate": 2.420796100731113e-06, + "loss": 1.5504, + "step": 1560 + }, + { + "epoch": 1.7095890410958905, + "eval_loss": 1.8282228708267212, + "eval_runtime": 37.2167, + "eval_samples_per_second": 26.87, + "eval_steps_per_second": 1.129, + "step": 1560 + }, + { + "epoch": 1.7106849315068493, + "grad_norm": 2.7341554164886475, + "learning_rate": 2.4187652315190904e-06, + "loss": 1.6141, + "step": 1561 + }, + { + "epoch": 1.7117808219178081, + "grad_norm": 2.5879876613616943, + "learning_rate": 2.4167343623070677e-06, + "loss": 1.6121, + "step": 1562 + }, + { + "epoch": 1.7128767123287671, + "grad_norm": 2.626166820526123, + "learning_rate": 2.414703493095045e-06, + "loss": 1.5993, + "step": 1563 + }, + { + "epoch": 1.7139726027397262, + "grad_norm": 2.620342254638672, + "learning_rate": 2.4126726238830222e-06, + "loss": 1.5972, + "step": 1564 + }, + { + "epoch": 1.7150684931506848, + "grad_norm": 2.596414566040039, + "learning_rate": 2.410641754670999e-06, + "loss": 1.5827, + "step": 1565 + }, + { + "epoch": 1.7161643835616438, + "grad_norm": 2.6863791942596436, + "learning_rate": 2.4086108854589763e-06, + "loss": 1.6427, + "step": 1566 + }, + { + "epoch": 1.7172602739726028, + "grad_norm": 2.741095781326294, + "learning_rate": 2.4065800162469536e-06, + "loss": 1.5882, + "step": 1567 + }, + { + "epoch": 1.7183561643835616, + "grad_norm": 2.890286922454834, + "learning_rate": 2.404549147034931e-06, + "loss": 1.6046, + "step": 1568 + }, + { + "epoch": 1.7194520547945205, + "grad_norm": 2.699467658996582, + "learning_rate": 2.402518277822908e-06, + "loss": 1.6395, + "step": 1569 + }, + { + "epoch": 1.7205479452054795, + "grad_norm": 2.6112730503082275, + "learning_rate": 2.4004874086108854e-06, + "loss": 1.5772, + "step": 1570 + }, + { + "epoch": 1.7205479452054795, + "eval_loss": 1.824203372001648, + "eval_runtime": 37.271, + "eval_samples_per_second": 26.831, + "eval_steps_per_second": 1.127, + "step": 1570 + }, + { + "epoch": 1.7216438356164385, + "grad_norm": 2.7761809825897217, + "learning_rate": 2.3984565393988627e-06, + "loss": 1.6177, + "step": 1571 + }, + { + "epoch": 1.7227397260273971, + "grad_norm": 2.691741704940796, + "learning_rate": 2.39642567018684e-06, + "loss": 1.5837, + "step": 1572 + }, + { + "epoch": 1.7238356164383561, + "grad_norm": 2.655440330505371, + "learning_rate": 2.3943948009748173e-06, + "loss": 1.5896, + "step": 1573 + }, + { + "epoch": 1.7249315068493152, + "grad_norm": 2.8032209873199463, + "learning_rate": 2.3923639317627945e-06, + "loss": 1.6267, + "step": 1574 + }, + { + "epoch": 1.726027397260274, + "grad_norm": 2.7980844974517822, + "learning_rate": 2.390333062550772e-06, + "loss": 1.5695, + "step": 1575 + }, + { + "epoch": 1.7271232876712328, + "grad_norm": 2.6170504093170166, + "learning_rate": 2.388302193338749e-06, + "loss": 1.4959, + "step": 1576 + }, + { + "epoch": 1.7282191780821918, + "grad_norm": 2.836310386657715, + "learning_rate": 2.3862713241267264e-06, + "loss": 1.6757, + "step": 1577 + }, + { + "epoch": 1.7293150684931506, + "grad_norm": 2.868952751159668, + "learning_rate": 2.3842404549147037e-06, + "loss": 1.6065, + "step": 1578 + }, + { + "epoch": 1.7304109589041095, + "grad_norm": 2.8427889347076416, + "learning_rate": 2.382209585702681e-06, + "loss": 1.5705, + "step": 1579 + }, + { + "epoch": 1.7315068493150685, + "grad_norm": 2.6732518672943115, + "learning_rate": 2.380178716490658e-06, + "loss": 1.6437, + "step": 1580 + }, + { + "epoch": 1.7315068493150685, + "eval_loss": 1.8212717771530151, + "eval_runtime": 37.3837, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.123, + "step": 1580 + }, + { + "epoch": 1.7326027397260275, + "grad_norm": 2.7127609252929688, + "learning_rate": 2.3781478472786355e-06, + "loss": 1.5995, + "step": 1581 + }, + { + "epoch": 1.7336986301369863, + "grad_norm": 2.822373390197754, + "learning_rate": 2.3761169780666128e-06, + "loss": 1.6416, + "step": 1582 + }, + { + "epoch": 1.7347945205479451, + "grad_norm": 2.7300949096679688, + "learning_rate": 2.37408610885459e-06, + "loss": 1.6504, + "step": 1583 + }, + { + "epoch": 1.7358904109589042, + "grad_norm": 2.6506187915802, + "learning_rate": 2.3720552396425673e-06, + "loss": 1.5873, + "step": 1584 + }, + { + "epoch": 1.736986301369863, + "grad_norm": 2.6960768699645996, + "learning_rate": 2.3700243704305446e-06, + "loss": 1.6743, + "step": 1585 + }, + { + "epoch": 1.7380821917808218, + "grad_norm": 2.640947103500366, + "learning_rate": 2.367993501218522e-06, + "loss": 1.6105, + "step": 1586 + }, + { + "epoch": 1.7391780821917808, + "grad_norm": 2.8700079917907715, + "learning_rate": 2.365962632006499e-06, + "loss": 1.5965, + "step": 1587 + }, + { + "epoch": 1.7402739726027399, + "grad_norm": 2.748927116394043, + "learning_rate": 2.3639317627944764e-06, + "loss": 1.6386, + "step": 1588 + }, + { + "epoch": 1.7413698630136987, + "grad_norm": 2.7105116844177246, + "learning_rate": 2.3619008935824537e-06, + "loss": 1.6392, + "step": 1589 + }, + { + "epoch": 1.7424657534246575, + "grad_norm": 2.6624841690063477, + "learning_rate": 2.3598700243704305e-06, + "loss": 1.6312, + "step": 1590 + }, + { + "epoch": 1.7424657534246575, + "eval_loss": 1.8209728002548218, + "eval_runtime": 37.3533, + "eval_samples_per_second": 26.771, + "eval_steps_per_second": 1.124, + "step": 1590 + }, + { + "epoch": 1.7435616438356165, + "grad_norm": 2.903249740600586, + "learning_rate": 2.357839155158408e-06, + "loss": 1.6263, + "step": 1591 + }, + { + "epoch": 1.7446575342465753, + "grad_norm": 2.792497158050537, + "learning_rate": 2.355808285946385e-06, + "loss": 1.6206, + "step": 1592 + }, + { + "epoch": 1.7457534246575341, + "grad_norm": 2.737065315246582, + "learning_rate": 2.3537774167343624e-06, + "loss": 1.6394, + "step": 1593 + }, + { + "epoch": 1.7468493150684932, + "grad_norm": 2.651978015899658, + "learning_rate": 2.3517465475223396e-06, + "loss": 1.6307, + "step": 1594 + }, + { + "epoch": 1.7479452054794522, + "grad_norm": 2.7566919326782227, + "learning_rate": 2.349715678310317e-06, + "loss": 1.6203, + "step": 1595 + }, + { + "epoch": 1.749041095890411, + "grad_norm": 2.7853057384490967, + "learning_rate": 2.347684809098294e-06, + "loss": 1.594, + "step": 1596 + }, + { + "epoch": 1.7501369863013698, + "grad_norm": 2.770509719848633, + "learning_rate": 2.3456539398862715e-06, + "loss": 1.6491, + "step": 1597 + }, + { + "epoch": 1.7512328767123289, + "grad_norm": 2.718459367752075, + "learning_rate": 2.3436230706742487e-06, + "loss": 1.5671, + "step": 1598 + }, + { + "epoch": 1.7523287671232877, + "grad_norm": 2.724194288253784, + "learning_rate": 2.341592201462226e-06, + "loss": 1.5771, + "step": 1599 + }, + { + "epoch": 1.7534246575342465, + "grad_norm": 2.691155195236206, + "learning_rate": 2.3395613322502033e-06, + "loss": 1.5889, + "step": 1600 + }, + { + "epoch": 1.7534246575342465, + "eval_loss": 1.8217312097549438, + "eval_runtime": 37.3364, + "eval_samples_per_second": 26.784, + "eval_steps_per_second": 1.125, + "step": 1600 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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": 4.987695054979072e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-1600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/latest new file mode 100644 index 0000000000000000000000000000000000000000..ca3fd898b349457b6d1b62649e5bdfc274b799e1 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..cf675ebc17137d730d084d55b9e2fd2f0f06c89d --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/trainer_state.json @@ -0,0 +1,21093 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.958904109589041, + "eval_steps": 10, + "global_step": 2700, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + }, + { + "epoch": 0.6586301369863014, + "grad_norm": 2.6008102893829346, + "learning_rate": 4.364337936636881e-06, + "loss": 1.7848, + "step": 601 + }, + { + "epoch": 0.6597260273972603, + "grad_norm": 2.630127429962158, + "learning_rate": 4.362307067424858e-06, + "loss": 1.7135, + "step": 602 + }, + { + "epoch": 0.6608219178082192, + "grad_norm": 2.720945358276367, + "learning_rate": 4.3602761982128355e-06, + "loss": 1.7897, + "step": 603 + }, + { + "epoch": 0.6619178082191781, + "grad_norm": 2.898836851119995, + "learning_rate": 4.358245329000813e-06, + "loss": 1.7434, + "step": 604 + }, + { + "epoch": 0.663013698630137, + "grad_norm": 2.6548235416412354, + "learning_rate": 4.35621445978879e-06, + "loss": 1.7683, + "step": 605 + }, + { + "epoch": 0.6641095890410958, + "grad_norm": 2.7218399047851562, + "learning_rate": 4.354183590576767e-06, + "loss": 1.7729, + "step": 606 + }, + { + "epoch": 0.6652054794520548, + "grad_norm": 2.5972483158111572, + "learning_rate": 4.352152721364745e-06, + "loss": 1.7517, + "step": 607 + }, + { + "epoch": 0.6663013698630137, + "grad_norm": 2.6741726398468018, + "learning_rate": 4.350121852152722e-06, + "loss": 1.7354, + "step": 608 + }, + { + "epoch": 0.6673972602739726, + "grad_norm": 2.898418426513672, + "learning_rate": 4.348090982940699e-06, + "loss": 1.7843, + "step": 609 + }, + { + "epoch": 0.6684931506849315, + "grad_norm": 2.3964333534240723, + "learning_rate": 4.3460601137286764e-06, + "loss": 1.808, + "step": 610 + }, + { + "epoch": 0.6684931506849315, + "eval_loss": 1.9525212049484253, + "eval_runtime": 36.8297, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 610 + }, + { + "epoch": 0.6695890410958905, + "grad_norm": 2.9952592849731445, + "learning_rate": 4.344029244516654e-06, + "loss": 1.7856, + "step": 611 + }, + { + "epoch": 0.6706849315068493, + "grad_norm": 2.81740140914917, + "learning_rate": 4.341998375304631e-06, + "loss": 1.7613, + "step": 612 + }, + { + "epoch": 0.6717808219178082, + "grad_norm": 2.6271541118621826, + "learning_rate": 4.339967506092607e-06, + "loss": 1.8409, + "step": 613 + }, + { + "epoch": 0.6728767123287671, + "grad_norm": 2.540395498275757, + "learning_rate": 4.3379366368805855e-06, + "loss": 1.7765, + "step": 614 + }, + { + "epoch": 0.673972602739726, + "grad_norm": 2.629131317138672, + "learning_rate": 4.335905767668562e-06, + "loss": 1.7634, + "step": 615 + }, + { + "epoch": 0.675068493150685, + "grad_norm": 2.5796725749969482, + "learning_rate": 4.33387489845654e-06, + "loss": 1.7414, + "step": 616 + }, + { + "epoch": 0.6761643835616439, + "grad_norm": 2.6980209350585938, + "learning_rate": 4.3318440292445165e-06, + "loss": 1.8048, + "step": 617 + }, + { + "epoch": 0.6772602739726027, + "grad_norm": 2.658919095993042, + "learning_rate": 4.329813160032495e-06, + "loss": 1.7725, + "step": 618 + }, + { + "epoch": 0.6783561643835616, + "grad_norm": 2.5237927436828613, + "learning_rate": 4.327782290820471e-06, + "loss": 1.7559, + "step": 619 + }, + { + "epoch": 0.6794520547945205, + "grad_norm": 2.5822298526763916, + "learning_rate": 4.325751421608449e-06, + "loss": 1.7129, + "step": 620 + }, + { + "epoch": 0.6794520547945205, + "eval_loss": 1.9492865800857544, + "eval_runtime": 37.0012, + "eval_samples_per_second": 27.026, + "eval_steps_per_second": 1.135, + "step": 620 + }, + { + "epoch": 0.6805479452054795, + "grad_norm": 2.762810468673706, + "learning_rate": 4.323720552396426e-06, + "loss": 1.7895, + "step": 621 + }, + { + "epoch": 0.6816438356164384, + "grad_norm": 2.48427414894104, + "learning_rate": 4.321689683184404e-06, + "loss": 1.7667, + "step": 622 + }, + { + "epoch": 0.6827397260273973, + "grad_norm": 2.790052890777588, + "learning_rate": 4.31965881397238e-06, + "loss": 1.7548, + "step": 623 + }, + { + "epoch": 0.6838356164383562, + "grad_norm": 2.927957773208618, + "learning_rate": 4.3176279447603574e-06, + "loss": 1.7489, + "step": 624 + }, + { + "epoch": 0.684931506849315, + "grad_norm": 2.4964418411254883, + "learning_rate": 4.315597075548335e-06, + "loss": 1.7947, + "step": 625 + }, + { + "epoch": 0.686027397260274, + "grad_norm": 2.4845147132873535, + "learning_rate": 4.313566206336312e-06, + "loss": 1.7886, + "step": 626 + }, + { + "epoch": 0.6871232876712329, + "grad_norm": 2.687361717224121, + "learning_rate": 4.311535337124289e-06, + "loss": 1.7968, + "step": 627 + }, + { + "epoch": 0.6882191780821918, + "grad_norm": 2.568202018737793, + "learning_rate": 4.3095044679122665e-06, + "loss": 1.7332, + "step": 628 + }, + { + "epoch": 0.6893150684931507, + "grad_norm": 2.496429443359375, + "learning_rate": 4.307473598700244e-06, + "loss": 1.7583, + "step": 629 + }, + { + "epoch": 0.6904109589041096, + "grad_norm": 2.6063425540924072, + "learning_rate": 4.305442729488221e-06, + "loss": 1.779, + "step": 630 + }, + { + "epoch": 0.6904109589041096, + "eval_loss": 1.9421359300613403, + "eval_runtime": 37.0712, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 630 + }, + { + "epoch": 0.6915068493150685, + "grad_norm": 2.5371274948120117, + "learning_rate": 4.303411860276198e-06, + "loss": 1.7133, + "step": 631 + }, + { + "epoch": 0.6926027397260274, + "grad_norm": 2.5601038932800293, + "learning_rate": 4.301380991064176e-06, + "loss": 1.8089, + "step": 632 + }, + { + "epoch": 0.6936986301369863, + "grad_norm": 2.6244118213653564, + "learning_rate": 4.299350121852153e-06, + "loss": 1.7983, + "step": 633 + }, + { + "epoch": 0.6947945205479452, + "grad_norm": 2.4798192977905273, + "learning_rate": 4.29731925264013e-06, + "loss": 1.8221, + "step": 634 + }, + { + "epoch": 0.6958904109589041, + "grad_norm": 2.483376979827881, + "learning_rate": 4.2952883834281075e-06, + "loss": 1.8032, + "step": 635 + }, + { + "epoch": 0.6969863013698631, + "grad_norm": 2.716745615005493, + "learning_rate": 4.293257514216085e-06, + "loss": 1.7583, + "step": 636 + }, + { + "epoch": 0.6980821917808219, + "grad_norm": 2.6685945987701416, + "learning_rate": 4.291226645004062e-06, + "loss": 1.7365, + "step": 637 + }, + { + "epoch": 0.6991780821917808, + "grad_norm": 2.465810537338257, + "learning_rate": 4.289195775792039e-06, + "loss": 1.7682, + "step": 638 + }, + { + "epoch": 0.7002739726027397, + "grad_norm": 2.7629129886627197, + "learning_rate": 4.287164906580017e-06, + "loss": 1.809, + "step": 639 + }, + { + "epoch": 0.7013698630136986, + "grad_norm": 2.5273563861846924, + "learning_rate": 4.285134037367994e-06, + "loss": 1.7464, + "step": 640 + }, + { + "epoch": 0.7013698630136986, + "eval_loss": 1.9359965324401855, + "eval_runtime": 37.1317, + "eval_samples_per_second": 26.931, + "eval_steps_per_second": 1.131, + "step": 640 + }, + { + "epoch": 0.7024657534246576, + "grad_norm": 2.514230966567993, + "learning_rate": 4.283103168155971e-06, + "loss": 1.7365, + "step": 641 + }, + { + "epoch": 0.7035616438356165, + "grad_norm": 2.704042911529541, + "learning_rate": 4.281072298943948e-06, + "loss": 1.7268, + "step": 642 + }, + { + "epoch": 0.7046575342465753, + "grad_norm": 2.4716262817382812, + "learning_rate": 4.279041429731926e-06, + "loss": 1.7212, + "step": 643 + }, + { + "epoch": 0.7057534246575342, + "grad_norm": 2.5695507526397705, + "learning_rate": 4.277010560519903e-06, + "loss": 1.7398, + "step": 644 + }, + { + "epoch": 0.7068493150684931, + "grad_norm": 2.762845039367676, + "learning_rate": 4.27497969130788e-06, + "loss": 1.7158, + "step": 645 + }, + { + "epoch": 0.707945205479452, + "grad_norm": 2.571054220199585, + "learning_rate": 4.2729488220958575e-06, + "loss": 1.7685, + "step": 646 + }, + { + "epoch": 0.709041095890411, + "grad_norm": 2.908184289932251, + "learning_rate": 4.270917952883835e-06, + "loss": 1.7865, + "step": 647 + }, + { + "epoch": 0.7101369863013699, + "grad_norm": 2.705479621887207, + "learning_rate": 4.268887083671812e-06, + "loss": 1.7922, + "step": 648 + }, + { + "epoch": 0.7112328767123288, + "grad_norm": 2.4418110847473145, + "learning_rate": 4.266856214459789e-06, + "loss": 1.7312, + "step": 649 + }, + { + "epoch": 0.7123287671232876, + "grad_norm": 2.4619884490966797, + "learning_rate": 4.264825345247766e-06, + "loss": 1.8343, + "step": 650 + }, + { + "epoch": 0.7123287671232876, + "eval_loss": 1.9352803230285645, + "eval_runtime": 37.0958, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 650 + }, + { + "epoch": 0.7134246575342466, + "grad_norm": 2.5386202335357666, + "learning_rate": 4.262794476035744e-06, + "loss": 1.6902, + "step": 651 + }, + { + "epoch": 0.7145205479452055, + "grad_norm": 2.9012045860290527, + "learning_rate": 4.26076360682372e-06, + "loss": 1.7552, + "step": 652 + }, + { + "epoch": 0.7156164383561644, + "grad_norm": 2.8028385639190674, + "learning_rate": 4.2587327376116984e-06, + "loss": 1.6872, + "step": 653 + }, + { + "epoch": 0.7167123287671233, + "grad_norm": 2.4505386352539062, + "learning_rate": 4.256701868399675e-06, + "loss": 1.75, + "step": 654 + }, + { + "epoch": 0.7178082191780822, + "grad_norm": 2.7551822662353516, + "learning_rate": 4.254670999187653e-06, + "loss": 1.7612, + "step": 655 + }, + { + "epoch": 0.718904109589041, + "grad_norm": 2.596270799636841, + "learning_rate": 4.252640129975629e-06, + "loss": 1.7694, + "step": 656 + }, + { + "epoch": 0.72, + "grad_norm": 2.387662172317505, + "learning_rate": 4.2506092607636075e-06, + "loss": 1.7776, + "step": 657 + }, + { + "epoch": 0.7210958904109589, + "grad_norm": 2.7208738327026367, + "learning_rate": 4.248578391551584e-06, + "loss": 1.7375, + "step": 658 + }, + { + "epoch": 0.7221917808219178, + "grad_norm": 2.5317776203155518, + "learning_rate": 4.246547522339562e-06, + "loss": 1.6922, + "step": 659 + }, + { + "epoch": 0.7232876712328767, + "grad_norm": 2.4803197383880615, + "learning_rate": 4.2445166531275385e-06, + "loss": 1.741, + "step": 660 + }, + { + "epoch": 0.7232876712328767, + "eval_loss": 1.9322715997695923, + "eval_runtime": 37.0962, + "eval_samples_per_second": 26.957, + "eval_steps_per_second": 1.132, + "step": 660 + }, + { + "epoch": 0.7243835616438357, + "grad_norm": 2.5688276290893555, + "learning_rate": 4.242485783915517e-06, + "loss": 1.7518, + "step": 661 + }, + { + "epoch": 0.7254794520547945, + "grad_norm": 2.4652810096740723, + "learning_rate": 4.240454914703493e-06, + "loss": 1.6853, + "step": 662 + }, + { + "epoch": 0.7265753424657534, + "grad_norm": 2.541903257369995, + "learning_rate": 4.23842404549147e-06, + "loss": 1.6841, + "step": 663 + }, + { + "epoch": 0.7276712328767123, + "grad_norm": 2.377857208251953, + "learning_rate": 4.236393176279448e-06, + "loss": 1.7312, + "step": 664 + }, + { + "epoch": 0.7287671232876712, + "grad_norm": 2.418323040008545, + "learning_rate": 4.234362307067425e-06, + "loss": 1.6892, + "step": 665 + }, + { + "epoch": 0.7298630136986302, + "grad_norm": 2.5374410152435303, + "learning_rate": 4.232331437855402e-06, + "loss": 1.7719, + "step": 666 + }, + { + "epoch": 0.7309589041095891, + "grad_norm": 2.7119228839874268, + "learning_rate": 4.2303005686433795e-06, + "loss": 1.7819, + "step": 667 + }, + { + "epoch": 0.7320547945205479, + "grad_norm": 2.5333240032196045, + "learning_rate": 4.228269699431357e-06, + "loss": 1.7758, + "step": 668 + }, + { + "epoch": 0.7331506849315068, + "grad_norm": 2.528696060180664, + "learning_rate": 4.226238830219334e-06, + "loss": 1.75, + "step": 669 + }, + { + "epoch": 0.7342465753424657, + "grad_norm": 2.539039134979248, + "learning_rate": 4.224207961007311e-06, + "loss": 1.7498, + "step": 670 + }, + { + "epoch": 0.7342465753424657, + "eval_loss": 1.934200644493103, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 670 + }, + { + "epoch": 0.7353424657534247, + "grad_norm": 2.5384178161621094, + "learning_rate": 4.2221770917952886e-06, + "loss": 1.7615, + "step": 671 + }, + { + "epoch": 0.7364383561643836, + "grad_norm": 2.601658821105957, + "learning_rate": 4.220146222583266e-06, + "loss": 1.7645, + "step": 672 + }, + { + "epoch": 0.7375342465753425, + "grad_norm": 2.5359764099121094, + "learning_rate": 4.218115353371243e-06, + "loss": 1.7549, + "step": 673 + }, + { + "epoch": 0.7386301369863014, + "grad_norm": 2.641118049621582, + "learning_rate": 4.21608448415922e-06, + "loss": 1.7487, + "step": 674 + }, + { + "epoch": 0.7397260273972602, + "grad_norm": 2.4146382808685303, + "learning_rate": 4.214053614947198e-06, + "loss": 1.7102, + "step": 675 + }, + { + "epoch": 0.7408219178082192, + "grad_norm": 2.4245245456695557, + "learning_rate": 4.212022745735175e-06, + "loss": 1.7842, + "step": 676 + }, + { + "epoch": 0.7419178082191781, + "grad_norm": 2.509824752807617, + "learning_rate": 4.209991876523152e-06, + "loss": 1.6876, + "step": 677 + }, + { + "epoch": 0.743013698630137, + "grad_norm": 2.636444330215454, + "learning_rate": 4.2079610073111295e-06, + "loss": 1.7681, + "step": 678 + }, + { + "epoch": 0.7441095890410959, + "grad_norm": 2.476858615875244, + "learning_rate": 4.205930138099107e-06, + "loss": 1.7049, + "step": 679 + }, + { + "epoch": 0.7452054794520548, + "grad_norm": 2.75732159614563, + "learning_rate": 4.203899268887084e-06, + "loss": 1.7833, + "step": 680 + }, + { + "epoch": 0.7452054794520548, + "eval_loss": 1.931411623954773, + "eval_runtime": 37.0994, + "eval_samples_per_second": 26.955, + "eval_steps_per_second": 1.132, + "step": 680 + }, + { + "epoch": 0.7463013698630137, + "grad_norm": 2.5437941551208496, + "learning_rate": 4.201868399675061e-06, + "loss": 1.7751, + "step": 681 + }, + { + "epoch": 0.7473972602739726, + "grad_norm": 2.5010673999786377, + "learning_rate": 4.199837530463039e-06, + "loss": 1.7715, + "step": 682 + }, + { + "epoch": 0.7484931506849315, + "grad_norm": 2.6352853775024414, + "learning_rate": 4.197806661251016e-06, + "loss": 1.6798, + "step": 683 + }, + { + "epoch": 0.7495890410958904, + "grad_norm": 2.588247776031494, + "learning_rate": 4.195775792038993e-06, + "loss": 1.7575, + "step": 684 + }, + { + "epoch": 0.7506849315068493, + "grad_norm": 2.349921226501465, + "learning_rate": 4.19374492282697e-06, + "loss": 1.7245, + "step": 685 + }, + { + "epoch": 0.7517808219178083, + "grad_norm": 2.394432783126831, + "learning_rate": 4.191714053614948e-06, + "loss": 1.6778, + "step": 686 + }, + { + "epoch": 0.7528767123287671, + "grad_norm": 2.883906602859497, + "learning_rate": 4.189683184402924e-06, + "loss": 1.7701, + "step": 687 + }, + { + "epoch": 0.753972602739726, + "grad_norm": 2.537799596786499, + "learning_rate": 4.187652315190902e-06, + "loss": 1.7286, + "step": 688 + }, + { + "epoch": 0.7550684931506849, + "grad_norm": 2.575700283050537, + "learning_rate": 4.185621445978879e-06, + "loss": 1.7812, + "step": 689 + }, + { + "epoch": 0.7561643835616438, + "grad_norm": 2.592252492904663, + "learning_rate": 4.183590576766857e-06, + "loss": 1.7176, + "step": 690 + }, + { + "epoch": 0.7561643835616438, + "eval_loss": 1.9259952306747437, + "eval_runtime": 37.144, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 690 + }, + { + "epoch": 0.7572602739726028, + "grad_norm": 2.511845111846924, + "learning_rate": 4.181559707554833e-06, + "loss": 1.7564, + "step": 691 + }, + { + "epoch": 0.7583561643835617, + "grad_norm": 2.575446367263794, + "learning_rate": 4.179528838342811e-06, + "loss": 1.7943, + "step": 692 + }, + { + "epoch": 0.7594520547945205, + "grad_norm": 2.379391670227051, + "learning_rate": 4.177497969130788e-06, + "loss": 1.7098, + "step": 693 + }, + { + "epoch": 0.7605479452054794, + "grad_norm": 2.701812982559204, + "learning_rate": 4.175467099918766e-06, + "loss": 1.7292, + "step": 694 + }, + { + "epoch": 0.7616438356164383, + "grad_norm": 2.486640453338623, + "learning_rate": 4.173436230706742e-06, + "loss": 1.7963, + "step": 695 + }, + { + "epoch": 0.7627397260273973, + "grad_norm": 2.561983823776245, + "learning_rate": 4.1714053614947205e-06, + "loss": 1.7153, + "step": 696 + }, + { + "epoch": 0.7638356164383562, + "grad_norm": 2.5709176063537598, + "learning_rate": 4.169374492282697e-06, + "loss": 1.7412, + "step": 697 + }, + { + "epoch": 0.7649315068493151, + "grad_norm": 2.3308839797973633, + "learning_rate": 4.167343623070675e-06, + "loss": 1.7188, + "step": 698 + }, + { + "epoch": 0.766027397260274, + "grad_norm": 2.592470169067383, + "learning_rate": 4.165312753858652e-06, + "loss": 1.7599, + "step": 699 + }, + { + "epoch": 0.7671232876712328, + "grad_norm": 2.3747010231018066, + "learning_rate": 4.163281884646629e-06, + "loss": 1.7361, + "step": 700 + }, + { + "epoch": 0.7671232876712328, + "eval_loss": 1.9229093790054321, + "eval_runtime": 37.1417, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 700 + }, + { + "epoch": 0.7682191780821918, + "grad_norm": 2.4062509536743164, + "learning_rate": 4.161251015434607e-06, + "loss": 1.7256, + "step": 701 + }, + { + "epoch": 0.7693150684931507, + "grad_norm": 2.4951181411743164, + "learning_rate": 4.159220146222583e-06, + "loss": 1.7622, + "step": 702 + }, + { + "epoch": 0.7704109589041096, + "grad_norm": 2.3453004360198975, + "learning_rate": 4.157189277010561e-06, + "loss": 1.6969, + "step": 703 + }, + { + "epoch": 0.7715068493150685, + "grad_norm": 2.382671594619751, + "learning_rate": 4.155158407798538e-06, + "loss": 1.7109, + "step": 704 + }, + { + "epoch": 0.7726027397260274, + "grad_norm": 2.4277336597442627, + "learning_rate": 4.153127538586516e-06, + "loss": 1.7471, + "step": 705 + }, + { + "epoch": 0.7736986301369863, + "grad_norm": 2.4476678371429443, + "learning_rate": 4.151096669374492e-06, + "loss": 1.7483, + "step": 706 + }, + { + "epoch": 0.7747945205479452, + "grad_norm": 2.451099395751953, + "learning_rate": 4.1490658001624705e-06, + "loss": 1.7366, + "step": 707 + }, + { + "epoch": 0.7758904109589041, + "grad_norm": 2.618217706680298, + "learning_rate": 4.147034930950447e-06, + "loss": 1.7282, + "step": 708 + }, + { + "epoch": 0.776986301369863, + "grad_norm": 2.414400100708008, + "learning_rate": 4.145004061738425e-06, + "loss": 1.7409, + "step": 709 + }, + { + "epoch": 0.7780821917808219, + "grad_norm": 2.535353422164917, + "learning_rate": 4.1429731925264015e-06, + "loss": 1.7346, + "step": 710 + }, + { + "epoch": 0.7780821917808219, + "eval_loss": 1.9240931272506714, + "eval_runtime": 36.844, + "eval_samples_per_second": 27.141, + "eval_steps_per_second": 1.14, + "step": 710 + }, + { + "epoch": 0.7791780821917809, + "grad_norm": 2.5075416564941406, + "learning_rate": 4.140942323314379e-06, + "loss": 1.6668, + "step": 711 + }, + { + "epoch": 0.7802739726027397, + "grad_norm": 2.4136979579925537, + "learning_rate": 4.138911454102356e-06, + "loss": 1.696, + "step": 712 + }, + { + "epoch": 0.7813698630136986, + "grad_norm": 2.4363577365875244, + "learning_rate": 4.136880584890333e-06, + "loss": 1.7174, + "step": 713 + }, + { + "epoch": 0.7824657534246575, + "grad_norm": 2.4578073024749756, + "learning_rate": 4.1348497156783106e-06, + "loss": 1.7621, + "step": 714 + }, + { + "epoch": 0.7835616438356164, + "grad_norm": 2.370976686477661, + "learning_rate": 4.132818846466288e-06, + "loss": 1.7443, + "step": 715 + }, + { + "epoch": 0.7846575342465754, + "grad_norm": 2.363755702972412, + "learning_rate": 4.130787977254265e-06, + "loss": 1.7399, + "step": 716 + }, + { + "epoch": 0.7857534246575343, + "grad_norm": 2.44704532623291, + "learning_rate": 4.128757108042242e-06, + "loss": 1.7481, + "step": 717 + }, + { + "epoch": 0.7868493150684932, + "grad_norm": 2.5463340282440186, + "learning_rate": 4.12672623883022e-06, + "loss": 1.7646, + "step": 718 + }, + { + "epoch": 0.787945205479452, + "grad_norm": 2.7088072299957275, + "learning_rate": 4.124695369618197e-06, + "loss": 1.8108, + "step": 719 + }, + { + "epoch": 0.7890410958904109, + "grad_norm": 2.463271141052246, + "learning_rate": 4.122664500406174e-06, + "loss": 1.7311, + "step": 720 + }, + { + "epoch": 0.7890410958904109, + "eval_loss": 1.918715238571167, + "eval_runtime": 37.0468, + "eval_samples_per_second": 26.993, + "eval_steps_per_second": 1.134, + "step": 720 + }, + { + "epoch": 0.7901369863013699, + "grad_norm": 2.6165266036987305, + "learning_rate": 4.1206336311941515e-06, + "loss": 1.7295, + "step": 721 + }, + { + "epoch": 0.7912328767123288, + "grad_norm": 2.3432133197784424, + "learning_rate": 4.118602761982129e-06, + "loss": 1.7512, + "step": 722 + }, + { + "epoch": 0.7923287671232877, + "grad_norm": 2.552751302719116, + "learning_rate": 4.116571892770106e-06, + "loss": 1.6976, + "step": 723 + }, + { + "epoch": 0.7934246575342466, + "grad_norm": 2.7586560249328613, + "learning_rate": 4.114541023558083e-06, + "loss": 1.6708, + "step": 724 + }, + { + "epoch": 0.7945205479452054, + "grad_norm": 2.5326449871063232, + "learning_rate": 4.112510154346061e-06, + "loss": 1.7824, + "step": 725 + }, + { + "epoch": 0.7956164383561644, + "grad_norm": 2.7658956050872803, + "learning_rate": 4.110479285134038e-06, + "loss": 1.7093, + "step": 726 + }, + { + "epoch": 0.7967123287671233, + "grad_norm": 2.6868369579315186, + "learning_rate": 4.108448415922015e-06, + "loss": 1.7475, + "step": 727 + }, + { + "epoch": 0.7978082191780822, + "grad_norm": 2.543748140335083, + "learning_rate": 4.1064175467099924e-06, + "loss": 1.7714, + "step": 728 + }, + { + "epoch": 0.7989041095890411, + "grad_norm": 2.7235589027404785, + "learning_rate": 4.10438667749797e-06, + "loss": 1.7593, + "step": 729 + }, + { + "epoch": 0.8, + "grad_norm": 2.697038173675537, + "learning_rate": 4.102355808285947e-06, + "loss": 1.6296, + "step": 730 + }, + { + "epoch": 0.8, + "eval_loss": 1.91779363155365, + "eval_runtime": 37.1168, + "eval_samples_per_second": 26.942, + "eval_steps_per_second": 1.132, + "step": 730 + }, + { + "epoch": 0.8010958904109589, + "grad_norm": 2.6746180057525635, + "learning_rate": 4.100324939073924e-06, + "loss": 1.7971, + "step": 731 + }, + { + "epoch": 0.8021917808219178, + "grad_norm": 2.6281094551086426, + "learning_rate": 4.0982940698619015e-06, + "loss": 1.7212, + "step": 732 + }, + { + "epoch": 0.8032876712328767, + "grad_norm": 2.446025848388672, + "learning_rate": 4.096263200649879e-06, + "loss": 1.7324, + "step": 733 + }, + { + "epoch": 0.8043835616438356, + "grad_norm": 2.5460824966430664, + "learning_rate": 4.094232331437856e-06, + "loss": 1.7435, + "step": 734 + }, + { + "epoch": 0.8054794520547945, + "grad_norm": 2.5419771671295166, + "learning_rate": 4.092201462225833e-06, + "loss": 1.6979, + "step": 735 + }, + { + "epoch": 0.8065753424657535, + "grad_norm": 2.5545520782470703, + "learning_rate": 4.090170593013811e-06, + "loss": 1.7042, + "step": 736 + }, + { + "epoch": 0.8076712328767123, + "grad_norm": 2.8353254795074463, + "learning_rate": 4.088139723801787e-06, + "loss": 1.7313, + "step": 737 + }, + { + "epoch": 0.8087671232876712, + "grad_norm": 2.6521668434143066, + "learning_rate": 4.086108854589765e-06, + "loss": 1.6901, + "step": 738 + }, + { + "epoch": 0.8098630136986301, + "grad_norm": 2.6611177921295166, + "learning_rate": 4.084077985377742e-06, + "loss": 1.7064, + "step": 739 + }, + { + "epoch": 0.810958904109589, + "grad_norm": 2.411482572555542, + "learning_rate": 4.08204711616572e-06, + "loss": 1.729, + "step": 740 + }, + { + "epoch": 0.810958904109589, + "eval_loss": 1.9176616668701172, + "eval_runtime": 37.1603, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 740 + }, + { + "epoch": 0.812054794520548, + "grad_norm": 2.689110279083252, + "learning_rate": 4.080016246953696e-06, + "loss": 1.7707, + "step": 741 + }, + { + "epoch": 0.8131506849315069, + "grad_norm": 2.544018268585205, + "learning_rate": 4.077985377741674e-06, + "loss": 1.6724, + "step": 742 + }, + { + "epoch": 0.8142465753424658, + "grad_norm": 2.55307936668396, + "learning_rate": 4.075954508529651e-06, + "loss": 1.7677, + "step": 743 + }, + { + "epoch": 0.8153424657534246, + "grad_norm": 2.3896875381469727, + "learning_rate": 4.073923639317629e-06, + "loss": 1.7729, + "step": 744 + }, + { + "epoch": 0.8164383561643835, + "grad_norm": 2.381887197494507, + "learning_rate": 4.071892770105605e-06, + "loss": 1.7737, + "step": 745 + }, + { + "epoch": 0.8175342465753425, + "grad_norm": 2.4873931407928467, + "learning_rate": 4.069861900893583e-06, + "loss": 1.6928, + "step": 746 + }, + { + "epoch": 0.8186301369863014, + "grad_norm": 2.4908432960510254, + "learning_rate": 4.06783103168156e-06, + "loss": 1.6674, + "step": 747 + }, + { + "epoch": 0.8197260273972603, + "grad_norm": 2.4346399307250977, + "learning_rate": 4.065800162469538e-06, + "loss": 1.6947, + "step": 748 + }, + { + "epoch": 0.8208219178082192, + "grad_norm": 2.6459081172943115, + "learning_rate": 4.063769293257514e-06, + "loss": 1.701, + "step": 749 + }, + { + "epoch": 0.821917808219178, + "grad_norm": 2.6092610359191895, + "learning_rate": 4.061738424045492e-06, + "loss": 1.7425, + "step": 750 + }, + { + "epoch": 0.821917808219178, + "eval_loss": 1.919799566268921, + "eval_runtime": 37.1632, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 750 + }, + { + "epoch": 0.823013698630137, + "grad_norm": 2.4443421363830566, + "learning_rate": 4.059707554833469e-06, + "loss": 1.7267, + "step": 751 + }, + { + "epoch": 0.8241095890410959, + "grad_norm": 2.8252694606781006, + "learning_rate": 4.057676685621446e-06, + "loss": 1.7412, + "step": 752 + }, + { + "epoch": 0.8252054794520548, + "grad_norm": 2.3391194343566895, + "learning_rate": 4.0556458164094235e-06, + "loss": 1.7916, + "step": 753 + }, + { + "epoch": 0.8263013698630137, + "grad_norm": 2.57979679107666, + "learning_rate": 4.053614947197401e-06, + "loss": 1.7438, + "step": 754 + }, + { + "epoch": 0.8273972602739726, + "grad_norm": 2.398383855819702, + "learning_rate": 4.051584077985378e-06, + "loss": 1.6606, + "step": 755 + }, + { + "epoch": 0.8284931506849315, + "grad_norm": 2.394498348236084, + "learning_rate": 4.049553208773355e-06, + "loss": 1.736, + "step": 756 + }, + { + "epoch": 0.8295890410958904, + "grad_norm": 2.5070464611053467, + "learning_rate": 4.047522339561333e-06, + "loss": 1.7559, + "step": 757 + }, + { + "epoch": 0.8306849315068493, + "grad_norm": 2.439722776412964, + "learning_rate": 4.04549147034931e-06, + "loss": 1.7133, + "step": 758 + }, + { + "epoch": 0.8317808219178082, + "grad_norm": 2.569152593612671, + "learning_rate": 4.043460601137287e-06, + "loss": 1.7123, + "step": 759 + }, + { + "epoch": 0.8328767123287671, + "grad_norm": 2.637657642364502, + "learning_rate": 4.041429731925264e-06, + "loss": 1.8267, + "step": 760 + }, + { + "epoch": 0.8328767123287671, + "eval_loss": 1.9148659706115723, + "eval_runtime": 37.2248, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 760 + }, + { + "epoch": 0.8339726027397261, + "grad_norm": 2.441269636154175, + "learning_rate": 4.039398862713242e-06, + "loss": 1.7106, + "step": 761 + }, + { + "epoch": 0.8350684931506849, + "grad_norm": 2.582172155380249, + "learning_rate": 4.037367993501219e-06, + "loss": 1.7167, + "step": 762 + }, + { + "epoch": 0.8361643835616438, + "grad_norm": 2.5083518028259277, + "learning_rate": 4.035337124289196e-06, + "loss": 1.8031, + "step": 763 + }, + { + "epoch": 0.8372602739726027, + "grad_norm": 2.504112958908081, + "learning_rate": 4.0333062550771735e-06, + "loss": 1.694, + "step": 764 + }, + { + "epoch": 0.8383561643835616, + "grad_norm": 2.5068695545196533, + "learning_rate": 4.031275385865151e-06, + "loss": 1.6798, + "step": 765 + }, + { + "epoch": 0.8394520547945206, + "grad_norm": 2.600541114807129, + "learning_rate": 4.029244516653128e-06, + "loss": 1.7346, + "step": 766 + }, + { + "epoch": 0.8405479452054795, + "grad_norm": 2.503020763397217, + "learning_rate": 4.027213647441105e-06, + "loss": 1.7507, + "step": 767 + }, + { + "epoch": 0.8416438356164384, + "grad_norm": 2.7774016857147217, + "learning_rate": 4.025182778229083e-06, + "loss": 1.6737, + "step": 768 + }, + { + "epoch": 0.8427397260273972, + "grad_norm": 2.5771589279174805, + "learning_rate": 4.02315190901706e-06, + "loss": 1.6943, + "step": 769 + }, + { + "epoch": 0.8438356164383561, + "grad_norm": 2.3841609954833984, + "learning_rate": 4.021121039805037e-06, + "loss": 1.7142, + "step": 770 + }, + { + "epoch": 0.8438356164383561, + "eval_loss": 1.9146945476531982, + "eval_runtime": 37.1701, + "eval_samples_per_second": 26.903, + "eval_steps_per_second": 1.13, + "step": 770 + }, + { + "epoch": 0.8449315068493151, + "grad_norm": 2.466207981109619, + "learning_rate": 4.0190901705930144e-06, + "loss": 1.7192, + "step": 771 + }, + { + "epoch": 0.846027397260274, + "grad_norm": 2.72037935256958, + "learning_rate": 4.017059301380992e-06, + "loss": 1.8082, + "step": 772 + }, + { + "epoch": 0.8471232876712329, + "grad_norm": 2.36811900138855, + "learning_rate": 4.015028432168969e-06, + "loss": 1.6964, + "step": 773 + }, + { + "epoch": 0.8482191780821918, + "grad_norm": 2.4803504943847656, + "learning_rate": 4.012997562956945e-06, + "loss": 1.779, + "step": 774 + }, + { + "epoch": 0.8493150684931506, + "grad_norm": 2.707174777984619, + "learning_rate": 4.0109666937449235e-06, + "loss": 1.6715, + "step": 775 + }, + { + "epoch": 0.8504109589041096, + "grad_norm": 2.462543249130249, + "learning_rate": 4.0089358245329e-06, + "loss": 1.7117, + "step": 776 + }, + { + "epoch": 0.8515068493150685, + "grad_norm": 2.481444835662842, + "learning_rate": 4.006904955320878e-06, + "loss": 1.7007, + "step": 777 + }, + { + "epoch": 0.8526027397260274, + "grad_norm": 2.4432222843170166, + "learning_rate": 4.0048740861088545e-06, + "loss": 1.703, + "step": 778 + }, + { + "epoch": 0.8536986301369863, + "grad_norm": 2.611387014389038, + "learning_rate": 4.002843216896833e-06, + "loss": 1.7722, + "step": 779 + }, + { + "epoch": 0.8547945205479452, + "grad_norm": 2.456305742263794, + "learning_rate": 4.000812347684809e-06, + "loss": 1.7615, + "step": 780 + }, + { + "epoch": 0.8547945205479452, + "eval_loss": 1.9089828729629517, + "eval_runtime": 37.1812, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 780 + }, + { + "epoch": 0.8558904109589041, + "grad_norm": 2.4287772178649902, + "learning_rate": 3.998781478472787e-06, + "loss": 1.7098, + "step": 781 + }, + { + "epoch": 0.856986301369863, + "grad_norm": 2.5659544467926025, + "learning_rate": 3.996750609260764e-06, + "loss": 1.7143, + "step": 782 + }, + { + "epoch": 0.8580821917808219, + "grad_norm": 2.467067003250122, + "learning_rate": 3.994719740048742e-06, + "loss": 1.6668, + "step": 783 + }, + { + "epoch": 0.8591780821917808, + "grad_norm": 2.5660688877105713, + "learning_rate": 3.992688870836718e-06, + "loss": 1.8185, + "step": 784 + }, + { + "epoch": 0.8602739726027397, + "grad_norm": 2.6185719966888428, + "learning_rate": 3.990658001624696e-06, + "loss": 1.7356, + "step": 785 + }, + { + "epoch": 0.8613698630136987, + "grad_norm": 2.5389695167541504, + "learning_rate": 3.988627132412673e-06, + "loss": 1.6973, + "step": 786 + }, + { + "epoch": 0.8624657534246576, + "grad_norm": 2.499234914779663, + "learning_rate": 3.98659626320065e-06, + "loss": 1.7539, + "step": 787 + }, + { + "epoch": 0.8635616438356164, + "grad_norm": 2.6246273517608643, + "learning_rate": 3.984565393988627e-06, + "loss": 1.7055, + "step": 788 + }, + { + "epoch": 0.8646575342465753, + "grad_norm": 2.6127917766571045, + "learning_rate": 3.9825345247766046e-06, + "loss": 1.7177, + "step": 789 + }, + { + "epoch": 0.8657534246575342, + "grad_norm": 2.4941787719726562, + "learning_rate": 3.980503655564582e-06, + "loss": 1.728, + "step": 790 + }, + { + "epoch": 0.8657534246575342, + "eval_loss": 1.9063301086425781, + "eval_runtime": 37.3284, + "eval_samples_per_second": 26.789, + "eval_steps_per_second": 1.125, + "step": 790 + }, + { + "epoch": 0.8668493150684932, + "grad_norm": 2.4694740772247314, + "learning_rate": 3.978472786352559e-06, + "loss": 1.7384, + "step": 791 + }, + { + "epoch": 0.8679452054794521, + "grad_norm": 2.8288261890411377, + "learning_rate": 3.976441917140536e-06, + "loss": 1.7409, + "step": 792 + }, + { + "epoch": 0.869041095890411, + "grad_norm": 2.518359661102295, + "learning_rate": 3.974411047928514e-06, + "loss": 1.714, + "step": 793 + }, + { + "epoch": 0.8701369863013698, + "grad_norm": 2.4874095916748047, + "learning_rate": 3.972380178716491e-06, + "loss": 1.781, + "step": 794 + }, + { + "epoch": 0.8712328767123287, + "grad_norm": 2.523772716522217, + "learning_rate": 3.970349309504468e-06, + "loss": 1.7474, + "step": 795 + }, + { + "epoch": 0.8723287671232877, + "grad_norm": 2.406132459640503, + "learning_rate": 3.9683184402924455e-06, + "loss": 1.7486, + "step": 796 + }, + { + "epoch": 0.8734246575342466, + "grad_norm": 2.388207197189331, + "learning_rate": 3.966287571080423e-06, + "loss": 1.67, + "step": 797 + }, + { + "epoch": 0.8745205479452055, + "grad_norm": 2.5564754009246826, + "learning_rate": 3.9642567018684e-06, + "loss": 1.6803, + "step": 798 + }, + { + "epoch": 0.8756164383561644, + "grad_norm": 2.531890869140625, + "learning_rate": 3.962225832656377e-06, + "loss": 1.7368, + "step": 799 + }, + { + "epoch": 0.8767123287671232, + "grad_norm": 2.404284954071045, + "learning_rate": 3.960194963444355e-06, + "loss": 1.7247, + "step": 800 + }, + { + "epoch": 0.8767123287671232, + "eval_loss": 1.9008547067642212, + "eval_runtime": 37.1962, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 800 + }, + { + "epoch": 0.8778082191780822, + "grad_norm": 2.3903231620788574, + "learning_rate": 3.958164094232332e-06, + "loss": 1.6465, + "step": 801 + }, + { + "epoch": 0.8789041095890411, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7126, + "step": 802 + }, + { + "epoch": 0.88, + "grad_norm": 2.4444072246551514, + "learning_rate": 3.956133225020309e-06, + "loss": 1.7174, + "step": 803 + }, + { + "epoch": 0.8810958904109589, + "grad_norm": 2.6466639041900635, + "learning_rate": 3.954102355808286e-06, + "loss": 1.7166, + "step": 804 + }, + { + "epoch": 0.8821917808219178, + "grad_norm": 2.530284881591797, + "learning_rate": 3.952071486596264e-06, + "loss": 1.7155, + "step": 805 + }, + { + "epoch": 0.8832876712328767, + "grad_norm": 2.601516008377075, + "learning_rate": 3.950040617384241e-06, + "loss": 1.7157, + "step": 806 + }, + { + "epoch": 0.8843835616438356, + "grad_norm": 2.6146240234375, + "learning_rate": 3.948009748172218e-06, + "loss": 1.6931, + "step": 807 + }, + { + "epoch": 0.8854794520547945, + "grad_norm": 2.4904634952545166, + "learning_rate": 3.9459788789601955e-06, + "loss": 1.7594, + "step": 808 + }, + { + "epoch": 0.8865753424657534, + "grad_norm": 2.345689296722412, + "learning_rate": 3.943948009748173e-06, + "loss": 1.7175, + "step": 809 + }, + { + "epoch": 0.8876712328767123, + "grad_norm": 2.4469082355499268, + "learning_rate": 3.94191714053615e-06, + "loss": 1.7161, + "step": 810 + }, + { + "epoch": 0.8876712328767123, + "eval_loss": 1.9021154642105103, + "eval_runtime": 36.8772, + "eval_samples_per_second": 27.117, + "eval_steps_per_second": 1.139, + "step": 810 + }, + { + "epoch": 0.8887671232876713, + "grad_norm": 2.473752737045288, + "learning_rate": 3.939886271324127e-06, + "loss": 1.7688, + "step": 811 + }, + { + "epoch": 0.8898630136986302, + "grad_norm": 2.461033582687378, + "learning_rate": 3.937855402112104e-06, + "loss": 1.6752, + "step": 812 + }, + { + "epoch": 0.890958904109589, + "grad_norm": 2.419179916381836, + "learning_rate": 3.935824532900082e-06, + "loss": 1.6685, + "step": 813 + }, + { + "epoch": 0.8920547945205479, + "grad_norm": 2.543102502822876, + "learning_rate": 3.933793663688058e-06, + "loss": 1.7348, + "step": 814 + }, + { + "epoch": 0.8931506849315068, + "grad_norm": 2.4932174682617188, + "learning_rate": 3.9317627944760365e-06, + "loss": 1.7572, + "step": 815 + }, + { + "epoch": 0.8942465753424658, + "grad_norm": 2.4563698768615723, + "learning_rate": 3.929731925264013e-06, + "loss": 1.6691, + "step": 816 + }, + { + "epoch": 0.8953424657534247, + "grad_norm": 2.566558837890625, + "learning_rate": 3.927701056051991e-06, + "loss": 1.697, + "step": 817 + }, + { + "epoch": 0.8964383561643836, + "grad_norm": 2.5550761222839355, + "learning_rate": 3.9256701868399674e-06, + "loss": 1.7194, + "step": 818 + }, + { + "epoch": 0.8975342465753424, + "grad_norm": 2.51362681388855, + "learning_rate": 3.9236393176279456e-06, + "loss": 1.6804, + "step": 819 + }, + { + "epoch": 0.8986301369863013, + "grad_norm": 2.448714017868042, + "learning_rate": 3.921608448415922e-06, + "loss": 1.6575, + "step": 820 + }, + { + "epoch": 0.8986301369863013, + "eval_loss": 1.904768943786621, + "eval_runtime": 37.0644, + "eval_samples_per_second": 26.98, + "eval_steps_per_second": 1.133, + "step": 820 + }, + { + "epoch": 0.8997260273972603, + "grad_norm": 2.582794666290283, + "learning_rate": 3.9195775792039e-06, + "loss": 1.717, + "step": 821 + }, + { + "epoch": 0.9008219178082192, + "grad_norm": 2.613327980041504, + "learning_rate": 3.9175467099918765e-06, + "loss": 1.6927, + "step": 822 + }, + { + "epoch": 0.9019178082191781, + "grad_norm": 2.522038221359253, + "learning_rate": 3.915515840779855e-06, + "loss": 1.7391, + "step": 823 + }, + { + "epoch": 0.903013698630137, + "grad_norm": 2.372134208679199, + "learning_rate": 3.913484971567831e-06, + "loss": 1.7448, + "step": 824 + }, + { + "epoch": 0.9041095890410958, + "grad_norm": 2.246591329574585, + "learning_rate": 3.911454102355808e-06, + "loss": 1.6531, + "step": 825 + }, + { + "epoch": 0.9052054794520548, + "grad_norm": 2.48276948928833, + "learning_rate": 3.909423233143786e-06, + "loss": 1.7268, + "step": 826 + }, + { + "epoch": 0.9063013698630137, + "grad_norm": 2.7920989990234375, + "learning_rate": 3.907392363931763e-06, + "loss": 1.7155, + "step": 827 + }, + { + "epoch": 0.9073972602739726, + "grad_norm": 2.5430562496185303, + "learning_rate": 3.90536149471974e-06, + "loss": 1.7866, + "step": 828 + }, + { + "epoch": 0.9084931506849315, + "grad_norm": 2.5819969177246094, + "learning_rate": 3.9033306255077175e-06, + "loss": 1.8024, + "step": 829 + }, + { + "epoch": 0.9095890410958904, + "grad_norm": 2.726024866104126, + "learning_rate": 3.901299756295695e-06, + "loss": 1.8467, + "step": 830 + }, + { + "epoch": 0.9095890410958904, + "eval_loss": 1.899052381515503, + "eval_runtime": 37.192, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 830 + }, + { + "epoch": 0.9106849315068493, + "grad_norm": 2.6913323402404785, + "learning_rate": 3.899268887083672e-06, + "loss": 1.6976, + "step": 831 + }, + { + "epoch": 0.9117808219178082, + "grad_norm": 2.5307703018188477, + "learning_rate": 3.897238017871649e-06, + "loss": 1.7072, + "step": 832 + }, + { + "epoch": 0.9128767123287671, + "grad_norm": 2.4976935386657715, + "learning_rate": 3.8952071486596266e-06, + "loss": 1.7183, + "step": 833 + }, + { + "epoch": 0.913972602739726, + "grad_norm": 2.435443878173828, + "learning_rate": 3.893176279447604e-06, + "loss": 1.6845, + "step": 834 + }, + { + "epoch": 0.915068493150685, + "grad_norm": 2.704554319381714, + "learning_rate": 3.891145410235581e-06, + "loss": 1.78, + "step": 835 + }, + { + "epoch": 0.9161643835616439, + "grad_norm": 2.4770138263702393, + "learning_rate": 3.889114541023558e-06, + "loss": 1.7436, + "step": 836 + }, + { + "epoch": 0.9172602739726028, + "grad_norm": 2.472780227661133, + "learning_rate": 3.887083671811536e-06, + "loss": 1.7007, + "step": 837 + }, + { + "epoch": 0.9183561643835616, + "grad_norm": 2.4693336486816406, + "learning_rate": 3.885052802599513e-06, + "loss": 1.6487, + "step": 838 + }, + { + "epoch": 0.9194520547945205, + "grad_norm": 2.458834648132324, + "learning_rate": 3.88302193338749e-06, + "loss": 1.7895, + "step": 839 + }, + { + "epoch": 0.9205479452054794, + "grad_norm": 2.4430108070373535, + "learning_rate": 3.8809910641754675e-06, + "loss": 1.72, + "step": 840 + }, + { + "epoch": 0.9205479452054794, + "eval_loss": 1.9009006023406982, + "eval_runtime": 37.1906, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 840 + }, + { + "epoch": 0.9216438356164384, + "grad_norm": 2.6104867458343506, + "learning_rate": 3.878960194963445e-06, + "loss": 1.6938, + "step": 841 + }, + { + "epoch": 0.9227397260273973, + "grad_norm": 2.5191142559051514, + "learning_rate": 3.876929325751422e-06, + "loss": 1.7528, + "step": 842 + }, + { + "epoch": 0.9238356164383562, + "grad_norm": 2.3788414001464844, + "learning_rate": 3.874898456539399e-06, + "loss": 1.6446, + "step": 843 + }, + { + "epoch": 0.924931506849315, + "grad_norm": 2.485631227493286, + "learning_rate": 3.872867587327377e-06, + "loss": 1.7087, + "step": 844 + }, + { + "epoch": 0.9260273972602739, + "grad_norm": 2.3721091747283936, + "learning_rate": 3.870836718115354e-06, + "loss": 1.6752, + "step": 845 + }, + { + "epoch": 0.9271232876712329, + "grad_norm": 2.4648489952087402, + "learning_rate": 3.868805848903331e-06, + "loss": 1.6883, + "step": 846 + }, + { + "epoch": 0.9282191780821918, + "grad_norm": 2.447054862976074, + "learning_rate": 3.8667749796913084e-06, + "loss": 1.7109, + "step": 847 + }, + { + "epoch": 0.9293150684931507, + "grad_norm": 2.70007061958313, + "learning_rate": 3.864744110479286e-06, + "loss": 1.7342, + "step": 848 + }, + { + "epoch": 0.9304109589041096, + "grad_norm": 2.6336557865142822, + "learning_rate": 3.862713241267262e-06, + "loss": 1.7308, + "step": 849 + }, + { + "epoch": 0.9315068493150684, + "grad_norm": 2.807671070098877, + "learning_rate": 3.86068237205524e-06, + "loss": 1.7519, + "step": 850 + }, + { + "epoch": 0.9315068493150684, + "eval_loss": 1.8990603685379028, + "eval_runtime": 37.1643, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 850 + }, + { + "epoch": 0.9326027397260274, + "grad_norm": 2.7177038192749023, + "learning_rate": 3.858651502843217e-06, + "loss": 1.7065, + "step": 851 + }, + { + "epoch": 0.9336986301369863, + "grad_norm": 2.570364236831665, + "learning_rate": 3.856620633631195e-06, + "loss": 1.6529, + "step": 852 + }, + { + "epoch": 0.9347945205479452, + "grad_norm": 2.590329170227051, + "learning_rate": 3.854589764419171e-06, + "loss": 1.6955, + "step": 853 + }, + { + "epoch": 0.9358904109589041, + "grad_norm": 2.542722702026367, + "learning_rate": 3.852558895207149e-06, + "loss": 1.7803, + "step": 854 + }, + { + "epoch": 0.936986301369863, + "grad_norm": 2.6095776557922363, + "learning_rate": 3.850528025995126e-06, + "loss": 1.7267, + "step": 855 + }, + { + "epoch": 0.938082191780822, + "grad_norm": 2.550161838531494, + "learning_rate": 3.848497156783104e-06, + "loss": 1.6767, + "step": 856 + }, + { + "epoch": 0.9391780821917808, + "grad_norm": 2.5079615116119385, + "learning_rate": 3.84646628757108e-06, + "loss": 1.6897, + "step": 857 + }, + { + "epoch": 0.9402739726027397, + "grad_norm": 2.5113067626953125, + "learning_rate": 3.8444354183590585e-06, + "loss": 1.7024, + "step": 858 + }, + { + "epoch": 0.9413698630136986, + "grad_norm": 2.559811592102051, + "learning_rate": 3.842404549147035e-06, + "loss": 1.695, + "step": 859 + }, + { + "epoch": 0.9424657534246575, + "grad_norm": 2.507901191711426, + "learning_rate": 3.840373679935013e-06, + "loss": 1.661, + "step": 860 + }, + { + "epoch": 0.9424657534246575, + "eval_loss": 1.8952044248580933, + "eval_runtime": 37.2303, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 860 + }, + { + "epoch": 0.9435616438356165, + "grad_norm": 2.288231134414673, + "learning_rate": 3.8383428107229894e-06, + "loss": 1.7095, + "step": 861 + }, + { + "epoch": 0.9446575342465754, + "grad_norm": 2.260040044784546, + "learning_rate": 3.836311941510967e-06, + "loss": 1.6858, + "step": 862 + }, + { + "epoch": 0.9457534246575342, + "grad_norm": 2.5524179935455322, + "learning_rate": 3.834281072298944e-06, + "loss": 1.7362, + "step": 863 + }, + { + "epoch": 0.9468493150684931, + "grad_norm": 2.419013738632202, + "learning_rate": 3.832250203086921e-06, + "loss": 1.6803, + "step": 864 + }, + { + "epoch": 0.947945205479452, + "grad_norm": 2.5345773696899414, + "learning_rate": 3.8302193338748985e-06, + "loss": 1.7435, + "step": 865 + }, + { + "epoch": 0.949041095890411, + "grad_norm": 2.5771822929382324, + "learning_rate": 3.828188464662876e-06, + "loss": 1.7932, + "step": 866 + }, + { + "epoch": 0.9501369863013699, + "grad_norm": 2.4880053997039795, + "learning_rate": 3.826157595450853e-06, + "loss": 1.711, + "step": 867 + }, + { + "epoch": 0.9512328767123288, + "grad_norm": 2.596400022506714, + "learning_rate": 3.82412672623883e-06, + "loss": 1.6569, + "step": 868 + }, + { + "epoch": 0.9523287671232876, + "grad_norm": 2.620776891708374, + "learning_rate": 3.822095857026808e-06, + "loss": 1.7159, + "step": 869 + }, + { + "epoch": 0.9534246575342465, + "grad_norm": 2.512397527694702, + "learning_rate": 3.820064987814785e-06, + "loss": 1.6832, + "step": 870 + }, + { + "epoch": 0.9534246575342465, + "eval_loss": 1.8950546979904175, + "eval_runtime": 37.1526, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 870 + }, + { + "epoch": 0.9545205479452055, + "grad_norm": 2.5925326347351074, + "learning_rate": 3.818034118602762e-06, + "loss": 1.6977, + "step": 871 + }, + { + "epoch": 0.9556164383561644, + "grad_norm": 2.3980817794799805, + "learning_rate": 3.8160032493907395e-06, + "loss": 1.6997, + "step": 872 + }, + { + "epoch": 0.9567123287671233, + "grad_norm": 2.4825663566589355, + "learning_rate": 3.813972380178717e-06, + "loss": 1.7063, + "step": 873 + }, + { + "epoch": 0.9578082191780822, + "grad_norm": 2.535006284713745, + "learning_rate": 3.811941510966694e-06, + "loss": 1.7457, + "step": 874 + }, + { + "epoch": 0.958904109589041, + "grad_norm": 2.450657606124878, + "learning_rate": 3.809910641754671e-06, + "loss": 1.6521, + "step": 875 + }, + { + "epoch": 0.96, + "grad_norm": 2.3942458629608154, + "learning_rate": 3.8078797725426486e-06, + "loss": 1.6194, + "step": 876 + }, + { + "epoch": 0.9610958904109589, + "grad_norm": 2.5781140327453613, + "learning_rate": 3.8058489033306254e-06, + "loss": 1.7105, + "step": 877 + }, + { + "epoch": 0.9621917808219178, + "grad_norm": 2.5196564197540283, + "learning_rate": 3.803818034118603e-06, + "loss": 1.6741, + "step": 878 + }, + { + "epoch": 0.9632876712328767, + "grad_norm": 2.4819812774658203, + "learning_rate": 3.80178716490658e-06, + "loss": 1.7559, + "step": 879 + }, + { + "epoch": 0.9643835616438357, + "grad_norm": 2.6393206119537354, + "learning_rate": 3.7997562956945577e-06, + "loss": 1.7421, + "step": 880 + }, + { + "epoch": 0.9643835616438357, + "eval_loss": 1.8945845365524292, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 880 + }, + { + "epoch": 0.9654794520547946, + "grad_norm": 2.5286684036254883, + "learning_rate": 3.7977254264825345e-06, + "loss": 1.6933, + "step": 881 + }, + { + "epoch": 0.9665753424657534, + "grad_norm": 2.690138339996338, + "learning_rate": 3.7956945572705122e-06, + "loss": 1.6804, + "step": 882 + }, + { + "epoch": 0.9676712328767123, + "grad_norm": 2.7789907455444336, + "learning_rate": 3.793663688058489e-06, + "loss": 1.7839, + "step": 883 + }, + { + "epoch": 0.9687671232876712, + "grad_norm": 2.5196211338043213, + "learning_rate": 3.791632818846467e-06, + "loss": 1.7564, + "step": 884 + }, + { + "epoch": 0.9698630136986301, + "grad_norm": 2.550898790359497, + "learning_rate": 3.7896019496344436e-06, + "loss": 1.758, + "step": 885 + }, + { + "epoch": 0.9709589041095891, + "grad_norm": 2.617361545562744, + "learning_rate": 3.787571080422421e-06, + "loss": 1.7497, + "step": 886 + }, + { + "epoch": 0.972054794520548, + "grad_norm": 2.533339500427246, + "learning_rate": 3.785540211210398e-06, + "loss": 1.7336, + "step": 887 + }, + { + "epoch": 0.9731506849315068, + "grad_norm": 2.475037097930908, + "learning_rate": 3.7835093419983755e-06, + "loss": 1.712, + "step": 888 + }, + { + "epoch": 0.9742465753424657, + "grad_norm": 2.485506534576416, + "learning_rate": 3.7814784727863527e-06, + "loss": 1.6985, + "step": 889 + }, + { + "epoch": 0.9753424657534246, + "grad_norm": 2.712912082672119, + "learning_rate": 3.77944760357433e-06, + "loss": 1.812, + "step": 890 + }, + { + "epoch": 0.9753424657534246, + "eval_loss": 1.8921257257461548, + "eval_runtime": 37.1309, + "eval_samples_per_second": 26.932, + "eval_steps_per_second": 1.131, + "step": 890 + }, + { + "epoch": 0.9764383561643836, + "grad_norm": 2.6280291080474854, + "learning_rate": 3.7774167343623073e-06, + "loss": 1.7504, + "step": 891 + }, + { + "epoch": 0.9775342465753425, + "grad_norm": 2.6162853240966797, + "learning_rate": 3.7753858651502846e-06, + "loss": 1.6793, + "step": 892 + }, + { + "epoch": 0.9786301369863014, + "grad_norm": 2.665316581726074, + "learning_rate": 3.773354995938262e-06, + "loss": 1.7333, + "step": 893 + }, + { + "epoch": 0.9797260273972602, + "grad_norm": 2.531297445297241, + "learning_rate": 3.771324126726239e-06, + "loss": 1.7298, + "step": 894 + }, + { + "epoch": 0.9808219178082191, + "grad_norm": 2.543412446975708, + "learning_rate": 3.7692932575142164e-06, + "loss": 1.7677, + "step": 895 + }, + { + "epoch": 0.9819178082191781, + "grad_norm": 2.3839635848999023, + "learning_rate": 3.7672623883021937e-06, + "loss": 1.7039, + "step": 896 + }, + { + "epoch": 0.983013698630137, + "grad_norm": 2.3417305946350098, + "learning_rate": 3.765231519090171e-06, + "loss": 1.6206, + "step": 897 + }, + { + "epoch": 0.9841095890410959, + "grad_norm": 2.4434802532196045, + "learning_rate": 3.7632006498781482e-06, + "loss": 1.713, + "step": 898 + }, + { + "epoch": 0.9852054794520548, + "grad_norm": 2.5223472118377686, + "learning_rate": 3.761169780666125e-06, + "loss": 1.6587, + "step": 899 + }, + { + "epoch": 0.9863013698630136, + "grad_norm": 2.577479124069214, + "learning_rate": 3.7591389114541028e-06, + "loss": 1.7032, + "step": 900 + }, + { + "epoch": 0.9863013698630136, + "eval_loss": 1.8891643285751343, + "eval_runtime": 37.1123, + "eval_samples_per_second": 26.945, + "eval_steps_per_second": 1.132, + "step": 900 + }, + { + "epoch": 0.9873972602739726, + "grad_norm": 2.468871593475342, + "learning_rate": 3.7571080422420796e-06, + "loss": 1.732, + "step": 901 + }, + { + "epoch": 0.9884931506849315, + "grad_norm": 2.5907723903656006, + "learning_rate": 3.7550771730300573e-06, + "loss": 1.6866, + "step": 902 + }, + { + "epoch": 0.9895890410958904, + "grad_norm": 2.4454426765441895, + "learning_rate": 3.753046303818034e-06, + "loss": 1.746, + "step": 903 + }, + { + "epoch": 0.9906849315068493, + "grad_norm": 2.4651715755462646, + "learning_rate": 3.751015434606012e-06, + "loss": 1.7375, + "step": 904 + }, + { + "epoch": 0.9917808219178083, + "grad_norm": 2.6342053413391113, + "learning_rate": 3.7489845653939887e-06, + "loss": 1.7693, + "step": 905 + }, + { + "epoch": 0.9928767123287672, + "grad_norm": 2.454977512359619, + "learning_rate": 3.7469536961819664e-06, + "loss": 1.6796, + "step": 906 + }, + { + "epoch": 0.993972602739726, + "grad_norm": 2.440626859664917, + "learning_rate": 3.7449228269699433e-06, + "loss": 1.7075, + "step": 907 + }, + { + "epoch": 0.9950684931506849, + "grad_norm": 2.4905781745910645, + "learning_rate": 3.742891957757921e-06, + "loss": 1.7422, + "step": 908 + }, + { + "epoch": 0.9961643835616438, + "grad_norm": 2.6843647956848145, + "learning_rate": 3.740861088545898e-06, + "loss": 1.7107, + "step": 909 + }, + { + "epoch": 0.9972602739726028, + "grad_norm": 2.5906362533569336, + "learning_rate": 3.7388302193338755e-06, + "loss": 1.7545, + "step": 910 + }, + { + "epoch": 0.9972602739726028, + "eval_loss": 1.8889163732528687, + "eval_runtime": 36.8136, + "eval_samples_per_second": 27.164, + "eval_steps_per_second": 1.141, + "step": 910 + }, + { + "epoch": 0.9983561643835617, + "grad_norm": 2.5185422897338867, + "learning_rate": 3.7367993501218524e-06, + "loss": 1.6857, + "step": 911 + }, + { + "epoch": 0.9994520547945206, + "grad_norm": 2.4425783157348633, + "learning_rate": 3.7347684809098297e-06, + "loss": 1.687, + "step": 912 + }, + { + "epoch": 1.0005479452054795, + "grad_norm": 2.5200576782226562, + "learning_rate": 3.732737611697807e-06, + "loss": 1.643, + "step": 913 + }, + { + "epoch": 1.0016438356164383, + "grad_norm": 2.4664227962493896, + "learning_rate": 3.7307067424857842e-06, + "loss": 1.5965, + "step": 914 + }, + { + "epoch": 1.0027397260273974, + "grad_norm": 2.495962381362915, + "learning_rate": 3.7286758732737615e-06, + "loss": 1.6663, + "step": 915 + }, + { + "epoch": 1.0038356164383562, + "grad_norm": 2.5607051849365234, + "learning_rate": 3.7266450040617388e-06, + "loss": 1.6388, + "step": 916 + }, + { + "epoch": 1.004931506849315, + "grad_norm": 2.9295718669891357, + "learning_rate": 3.724614134849716e-06, + "loss": 1.7541, + "step": 917 + }, + { + "epoch": 1.006027397260274, + "grad_norm": 2.5632426738739014, + "learning_rate": 3.7225832656376933e-06, + "loss": 1.6359, + "step": 918 + }, + { + "epoch": 1.0071232876712328, + "grad_norm": 2.6059279441833496, + "learning_rate": 3.7205523964256706e-06, + "loss": 1.7427, + "step": 919 + }, + { + "epoch": 1.0082191780821919, + "grad_norm": 2.6244208812713623, + "learning_rate": 3.718521527213648e-06, + "loss": 1.7097, + "step": 920 + }, + { + "epoch": 1.0082191780821919, + "eval_loss": 1.8882938623428345, + "eval_runtime": 36.9853, + "eval_samples_per_second": 27.038, + "eval_steps_per_second": 1.136, + "step": 920 + }, + { + "epoch": 1.0093150684931507, + "grad_norm": 2.434684991836548, + "learning_rate": 3.716490658001625e-06, + "loss": 1.6875, + "step": 921 + }, + { + "epoch": 1.0104109589041095, + "grad_norm": 2.6481709480285645, + "learning_rate": 3.7144597887896024e-06, + "loss": 1.5726, + "step": 922 + }, + { + "epoch": 1.0115068493150685, + "grad_norm": 2.5398213863372803, + "learning_rate": 3.7124289195775797e-06, + "loss": 1.6881, + "step": 923 + }, + { + "epoch": 1.0126027397260273, + "grad_norm": 2.684122085571289, + "learning_rate": 3.710398050365557e-06, + "loss": 1.691, + "step": 924 + }, + { + "epoch": 1.0136986301369864, + "grad_norm": 2.712214946746826, + "learning_rate": 3.708367181153534e-06, + "loss": 1.653, + "step": 925 + }, + { + "epoch": 1.0147945205479452, + "grad_norm": 2.8427517414093018, + "learning_rate": 3.7063363119415115e-06, + "loss": 1.6853, + "step": 926 + }, + { + "epoch": 1.0158904109589042, + "grad_norm": 2.676410436630249, + "learning_rate": 3.7043054427294884e-06, + "loss": 1.645, + "step": 927 + }, + { + "epoch": 1.016986301369863, + "grad_norm": 2.5397610664367676, + "learning_rate": 3.702274573517466e-06, + "loss": 1.6171, + "step": 928 + }, + { + "epoch": 1.0180821917808218, + "grad_norm": 2.606250286102295, + "learning_rate": 3.700243704305443e-06, + "loss": 1.6729, + "step": 929 + }, + { + "epoch": 1.0191780821917809, + "grad_norm": 2.3617212772369385, + "learning_rate": 3.6982128350934206e-06, + "loss": 1.6546, + "step": 930 + }, + { + "epoch": 1.0191780821917809, + "eval_loss": 1.8915373086929321, + "eval_runtime": 37.094, + "eval_samples_per_second": 26.959, + "eval_steps_per_second": 1.132, + "step": 930 + }, + { + "epoch": 1.0202739726027397, + "grad_norm": 2.6635043621063232, + "learning_rate": 3.6961819658813975e-06, + "loss": 1.628, + "step": 931 + }, + { + "epoch": 1.0213698630136987, + "grad_norm": 2.724837303161621, + "learning_rate": 3.694151096669375e-06, + "loss": 1.7537, + "step": 932 + }, + { + "epoch": 1.0224657534246575, + "grad_norm": 2.56447434425354, + "learning_rate": 3.692120227457352e-06, + "loss": 1.6638, + "step": 933 + }, + { + "epoch": 1.0235616438356165, + "grad_norm": 2.6624584197998047, + "learning_rate": 3.6900893582453297e-06, + "loss": 1.6806, + "step": 934 + }, + { + "epoch": 1.0246575342465754, + "grad_norm": 2.7147905826568604, + "learning_rate": 3.6880584890333066e-06, + "loss": 1.5887, + "step": 935 + }, + { + "epoch": 1.0257534246575342, + "grad_norm": 2.683840751647949, + "learning_rate": 3.6860276198212834e-06, + "loss": 1.6664, + "step": 936 + }, + { + "epoch": 1.0268493150684932, + "grad_norm": 2.8141531944274902, + "learning_rate": 3.683996750609261e-06, + "loss": 1.6744, + "step": 937 + }, + { + "epoch": 1.027945205479452, + "grad_norm": 2.6172802448272705, + "learning_rate": 3.681965881397238e-06, + "loss": 1.6866, + "step": 938 + }, + { + "epoch": 1.029041095890411, + "grad_norm": 2.7853286266326904, + "learning_rate": 3.6799350121852157e-06, + "loss": 1.7034, + "step": 939 + }, + { + "epoch": 1.0301369863013699, + "grad_norm": 2.7121260166168213, + "learning_rate": 3.6779041429731925e-06, + "loss": 1.6573, + "step": 940 + }, + { + "epoch": 1.0301369863013699, + "eval_loss": 1.883426308631897, + "eval_runtime": 37.1964, + "eval_samples_per_second": 26.884, + "eval_steps_per_second": 1.129, + "step": 940 + }, + { + "epoch": 1.0312328767123287, + "grad_norm": 2.4775121212005615, + "learning_rate": 3.6758732737611702e-06, + "loss": 1.6651, + "step": 941 + }, + { + "epoch": 1.0323287671232877, + "grad_norm": 2.5361502170562744, + "learning_rate": 3.673842404549147e-06, + "loss": 1.6524, + "step": 942 + }, + { + "epoch": 1.0334246575342465, + "grad_norm": 2.7969970703125, + "learning_rate": 3.671811535337125e-06, + "loss": 1.6598, + "step": 943 + }, + { + "epoch": 1.0345205479452055, + "grad_norm": 2.88427996635437, + "learning_rate": 3.6697806661251016e-06, + "loss": 1.6821, + "step": 944 + }, + { + "epoch": 1.0356164383561643, + "grad_norm": 2.9321346282958984, + "learning_rate": 3.6677497969130793e-06, + "loss": 1.7777, + "step": 945 + }, + { + "epoch": 1.0367123287671234, + "grad_norm": 2.802886724472046, + "learning_rate": 3.665718927701056e-06, + "loss": 1.6808, + "step": 946 + }, + { + "epoch": 1.0378082191780822, + "grad_norm": 2.85810923576355, + "learning_rate": 3.663688058489034e-06, + "loss": 1.7375, + "step": 947 + }, + { + "epoch": 1.038904109589041, + "grad_norm": 2.6501998901367188, + "learning_rate": 3.6616571892770107e-06, + "loss": 1.6662, + "step": 948 + }, + { + "epoch": 1.04, + "grad_norm": 2.550652265548706, + "learning_rate": 3.659626320064988e-06, + "loss": 1.6084, + "step": 949 + }, + { + "epoch": 1.0410958904109588, + "grad_norm": 2.5003185272216797, + "learning_rate": 3.6575954508529653e-06, + "loss": 1.681, + "step": 950 + }, + { + "epoch": 1.0410958904109588, + "eval_loss": 1.8800761699676514, + "eval_runtime": 37.2014, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 950 + }, + { + "epoch": 1.0421917808219179, + "grad_norm": 2.4262983798980713, + "learning_rate": 3.6555645816409426e-06, + "loss": 1.6608, + "step": 951 + }, + { + "epoch": 1.0432876712328767, + "grad_norm": 2.6474270820617676, + "learning_rate": 3.65353371242892e-06, + "loss": 1.6972, + "step": 952 + }, + { + "epoch": 1.0443835616438357, + "grad_norm": 2.5103557109832764, + "learning_rate": 3.651502843216897e-06, + "loss": 1.5581, + "step": 953 + }, + { + "epoch": 1.0454794520547945, + "grad_norm": 2.529756546020508, + "learning_rate": 3.6494719740048744e-06, + "loss": 1.693, + "step": 954 + }, + { + "epoch": 1.0465753424657533, + "grad_norm": 2.578362464904785, + "learning_rate": 3.6474411047928517e-06, + "loss": 1.6302, + "step": 955 + }, + { + "epoch": 1.0476712328767124, + "grad_norm": 2.8062336444854736, + "learning_rate": 3.645410235580829e-06, + "loss": 1.6855, + "step": 956 + }, + { + "epoch": 1.0487671232876712, + "grad_norm": 2.935694932937622, + "learning_rate": 3.6433793663688062e-06, + "loss": 1.6861, + "step": 957 + }, + { + "epoch": 1.0498630136986302, + "grad_norm": 2.5260093212127686, + "learning_rate": 3.6413484971567835e-06, + "loss": 1.6534, + "step": 958 + }, + { + "epoch": 1.050958904109589, + "grad_norm": 2.7627487182617188, + "learning_rate": 3.6393176279447608e-06, + "loss": 1.6833, + "step": 959 + }, + { + "epoch": 1.0520547945205478, + "grad_norm": 2.9553534984588623, + "learning_rate": 3.637286758732738e-06, + "loss": 1.6887, + "step": 960 + }, + { + "epoch": 1.0520547945205478, + "eval_loss": 1.8832038640975952, + "eval_runtime": 37.1828, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 960 + }, + { + "epoch": 1.0531506849315069, + "grad_norm": 2.5526084899902344, + "learning_rate": 3.6352558895207153e-06, + "loss": 1.7162, + "step": 961 + }, + { + "epoch": 1.0542465753424657, + "grad_norm": 2.761218786239624, + "learning_rate": 3.633225020308692e-06, + "loss": 1.6823, + "step": 962 + }, + { + "epoch": 1.0553424657534247, + "grad_norm": 2.672189950942993, + "learning_rate": 3.63119415109667e-06, + "loss": 1.6489, + "step": 963 + }, + { + "epoch": 1.0564383561643835, + "grad_norm": 2.7650046348571777, + "learning_rate": 3.6291632818846467e-06, + "loss": 1.7201, + "step": 964 + }, + { + "epoch": 1.0575342465753426, + "grad_norm": 2.6351473331451416, + "learning_rate": 3.6271324126726244e-06, + "loss": 1.6164, + "step": 965 + }, + { + "epoch": 1.0586301369863014, + "grad_norm": 2.5959131717681885, + "learning_rate": 3.6251015434606013e-06, + "loss": 1.603, + "step": 966 + }, + { + "epoch": 1.0597260273972602, + "grad_norm": 2.525513172149658, + "learning_rate": 3.623070674248579e-06, + "loss": 1.6671, + "step": 967 + }, + { + "epoch": 1.0608219178082192, + "grad_norm": 2.5529232025146484, + "learning_rate": 3.621039805036556e-06, + "loss": 1.6446, + "step": 968 + }, + { + "epoch": 1.061917808219178, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6846, + "step": 969 + }, + { + "epoch": 1.063013698630137, + "grad_norm": 2.648120403289795, + "learning_rate": 3.6190089358245335e-06, + "loss": 1.6786, + "step": 970 + }, + { + "epoch": 1.063013698630137, + "eval_loss": 1.885197639465332, + "eval_runtime": 37.2374, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 970 + }, + { + "epoch": 1.0641095890410959, + "grad_norm": 2.50577449798584, + "learning_rate": 3.6169780666125104e-06, + "loss": 1.6892, + "step": 971 + }, + { + "epoch": 1.0652054794520547, + "grad_norm": 2.57772159576416, + "learning_rate": 3.614947197400488e-06, + "loss": 1.5863, + "step": 972 + }, + { + "epoch": 1.0663013698630137, + "grad_norm": 2.5455222129821777, + "learning_rate": 3.612916328188465e-06, + "loss": 1.692, + "step": 973 + }, + { + "epoch": 1.0673972602739725, + "grad_norm": 2.6154329776763916, + "learning_rate": 3.610885458976442e-06, + "loss": 1.7027, + "step": 974 + }, + { + "epoch": 1.0684931506849316, + "grad_norm": 2.5776023864746094, + "learning_rate": 3.6088545897644195e-06, + "loss": 1.6353, + "step": 975 + }, + { + "epoch": 1.0695890410958904, + "grad_norm": 2.5480730533599854, + "learning_rate": 3.6068237205523963e-06, + "loss": 1.6358, + "step": 976 + }, + { + "epoch": 1.0706849315068494, + "grad_norm": 2.5939579010009766, + "learning_rate": 3.604792851340374e-06, + "loss": 1.7099, + "step": 977 + }, + { + "epoch": 1.0717808219178082, + "grad_norm": 2.65395188331604, + "learning_rate": 3.602761982128351e-06, + "loss": 1.6763, + "step": 978 + }, + { + "epoch": 1.072876712328767, + "grad_norm": 2.5302748680114746, + "learning_rate": 3.6007311129163286e-06, + "loss": 1.613, + "step": 979 + }, + { + "epoch": 1.073972602739726, + "grad_norm": 2.669433832168579, + "learning_rate": 3.5987002437043054e-06, + "loss": 1.6584, + "step": 980 + }, + { + "epoch": 1.073972602739726, + "eval_loss": 1.8816337585449219, + "eval_runtime": 37.2174, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 980 + }, + { + "epoch": 1.0750684931506849, + "grad_norm": 2.859063148498535, + "learning_rate": 3.596669374492283e-06, + "loss": 1.7096, + "step": 981 + }, + { + "epoch": 1.076164383561644, + "grad_norm": 2.531301259994507, + "learning_rate": 3.59463850528026e-06, + "loss": 1.6168, + "step": 982 + }, + { + "epoch": 1.0772602739726027, + "grad_norm": 2.5920488834381104, + "learning_rate": 3.5926076360682377e-06, + "loss": 1.6953, + "step": 983 + }, + { + "epoch": 1.0783561643835617, + "grad_norm": 2.698284387588501, + "learning_rate": 3.5905767668562145e-06, + "loss": 1.6602, + "step": 984 + }, + { + "epoch": 1.0794520547945206, + "grad_norm": 2.762387752532959, + "learning_rate": 3.5885458976441922e-06, + "loss": 1.7073, + "step": 985 + }, + { + "epoch": 1.0805479452054794, + "grad_norm": 2.554579019546509, + "learning_rate": 3.586515028432169e-06, + "loss": 1.7151, + "step": 986 + }, + { + "epoch": 1.0816438356164384, + "grad_norm": 2.644576072692871, + "learning_rate": 3.5844841592201464e-06, + "loss": 1.6088, + "step": 987 + }, + { + "epoch": 1.0827397260273972, + "grad_norm": 2.6375787258148193, + "learning_rate": 3.5824532900081237e-06, + "loss": 1.6434, + "step": 988 + }, + { + "epoch": 1.0838356164383562, + "grad_norm": 2.9475791454315186, + "learning_rate": 3.580422420796101e-06, + "loss": 1.6936, + "step": 989 + }, + { + "epoch": 1.084931506849315, + "grad_norm": 2.5753467082977295, + "learning_rate": 3.578391551584078e-06, + "loss": 1.5915, + "step": 990 + }, + { + "epoch": 1.084931506849315, + "eval_loss": 1.8799077272415161, + "eval_runtime": 37.2335, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 990 + }, + { + "epoch": 1.086027397260274, + "grad_norm": 2.8163156509399414, + "learning_rate": 3.5763606823720555e-06, + "loss": 1.6824, + "step": 991 + }, + { + "epoch": 1.087123287671233, + "grad_norm": 2.6741368770599365, + "learning_rate": 3.5743298131600328e-06, + "loss": 1.7066, + "step": 992 + }, + { + "epoch": 1.0882191780821917, + "grad_norm": 2.5235214233398438, + "learning_rate": 3.57229894394801e-06, + "loss": 1.6779, + "step": 993 + }, + { + "epoch": 1.0893150684931507, + "grad_norm": 2.620927095413208, + "learning_rate": 3.5702680747359873e-06, + "loss": 1.6392, + "step": 994 + }, + { + "epoch": 1.0904109589041096, + "grad_norm": 2.7678003311157227, + "learning_rate": 3.5682372055239646e-06, + "loss": 1.6768, + "step": 995 + }, + { + "epoch": 1.0915068493150686, + "grad_norm": 2.5900607109069824, + "learning_rate": 3.566206336311942e-06, + "loss": 1.618, + "step": 996 + }, + { + "epoch": 1.0926027397260274, + "grad_norm": 2.4440128803253174, + "learning_rate": 3.564175467099919e-06, + "loss": 1.5981, + "step": 997 + }, + { + "epoch": 1.0936986301369862, + "grad_norm": 2.686119318008423, + "learning_rate": 3.5621445978878964e-06, + "loss": 1.6649, + "step": 998 + }, + { + "epoch": 1.0947945205479452, + "grad_norm": 2.6912872791290283, + "learning_rate": 3.5601137286758737e-06, + "loss": 1.6426, + "step": 999 + }, + { + "epoch": 1.095890410958904, + "grad_norm": 2.7249274253845215, + "learning_rate": 3.5580828594638505e-06, + "loss": 1.6639, + "step": 1000 + }, + { + "epoch": 1.095890410958904, + "eval_loss": 1.8750032186508179, + "eval_runtime": 37.2319, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1000 + }, + { + "epoch": 1.096986301369863, + "grad_norm": 2.750321626663208, + "learning_rate": 3.5560519902518282e-06, + "loss": 1.7014, + "step": 1001 + }, + { + "epoch": 1.098082191780822, + "grad_norm": 2.593047618865967, + "learning_rate": 3.554021121039805e-06, + "loss": 1.6557, + "step": 1002 + }, + { + "epoch": 1.099178082191781, + "grad_norm": 2.744931936264038, + "learning_rate": 3.551990251827783e-06, + "loss": 1.6257, + "step": 1003 + }, + { + "epoch": 1.1002739726027397, + "grad_norm": 2.4940085411071777, + "learning_rate": 3.5499593826157596e-06, + "loss": 1.6546, + "step": 1004 + }, + { + "epoch": 1.1013698630136985, + "grad_norm": 2.6905856132507324, + "learning_rate": 3.5479285134037373e-06, + "loss": 1.6395, + "step": 1005 + }, + { + "epoch": 1.1024657534246576, + "grad_norm": 2.6744513511657715, + "learning_rate": 3.545897644191714e-06, + "loss": 1.6159, + "step": 1006 + }, + { + "epoch": 1.1035616438356164, + "grad_norm": 2.6766908168792725, + "learning_rate": 3.543866774979692e-06, + "loss": 1.6754, + "step": 1007 + }, + { + "epoch": 1.1046575342465754, + "grad_norm": 2.743103504180908, + "learning_rate": 3.5418359057676687e-06, + "loss": 1.6273, + "step": 1008 + }, + { + "epoch": 1.1057534246575342, + "grad_norm": 2.886472463607788, + "learning_rate": 3.5398050365556464e-06, + "loss": 1.6836, + "step": 1009 + }, + { + "epoch": 1.106849315068493, + "grad_norm": 2.823476791381836, + "learning_rate": 3.5377741673436233e-06, + "loss": 1.6475, + "step": 1010 + }, + { + "epoch": 1.106849315068493, + "eval_loss": 1.8762013912200928, + "eval_runtime": 36.8594, + "eval_samples_per_second": 27.13, + "eval_steps_per_second": 1.139, + "step": 1010 + }, + { + "epoch": 1.107945205479452, + "grad_norm": 2.741985559463501, + "learning_rate": 3.535743298131601e-06, + "loss": 1.6765, + "step": 1011 + }, + { + "epoch": 1.1090410958904109, + "grad_norm": 2.6321418285369873, + "learning_rate": 3.533712428919578e-06, + "loss": 1.6997, + "step": 1012 + }, + { + "epoch": 1.11013698630137, + "grad_norm": 2.7359395027160645, + "learning_rate": 3.5316815597075547e-06, + "loss": 1.6529, + "step": 1013 + }, + { + "epoch": 1.1112328767123287, + "grad_norm": 2.610210657119751, + "learning_rate": 3.5296506904955324e-06, + "loss": 1.68, + "step": 1014 + }, + { + "epoch": 1.1123287671232878, + "grad_norm": 2.7770233154296875, + "learning_rate": 3.5276198212835093e-06, + "loss": 1.6836, + "step": 1015 + }, + { + "epoch": 1.1134246575342466, + "grad_norm": 2.7146646976470947, + "learning_rate": 3.525588952071487e-06, + "loss": 1.699, + "step": 1016 + }, + { + "epoch": 1.1145205479452054, + "grad_norm": 2.782172679901123, + "learning_rate": 3.523558082859464e-06, + "loss": 1.6941, + "step": 1017 + }, + { + "epoch": 1.1156164383561644, + "grad_norm": 2.577692747116089, + "learning_rate": 3.5215272136474415e-06, + "loss": 1.6303, + "step": 1018 + }, + { + "epoch": 1.1167123287671232, + "grad_norm": 2.721348285675049, + "learning_rate": 3.5194963444354184e-06, + "loss": 1.6381, + "step": 1019 + }, + { + "epoch": 1.1178082191780823, + "grad_norm": 2.825709342956543, + "learning_rate": 3.517465475223396e-06, + "loss": 1.6144, + "step": 1020 + }, + { + "epoch": 1.1178082191780823, + "eval_loss": 1.8759764432907104, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 1020 + }, + { + "epoch": 1.118904109589041, + "grad_norm": 2.8446385860443115, + "learning_rate": 3.515434606011373e-06, + "loss": 1.685, + "step": 1021 + }, + { + "epoch": 1.12, + "grad_norm": 2.70577335357666, + "learning_rate": 3.5134037367993506e-06, + "loss": 1.7276, + "step": 1022 + }, + { + "epoch": 1.121095890410959, + "grad_norm": 2.6624722480773926, + "learning_rate": 3.5113728675873275e-06, + "loss": 1.6735, + "step": 1023 + }, + { + "epoch": 1.1221917808219177, + "grad_norm": 2.7956180572509766, + "learning_rate": 3.5093419983753047e-06, + "loss": 1.6413, + "step": 1024 + }, + { + "epoch": 1.1232876712328768, + "grad_norm": 2.792559862136841, + "learning_rate": 3.507311129163282e-06, + "loss": 1.6035, + "step": 1025 + }, + { + "epoch": 1.1243835616438356, + "grad_norm": 2.7113680839538574, + "learning_rate": 3.5052802599512593e-06, + "loss": 1.6461, + "step": 1026 + }, + { + "epoch": 1.1254794520547946, + "grad_norm": 2.656461715698242, + "learning_rate": 3.5032493907392366e-06, + "loss": 1.662, + "step": 1027 + }, + { + "epoch": 1.1265753424657534, + "grad_norm": 2.8650670051574707, + "learning_rate": 3.501218521527214e-06, + "loss": 1.5952, + "step": 1028 + }, + { + "epoch": 1.1276712328767124, + "grad_norm": 2.8043570518493652, + "learning_rate": 3.499187652315191e-06, + "loss": 1.6738, + "step": 1029 + }, + { + "epoch": 1.1287671232876713, + "grad_norm": 2.607752799987793, + "learning_rate": 3.4971567831031684e-06, + "loss": 1.6828, + "step": 1030 + }, + { + "epoch": 1.1287671232876713, + "eval_loss": 1.8729931116104126, + "eval_runtime": 37.2275, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 1030 + }, + { + "epoch": 1.12986301369863, + "grad_norm": 2.5819315910339355, + "learning_rate": 3.4951259138911457e-06, + "loss": 1.6108, + "step": 1031 + }, + { + "epoch": 1.130958904109589, + "grad_norm": 2.6494336128234863, + "learning_rate": 3.493095044679123e-06, + "loss": 1.7348, + "step": 1032 + }, + { + "epoch": 1.132054794520548, + "grad_norm": 2.552762508392334, + "learning_rate": 3.4910641754671002e-06, + "loss": 1.5534, + "step": 1033 + }, + { + "epoch": 1.133150684931507, + "grad_norm": 2.6580469608306885, + "learning_rate": 3.4890333062550775e-06, + "loss": 1.6675, + "step": 1034 + }, + { + "epoch": 1.1342465753424658, + "grad_norm": 2.922295331954956, + "learning_rate": 3.4870024370430548e-06, + "loss": 1.5819, + "step": 1035 + }, + { + "epoch": 1.1353424657534246, + "grad_norm": 2.7051825523376465, + "learning_rate": 3.484971567831032e-06, + "loss": 1.6573, + "step": 1036 + }, + { + "epoch": 1.1364383561643836, + "grad_norm": 2.738743782043457, + "learning_rate": 3.482940698619009e-06, + "loss": 1.6939, + "step": 1037 + }, + { + "epoch": 1.1375342465753424, + "grad_norm": 2.51552677154541, + "learning_rate": 3.4809098294069866e-06, + "loss": 1.665, + "step": 1038 + }, + { + "epoch": 1.1386301369863014, + "grad_norm": 2.578549385070801, + "learning_rate": 3.4788789601949634e-06, + "loss": 1.6642, + "step": 1039 + }, + { + "epoch": 1.1397260273972603, + "grad_norm": 2.7047011852264404, + "learning_rate": 3.476848090982941e-06, + "loss": 1.6298, + "step": 1040 + }, + { + "epoch": 1.1397260273972603, + "eval_loss": 1.8726528882980347, + "eval_runtime": 37.25, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1040 + }, + { + "epoch": 1.140821917808219, + "grad_norm": 2.722877025604248, + "learning_rate": 3.474817221770918e-06, + "loss": 1.7522, + "step": 1041 + }, + { + "epoch": 1.141917808219178, + "grad_norm": 2.464240074157715, + "learning_rate": 3.4727863525588957e-06, + "loss": 1.6876, + "step": 1042 + }, + { + "epoch": 1.143013698630137, + "grad_norm": 2.6737945079803467, + "learning_rate": 3.4707554833468725e-06, + "loss": 1.6453, + "step": 1043 + }, + { + "epoch": 1.144109589041096, + "grad_norm": 2.7960164546966553, + "learning_rate": 3.4687246141348502e-06, + "loss": 1.607, + "step": 1044 + }, + { + "epoch": 1.1452054794520548, + "grad_norm": 2.6985723972320557, + "learning_rate": 3.466693744922827e-06, + "loss": 1.6385, + "step": 1045 + }, + { + "epoch": 1.1463013698630138, + "grad_norm": 2.7561540603637695, + "learning_rate": 3.464662875710805e-06, + "loss": 1.6753, + "step": 1046 + }, + { + "epoch": 1.1473972602739726, + "grad_norm": 2.828885555267334, + "learning_rate": 3.4626320064987817e-06, + "loss": 1.6402, + "step": 1047 + }, + { + "epoch": 1.1484931506849314, + "grad_norm": 2.653183698654175, + "learning_rate": 3.4606011372867594e-06, + "loss": 1.6889, + "step": 1048 + }, + { + "epoch": 1.1495890410958904, + "grad_norm": 2.5803189277648926, + "learning_rate": 3.458570268074736e-06, + "loss": 1.654, + "step": 1049 + }, + { + "epoch": 1.1506849315068493, + "grad_norm": 2.5868985652923584, + "learning_rate": 3.4565393988627135e-06, + "loss": 1.6366, + "step": 1050 + }, + { + "epoch": 1.1506849315068493, + "eval_loss": 1.8656586408615112, + "eval_runtime": 37.2805, + "eval_samples_per_second": 26.824, + "eval_steps_per_second": 1.127, + "step": 1050 + }, + { + "epoch": 1.1517808219178083, + "grad_norm": 2.648036241531372, + "learning_rate": 3.4545085296506908e-06, + "loss": 1.6019, + "step": 1051 + }, + { + "epoch": 1.152876712328767, + "grad_norm": 2.5482382774353027, + "learning_rate": 3.452477660438668e-06, + "loss": 1.6597, + "step": 1052 + }, + { + "epoch": 1.1539726027397261, + "grad_norm": 2.700294017791748, + "learning_rate": 3.4504467912266453e-06, + "loss": 1.6075, + "step": 1053 + }, + { + "epoch": 1.155068493150685, + "grad_norm": 2.9346179962158203, + "learning_rate": 3.4484159220146226e-06, + "loss": 1.6558, + "step": 1054 + }, + { + "epoch": 1.1561643835616437, + "grad_norm": 2.893852710723877, + "learning_rate": 3.4463850528026e-06, + "loss": 1.6395, + "step": 1055 + }, + { + "epoch": 1.1572602739726028, + "grad_norm": 2.789057970046997, + "learning_rate": 3.444354183590577e-06, + "loss": 1.6833, + "step": 1056 + }, + { + "epoch": 1.1583561643835616, + "grad_norm": 2.745492935180664, + "learning_rate": 3.4423233143785544e-06, + "loss": 1.6266, + "step": 1057 + }, + { + "epoch": 1.1594520547945206, + "grad_norm": 2.7930359840393066, + "learning_rate": 3.4402924451665317e-06, + "loss": 1.5781, + "step": 1058 + }, + { + "epoch": 1.1605479452054794, + "grad_norm": 2.770575761795044, + "learning_rate": 3.438261575954509e-06, + "loss": 1.6518, + "step": 1059 + }, + { + "epoch": 1.1616438356164385, + "grad_norm": 2.7367143630981445, + "learning_rate": 3.4362307067424862e-06, + "loss": 1.6982, + "step": 1060 + }, + { + "epoch": 1.1616438356164385, + "eval_loss": 1.8708553314208984, + "eval_runtime": 37.2843, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1060 + }, + { + "epoch": 1.1627397260273973, + "grad_norm": 2.89182448387146, + "learning_rate": 3.434199837530463e-06, + "loss": 1.6587, + "step": 1061 + }, + { + "epoch": 1.163835616438356, + "grad_norm": 2.5572729110717773, + "learning_rate": 3.432168968318441e-06, + "loss": 1.5974, + "step": 1062 + }, + { + "epoch": 1.1649315068493151, + "grad_norm": 2.8172996044158936, + "learning_rate": 3.4301380991064176e-06, + "loss": 1.6941, + "step": 1063 + }, + { + "epoch": 1.166027397260274, + "grad_norm": 2.6480889320373535, + "learning_rate": 3.4281072298943953e-06, + "loss": 1.6902, + "step": 1064 + }, + { + "epoch": 1.167123287671233, + "grad_norm": 2.654735803604126, + "learning_rate": 3.426076360682372e-06, + "loss": 1.7083, + "step": 1065 + }, + { + "epoch": 1.1682191780821918, + "grad_norm": 2.4182465076446533, + "learning_rate": 3.42404549147035e-06, + "loss": 1.6494, + "step": 1066 + }, + { + "epoch": 1.1693150684931506, + "grad_norm": 2.6808230876922607, + "learning_rate": 3.4220146222583267e-06, + "loss": 1.5782, + "step": 1067 + }, + { + "epoch": 1.1704109589041096, + "grad_norm": 2.648282527923584, + "learning_rate": 3.4199837530463044e-06, + "loss": 1.6695, + "step": 1068 + }, + { + "epoch": 1.1715068493150684, + "grad_norm": 2.5920112133026123, + "learning_rate": 3.4179528838342813e-06, + "loss": 1.7234, + "step": 1069 + }, + { + "epoch": 1.1726027397260275, + "grad_norm": 2.690538167953491, + "learning_rate": 3.415922014622259e-06, + "loss": 1.686, + "step": 1070 + }, + { + "epoch": 1.1726027397260275, + "eval_loss": 1.8717143535614014, + "eval_runtime": 37.2714, + "eval_samples_per_second": 26.83, + "eval_steps_per_second": 1.127, + "step": 1070 + }, + { + "epoch": 1.1736986301369863, + "grad_norm": 2.672515392303467, + "learning_rate": 3.413891145410236e-06, + "loss": 1.6883, + "step": 1071 + }, + { + "epoch": 1.174794520547945, + "grad_norm": 2.8012726306915283, + "learning_rate": 3.4118602761982135e-06, + "loss": 1.6762, + "step": 1072 + }, + { + "epoch": 1.1758904109589041, + "grad_norm": 2.7551519870758057, + "learning_rate": 3.4098294069861904e-06, + "loss": 1.6694, + "step": 1073 + }, + { + "epoch": 1.176986301369863, + "grad_norm": 2.687445878982544, + "learning_rate": 3.4077985377741673e-06, + "loss": 1.6267, + "step": 1074 + }, + { + "epoch": 1.178082191780822, + "grad_norm": 2.674445390701294, + "learning_rate": 3.405767668562145e-06, + "loss": 1.6944, + "step": 1075 + }, + { + "epoch": 1.1791780821917808, + "grad_norm": 2.797832489013672, + "learning_rate": 3.403736799350122e-06, + "loss": 1.6735, + "step": 1076 + }, + { + "epoch": 1.1802739726027398, + "grad_norm": 2.7843174934387207, + "learning_rate": 3.4017059301380995e-06, + "loss": 1.6183, + "step": 1077 + }, + { + "epoch": 1.1813698630136986, + "grad_norm": 2.678809404373169, + "learning_rate": 3.3996750609260764e-06, + "loss": 1.7336, + "step": 1078 + }, + { + "epoch": 1.1824657534246574, + "grad_norm": 2.748838424682617, + "learning_rate": 3.397644191714054e-06, + "loss": 1.6499, + "step": 1079 + }, + { + "epoch": 1.1835616438356165, + "grad_norm": 2.5785276889801025, + "learning_rate": 3.395613322502031e-06, + "loss": 1.6589, + "step": 1080 + }, + { + "epoch": 1.1835616438356165, + "eval_loss": 1.8695456981658936, + "eval_runtime": 37.2493, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1080 + }, + { + "epoch": 1.1846575342465753, + "grad_norm": 2.7307844161987305, + "learning_rate": 3.3935824532900086e-06, + "loss": 1.5989, + "step": 1081 + }, + { + "epoch": 1.1857534246575343, + "grad_norm": 2.83357572555542, + "learning_rate": 3.3915515840779855e-06, + "loss": 1.6568, + "step": 1082 + }, + { + "epoch": 1.1868493150684931, + "grad_norm": 2.6178860664367676, + "learning_rate": 3.389520714865963e-06, + "loss": 1.6654, + "step": 1083 + }, + { + "epoch": 1.1879452054794521, + "grad_norm": 2.699328899383545, + "learning_rate": 3.38748984565394e-06, + "loss": 1.6657, + "step": 1084 + }, + { + "epoch": 1.189041095890411, + "grad_norm": 2.7418158054351807, + "learning_rate": 3.3854589764419177e-06, + "loss": 1.6351, + "step": 1085 + }, + { + "epoch": 1.1901369863013698, + "grad_norm": 2.5589001178741455, + "learning_rate": 3.3834281072298946e-06, + "loss": 1.6351, + "step": 1086 + }, + { + "epoch": 1.1912328767123288, + "grad_norm": 2.492795944213867, + "learning_rate": 3.381397238017872e-06, + "loss": 1.6188, + "step": 1087 + }, + { + "epoch": 1.1923287671232876, + "grad_norm": 2.8586132526397705, + "learning_rate": 3.379366368805849e-06, + "loss": 1.7127, + "step": 1088 + }, + { + "epoch": 1.1934246575342466, + "grad_norm": 2.7773218154907227, + "learning_rate": 3.3773354995938264e-06, + "loss": 1.6529, + "step": 1089 + }, + { + "epoch": 1.1945205479452055, + "grad_norm": 2.6956539154052734, + "learning_rate": 3.3753046303818037e-06, + "loss": 1.6566, + "step": 1090 + }, + { + "epoch": 1.1945205479452055, + "eval_loss": 1.8692638874053955, + "eval_runtime": 37.2317, + "eval_samples_per_second": 26.859, + "eval_steps_per_second": 1.128, + "step": 1090 + }, + { + "epoch": 1.1956164383561645, + "grad_norm": 2.7406537532806396, + "learning_rate": 3.373273761169781e-06, + "loss": 1.6482, + "step": 1091 + }, + { + "epoch": 1.1967123287671233, + "grad_norm": 2.746194362640381, + "learning_rate": 3.3712428919577582e-06, + "loss": 1.6343, + "step": 1092 + }, + { + "epoch": 1.1978082191780821, + "grad_norm": 2.767970323562622, + "learning_rate": 3.3692120227457355e-06, + "loss": 1.6577, + "step": 1093 + }, + { + "epoch": 1.1989041095890411, + "grad_norm": 2.730640172958374, + "learning_rate": 3.3671811535337128e-06, + "loss": 1.5941, + "step": 1094 + }, + { + "epoch": 1.2, + "grad_norm": 2.5661444664001465, + "learning_rate": 3.36515028432169e-06, + "loss": 1.6362, + "step": 1095 + }, + { + "epoch": 1.201095890410959, + "grad_norm": 2.5915091037750244, + "learning_rate": 3.3631194151096673e-06, + "loss": 1.6526, + "step": 1096 + }, + { + "epoch": 1.2021917808219178, + "grad_norm": 2.7135071754455566, + "learning_rate": 3.3610885458976446e-06, + "loss": 1.6999, + "step": 1097 + }, + { + "epoch": 1.2032876712328768, + "grad_norm": 2.591201066970825, + "learning_rate": 3.359057676685622e-06, + "loss": 1.684, + "step": 1098 + }, + { + "epoch": 1.2043835616438356, + "grad_norm": 2.564439296722412, + "learning_rate": 3.357026807473599e-06, + "loss": 1.5815, + "step": 1099 + }, + { + "epoch": 1.2054794520547945, + "grad_norm": 2.4480631351470947, + "learning_rate": 3.354995938261576e-06, + "loss": 1.6284, + "step": 1100 + }, + { + "epoch": 1.2054794520547945, + "eval_loss": 1.8628013134002686, + "eval_runtime": 37.23, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1100 + }, + { + "epoch": 1.2065753424657535, + "grad_norm": 2.5392649173736572, + "learning_rate": 3.3529650690495537e-06, + "loss": 1.5841, + "step": 1101 + }, + { + "epoch": 1.2076712328767123, + "grad_norm": 2.6359004974365234, + "learning_rate": 3.3509341998375305e-06, + "loss": 1.6278, + "step": 1102 + }, + { + "epoch": 1.208767123287671, + "grad_norm": 2.803671360015869, + "learning_rate": 3.3489033306255082e-06, + "loss": 1.6602, + "step": 1103 + }, + { + "epoch": 1.2098630136986301, + "grad_norm": 2.6031086444854736, + "learning_rate": 3.346872461413485e-06, + "loss": 1.6021, + "step": 1104 + }, + { + "epoch": 1.210958904109589, + "grad_norm": 2.687382459640503, + "learning_rate": 3.344841592201463e-06, + "loss": 1.6207, + "step": 1105 + }, + { + "epoch": 1.212054794520548, + "grad_norm": 2.6388001441955566, + "learning_rate": 3.3428107229894397e-06, + "loss": 1.6194, + "step": 1106 + }, + { + "epoch": 1.2131506849315068, + "grad_norm": 2.603121042251587, + "learning_rate": 3.3407798537774174e-06, + "loss": 1.6532, + "step": 1107 + }, + { + "epoch": 1.2142465753424658, + "grad_norm": 2.8053836822509766, + "learning_rate": 3.338748984565394e-06, + "loss": 1.68, + "step": 1108 + }, + { + "epoch": 1.2153424657534246, + "grad_norm": 2.5472421646118164, + "learning_rate": 3.336718115353372e-06, + "loss": 1.7164, + "step": 1109 + }, + { + "epoch": 1.2164383561643834, + "grad_norm": 2.5142838954925537, + "learning_rate": 3.3346872461413488e-06, + "loss": 1.6441, + "step": 1110 + }, + { + "epoch": 1.2164383561643834, + "eval_loss": 1.8628681898117065, + "eval_runtime": 36.8298, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 1110 + }, + { + "epoch": 1.2175342465753425, + "grad_norm": 2.6917803287506104, + "learning_rate": 3.3326563769293256e-06, + "loss": 1.6235, + "step": 1111 + }, + { + "epoch": 1.2186301369863013, + "grad_norm": 2.6062285900115967, + "learning_rate": 3.3306255077173033e-06, + "loss": 1.6488, + "step": 1112 + }, + { + "epoch": 1.2197260273972603, + "grad_norm": 2.5503156185150146, + "learning_rate": 3.32859463850528e-06, + "loss": 1.639, + "step": 1113 + }, + { + "epoch": 1.2208219178082191, + "grad_norm": 2.433518648147583, + "learning_rate": 3.326563769293258e-06, + "loss": 1.5951, + "step": 1114 + }, + { + "epoch": 1.2219178082191782, + "grad_norm": 2.600661516189575, + "learning_rate": 3.3245329000812347e-06, + "loss": 1.6148, + "step": 1115 + }, + { + "epoch": 1.223013698630137, + "grad_norm": 2.5558395385742188, + "learning_rate": 3.3225020308692124e-06, + "loss": 1.6334, + "step": 1116 + }, + { + "epoch": 1.2241095890410958, + "grad_norm": 2.6363396644592285, + "learning_rate": 3.3204711616571893e-06, + "loss": 1.5858, + "step": 1117 + }, + { + "epoch": 1.2252054794520548, + "grad_norm": 2.819617748260498, + "learning_rate": 3.318440292445167e-06, + "loss": 1.6274, + "step": 1118 + }, + { + "epoch": 1.2263013698630136, + "grad_norm": 2.9121081829071045, + "learning_rate": 3.316409423233144e-06, + "loss": 1.6081, + "step": 1119 + }, + { + "epoch": 1.2273972602739727, + "grad_norm": 2.734858989715576, + "learning_rate": 3.3143785540211215e-06, + "loss": 1.6889, + "step": 1120 + }, + { + "epoch": 1.2273972602739727, + "eval_loss": 1.8644160032272339, + "eval_runtime": 37.0389, + "eval_samples_per_second": 26.999, + "eval_steps_per_second": 1.134, + "step": 1120 + }, + { + "epoch": 1.2284931506849315, + "grad_norm": 2.6940014362335205, + "learning_rate": 3.3123476848090984e-06, + "loss": 1.6463, + "step": 1121 + }, + { + "epoch": 1.2295890410958905, + "grad_norm": 2.82663631439209, + "learning_rate": 3.310316815597076e-06, + "loss": 1.6474, + "step": 1122 + }, + { + "epoch": 1.2306849315068493, + "grad_norm": 2.634434223175049, + "learning_rate": 3.308285946385053e-06, + "loss": 1.7172, + "step": 1123 + }, + { + "epoch": 1.2317808219178081, + "grad_norm": 2.6721079349517822, + "learning_rate": 3.30625507717303e-06, + "loss": 1.5795, + "step": 1124 + }, + { + "epoch": 1.2328767123287672, + "grad_norm": 2.786130666732788, + "learning_rate": 3.3042242079610075e-06, + "loss": 1.6634, + "step": 1125 + }, + { + "epoch": 1.233972602739726, + "grad_norm": 2.7698960304260254, + "learning_rate": 3.3021933387489847e-06, + "loss": 1.687, + "step": 1126 + }, + { + "epoch": 1.235068493150685, + "grad_norm": 2.6305580139160156, + "learning_rate": 3.300162469536962e-06, + "loss": 1.6372, + "step": 1127 + }, + { + "epoch": 1.2361643835616438, + "grad_norm": 2.5221288204193115, + "learning_rate": 3.2981316003249393e-06, + "loss": 1.6746, + "step": 1128 + }, + { + "epoch": 1.2372602739726029, + "grad_norm": 2.583643674850464, + "learning_rate": 3.2961007311129166e-06, + "loss": 1.6538, + "step": 1129 + }, + { + "epoch": 1.2383561643835617, + "grad_norm": 2.5978968143463135, + "learning_rate": 3.294069861900894e-06, + "loss": 1.7177, + "step": 1130 + }, + { + "epoch": 1.2383561643835617, + "eval_loss": 1.8665071725845337, + "eval_runtime": 37.1194, + "eval_samples_per_second": 26.94, + "eval_steps_per_second": 1.131, + "step": 1130 + }, + { + "epoch": 1.2394520547945205, + "grad_norm": 2.6446492671966553, + "learning_rate": 3.292038992688871e-06, + "loss": 1.657, + "step": 1131 + }, + { + "epoch": 1.2405479452054795, + "grad_norm": 2.646174192428589, + "learning_rate": 3.2900081234768484e-06, + "loss": 1.682, + "step": 1132 + }, + { + "epoch": 1.2416438356164383, + "grad_norm": 2.671532392501831, + "learning_rate": 3.2879772542648257e-06, + "loss": 1.7033, + "step": 1133 + }, + { + "epoch": 1.2427397260273974, + "grad_norm": 2.616450309753418, + "learning_rate": 3.285946385052803e-06, + "loss": 1.586, + "step": 1134 + }, + { + "epoch": 1.2438356164383562, + "grad_norm": 2.552800416946411, + "learning_rate": 3.2839155158407802e-06, + "loss": 1.6599, + "step": 1135 + }, + { + "epoch": 1.244931506849315, + "grad_norm": 2.5700533390045166, + "learning_rate": 3.2818846466287575e-06, + "loss": 1.5593, + "step": 1136 + }, + { + "epoch": 1.246027397260274, + "grad_norm": 2.675553560256958, + "learning_rate": 3.2798537774167344e-06, + "loss": 1.7097, + "step": 1137 + }, + { + "epoch": 1.2471232876712328, + "grad_norm": 2.6610019207000732, + "learning_rate": 3.277822908204712e-06, + "loss": 1.6408, + "step": 1138 + }, + { + "epoch": 1.2482191780821918, + "grad_norm": 2.55435848236084, + "learning_rate": 3.275792038992689e-06, + "loss": 1.5909, + "step": 1139 + }, + { + "epoch": 1.2493150684931507, + "grad_norm": 2.811556816101074, + "learning_rate": 3.2737611697806666e-06, + "loss": 1.6552, + "step": 1140 + }, + { + "epoch": 1.2493150684931507, + "eval_loss": 1.8618569374084473, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 1140 + }, + { + "epoch": 1.2504109589041095, + "grad_norm": 2.7060115337371826, + "learning_rate": 3.2717303005686435e-06, + "loss": 1.6144, + "step": 1141 + }, + { + "epoch": 1.2515068493150685, + "grad_norm": 2.6817731857299805, + "learning_rate": 3.269699431356621e-06, + "loss": 1.6481, + "step": 1142 + }, + { + "epoch": 1.2526027397260273, + "grad_norm": 2.8206212520599365, + "learning_rate": 3.267668562144598e-06, + "loss": 1.6294, + "step": 1143 + }, + { + "epoch": 1.2536986301369863, + "grad_norm": 2.834332227706909, + "learning_rate": 3.2656376929325757e-06, + "loss": 1.6103, + "step": 1144 + }, + { + "epoch": 1.2547945205479452, + "grad_norm": 2.754049062728882, + "learning_rate": 3.2636068237205526e-06, + "loss": 1.6788, + "step": 1145 + }, + { + "epoch": 1.2558904109589042, + "grad_norm": 2.6089437007904053, + "learning_rate": 3.2615759545085303e-06, + "loss": 1.6469, + "step": 1146 + }, + { + "epoch": 1.256986301369863, + "grad_norm": 2.6095848083496094, + "learning_rate": 3.259545085296507e-06, + "loss": 1.7122, + "step": 1147 + }, + { + "epoch": 1.2580821917808218, + "grad_norm": 2.606787919998169, + "learning_rate": 3.257514216084484e-06, + "loss": 1.5851, + "step": 1148 + }, + { + "epoch": 1.2591780821917808, + "grad_norm": 2.643245220184326, + "learning_rate": 3.2554833468724617e-06, + "loss": 1.7042, + "step": 1149 + }, + { + "epoch": 1.2602739726027397, + "grad_norm": 2.8400895595550537, + "learning_rate": 3.2534524776604385e-06, + "loss": 1.6526, + "step": 1150 + }, + { + "epoch": 1.2602739726027397, + "eval_loss": 1.8640165328979492, + "eval_runtime": 37.162, + "eval_samples_per_second": 26.909, + "eval_steps_per_second": 1.13, + "step": 1150 + }, + { + "epoch": 1.2613698630136987, + "grad_norm": 2.519524335861206, + "learning_rate": 3.2514216084484162e-06, + "loss": 1.6268, + "step": 1151 + }, + { + "epoch": 1.2624657534246575, + "grad_norm": 2.565365791320801, + "learning_rate": 3.249390739236393e-06, + "loss": 1.6562, + "step": 1152 + }, + { + "epoch": 1.2635616438356165, + "grad_norm": 2.6364638805389404, + "learning_rate": 3.2473598700243708e-06, + "loss": 1.6143, + "step": 1153 + }, + { + "epoch": 1.2646575342465753, + "grad_norm": 2.6235382556915283, + "learning_rate": 3.2453290008123476e-06, + "loss": 1.64, + "step": 1154 + }, + { + "epoch": 1.2657534246575342, + "grad_norm": 2.7358779907226562, + "learning_rate": 3.2432981316003253e-06, + "loss": 1.6509, + "step": 1155 + }, + { + "epoch": 1.2668493150684932, + "grad_norm": 2.6720035076141357, + "learning_rate": 3.241267262388302e-06, + "loss": 1.5902, + "step": 1156 + }, + { + "epoch": 1.267945205479452, + "grad_norm": 2.5960681438446045, + "learning_rate": 3.23923639317628e-06, + "loss": 1.6378, + "step": 1157 + }, + { + "epoch": 1.269041095890411, + "grad_norm": 2.6655092239379883, + "learning_rate": 3.2372055239642567e-06, + "loss": 1.6102, + "step": 1158 + }, + { + "epoch": 1.2701369863013698, + "grad_norm": 2.7873294353485107, + "learning_rate": 3.2351746547522344e-06, + "loss": 1.6558, + "step": 1159 + }, + { + "epoch": 1.2712328767123289, + "grad_norm": 2.746189594268799, + "learning_rate": 3.2331437855402113e-06, + "loss": 1.7116, + "step": 1160 + }, + { + "epoch": 1.2712328767123289, + "eval_loss": 1.8605948686599731, + "eval_runtime": 37.1486, + "eval_samples_per_second": 26.919, + "eval_steps_per_second": 1.131, + "step": 1160 + }, + { + "epoch": 1.2723287671232877, + "grad_norm": 2.5058770179748535, + "learning_rate": 3.2311129163281885e-06, + "loss": 1.6156, + "step": 1161 + }, + { + "epoch": 1.2734246575342465, + "grad_norm": 2.8002450466156006, + "learning_rate": 3.229082047116166e-06, + "loss": 1.6884, + "step": 1162 + }, + { + "epoch": 1.2745205479452055, + "grad_norm": 2.696186065673828, + "learning_rate": 3.227051177904143e-06, + "loss": 1.6403, + "step": 1163 + }, + { + "epoch": 1.2756164383561643, + "grad_norm": 2.5473484992980957, + "learning_rate": 3.2250203086921204e-06, + "loss": 1.6375, + "step": 1164 + }, + { + "epoch": 1.2767123287671232, + "grad_norm": 2.6391875743865967, + "learning_rate": 3.2229894394800977e-06, + "loss": 1.6597, + "step": 1165 + }, + { + "epoch": 1.2778082191780822, + "grad_norm": 2.647244453430176, + "learning_rate": 3.220958570268075e-06, + "loss": 1.5666, + "step": 1166 + }, + { + "epoch": 1.2789041095890412, + "grad_norm": 2.849031448364258, + "learning_rate": 3.218927701056052e-06, + "loss": 1.6211, + "step": 1167 + }, + { + "epoch": 1.28, + "grad_norm": 2.7565410137176514, + "learning_rate": 3.2168968318440295e-06, + "loss": 1.5686, + "step": 1168 + }, + { + "epoch": 1.2810958904109588, + "grad_norm": 2.938270092010498, + "learning_rate": 3.2148659626320068e-06, + "loss": 1.7046, + "step": 1169 + }, + { + "epoch": 1.2821917808219179, + "grad_norm": 2.6462385654449463, + "learning_rate": 3.212835093419984e-06, + "loss": 1.602, + "step": 1170 + }, + { + "epoch": 1.2821917808219179, + "eval_loss": 1.8577098846435547, + "eval_runtime": 37.1448, + "eval_samples_per_second": 26.922, + "eval_steps_per_second": 1.131, + "step": 1170 + }, + { + "epoch": 1.2832876712328767, + "grad_norm": 2.679330587387085, + "learning_rate": 3.2108042242079613e-06, + "loss": 1.636, + "step": 1171 + }, + { + "epoch": 1.2843835616438355, + "grad_norm": 2.6410343647003174, + "learning_rate": 3.2087733549959386e-06, + "loss": 1.6904, + "step": 1172 + }, + { + "epoch": 1.2854794520547945, + "grad_norm": 2.6398510932922363, + "learning_rate": 3.206742485783916e-06, + "loss": 1.597, + "step": 1173 + }, + { + "epoch": 1.2865753424657536, + "grad_norm": 2.4636199474334717, + "learning_rate": 3.2047116165718927e-06, + "loss": 1.5823, + "step": 1174 + }, + { + "epoch": 1.2876712328767124, + "grad_norm": 2.646580457687378, + "learning_rate": 3.2026807473598704e-06, + "loss": 1.6591, + "step": 1175 + }, + { + "epoch": 1.2887671232876712, + "grad_norm": 2.8582923412323, + "learning_rate": 3.2006498781478473e-06, + "loss": 1.6731, + "step": 1176 + }, + { + "epoch": 1.2898630136986302, + "grad_norm": 2.5474820137023926, + "learning_rate": 3.198619008935825e-06, + "loss": 1.5656, + "step": 1177 + }, + { + "epoch": 1.290958904109589, + "grad_norm": 2.6609606742858887, + "learning_rate": 3.196588139723802e-06, + "loss": 1.6374, + "step": 1178 + }, + { + "epoch": 1.2920547945205478, + "grad_norm": 2.651947021484375, + "learning_rate": 3.1945572705117795e-06, + "loss": 1.6557, + "step": 1179 + }, + { + "epoch": 1.2931506849315069, + "grad_norm": 2.6957671642303467, + "learning_rate": 3.1925264012997564e-06, + "loss": 1.6436, + "step": 1180 + }, + { + "epoch": 1.2931506849315069, + "eval_loss": 1.8583322763442993, + "eval_runtime": 37.1374, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1180 + }, + { + "epoch": 1.2942465753424657, + "grad_norm": 2.5442283153533936, + "learning_rate": 3.190495532087734e-06, + "loss": 1.6424, + "step": 1181 + }, + { + "epoch": 1.2953424657534247, + "grad_norm": 2.593792200088501, + "learning_rate": 3.188464662875711e-06, + "loss": 1.7012, + "step": 1182 + }, + { + "epoch": 1.2964383561643835, + "grad_norm": 2.5021846294403076, + "learning_rate": 3.1864337936636886e-06, + "loss": 1.6268, + "step": 1183 + }, + { + "epoch": 1.2975342465753426, + "grad_norm": 2.7368624210357666, + "learning_rate": 3.1844029244516655e-06, + "loss": 1.5953, + "step": 1184 + }, + { + "epoch": 1.2986301369863014, + "grad_norm": 2.5100302696228027, + "learning_rate": 3.182372055239643e-06, + "loss": 1.5866, + "step": 1185 + }, + { + "epoch": 1.2997260273972602, + "grad_norm": 2.5767033100128174, + "learning_rate": 3.18034118602762e-06, + "loss": 1.6621, + "step": 1186 + }, + { + "epoch": 1.3008219178082192, + "grad_norm": 2.4694652557373047, + "learning_rate": 3.1783103168155973e-06, + "loss": 1.6554, + "step": 1187 + }, + { + "epoch": 1.301917808219178, + "grad_norm": 2.606158971786499, + "learning_rate": 3.1762794476035746e-06, + "loss": 1.6238, + "step": 1188 + }, + { + "epoch": 1.303013698630137, + "grad_norm": 2.857424020767212, + "learning_rate": 3.174248578391552e-06, + "loss": 1.626, + "step": 1189 + }, + { + "epoch": 1.3041095890410959, + "grad_norm": 2.6120405197143555, + "learning_rate": 3.172217709179529e-06, + "loss": 1.6134, + "step": 1190 + }, + { + "epoch": 1.3041095890410959, + "eval_loss": 1.8599685430526733, + "eval_runtime": 37.1373, + "eval_samples_per_second": 26.927, + "eval_steps_per_second": 1.131, + "step": 1190 + }, + { + "epoch": 1.305205479452055, + "grad_norm": 2.61995530128479, + "learning_rate": 3.1701868399675064e-06, + "loss": 1.6303, + "step": 1191 + }, + { + "epoch": 1.3063013698630137, + "grad_norm": 2.6253163814544678, + "learning_rate": 3.1681559707554837e-06, + "loss": 1.5886, + "step": 1192 + }, + { + "epoch": 1.3073972602739725, + "grad_norm": 2.627302885055542, + "learning_rate": 3.166125101543461e-06, + "loss": 1.6526, + "step": 1193 + }, + { + "epoch": 1.3084931506849315, + "grad_norm": 2.8007607460021973, + "learning_rate": 3.1640942323314382e-06, + "loss": 1.6581, + "step": 1194 + }, + { + "epoch": 1.3095890410958904, + "grad_norm": 2.551227331161499, + "learning_rate": 3.1620633631194155e-06, + "loss": 1.5921, + "step": 1195 + }, + { + "epoch": 1.3106849315068494, + "grad_norm": 2.7288761138916016, + "learning_rate": 3.1600324939073928e-06, + "loss": 1.6478, + "step": 1196 + }, + { + "epoch": 1.3117808219178082, + "grad_norm": 2.5698401927948, + "learning_rate": 3.15800162469537e-06, + "loss": 1.6552, + "step": 1197 + }, + { + "epoch": 1.3128767123287672, + "grad_norm": 2.796701192855835, + "learning_rate": 3.155970755483347e-06, + "loss": 1.6391, + "step": 1198 + }, + { + "epoch": 1.313972602739726, + "grad_norm": 2.6523337364196777, + "learning_rate": 3.1539398862713246e-06, + "loss": 1.6495, + "step": 1199 + }, + { + "epoch": 1.3150684931506849, + "grad_norm": 2.5528745651245117, + "learning_rate": 3.1519090170593015e-06, + "loss": 1.6396, + "step": 1200 + }, + { + "epoch": 1.3150684931506849, + "eval_loss": 1.8551454544067383, + "eval_runtime": 37.1721, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 1200 + }, + { + "epoch": 1.316164383561644, + "grad_norm": 2.650125741958618, + "learning_rate": 3.149878147847279e-06, + "loss": 1.6803, + "step": 1201 + }, + { + "epoch": 1.3172602739726027, + "grad_norm": 2.5522522926330566, + "learning_rate": 3.147847278635256e-06, + "loss": 1.726, + "step": 1202 + }, + { + "epoch": 1.3183561643835615, + "grad_norm": 2.5302486419677734, + "learning_rate": 3.1458164094232337e-06, + "loss": 1.6238, + "step": 1203 + }, + { + "epoch": 1.3194520547945205, + "grad_norm": 2.517793893814087, + "learning_rate": 3.1437855402112106e-06, + "loss": 1.5879, + "step": 1204 + }, + { + "epoch": 1.3205479452054796, + "grad_norm": 2.5749523639678955, + "learning_rate": 3.1417546709991883e-06, + "loss": 1.6361, + "step": 1205 + }, + { + "epoch": 1.3216438356164384, + "grad_norm": 2.6382429599761963, + "learning_rate": 3.139723801787165e-06, + "loss": 1.6143, + "step": 1206 + }, + { + "epoch": 1.3227397260273972, + "grad_norm": 2.7513139247894287, + "learning_rate": 3.137692932575143e-06, + "loss": 1.6608, + "step": 1207 + }, + { + "epoch": 1.3238356164383562, + "grad_norm": 2.616007089614868, + "learning_rate": 3.1356620633631197e-06, + "loss": 1.5842, + "step": 1208 + }, + { + "epoch": 1.324931506849315, + "grad_norm": 2.619389772415161, + "learning_rate": 3.1336311941510974e-06, + "loss": 1.6419, + "step": 1209 + }, + { + "epoch": 1.3260273972602739, + "grad_norm": 2.6845619678497314, + "learning_rate": 3.1316003249390742e-06, + "loss": 1.6726, + "step": 1210 + }, + { + "epoch": 1.3260273972602739, + "eval_loss": 1.8519575595855713, + "eval_runtime": 36.823, + "eval_samples_per_second": 27.157, + "eval_steps_per_second": 1.141, + "step": 1210 + }, + { + "epoch": 1.3271232876712329, + "grad_norm": 2.9907562732696533, + "learning_rate": 3.129569455727051e-06, + "loss": 1.6813, + "step": 1211 + }, + { + "epoch": 1.3282191780821917, + "grad_norm": 2.5070035457611084, + "learning_rate": 3.1275385865150288e-06, + "loss": 1.6449, + "step": 1212 + }, + { + "epoch": 1.3293150684931507, + "grad_norm": 2.6286256313323975, + "learning_rate": 3.1255077173030056e-06, + "loss": 1.6614, + "step": 1213 + }, + { + "epoch": 1.3304109589041095, + "grad_norm": 2.5735433101654053, + "learning_rate": 3.1234768480909833e-06, + "loss": 1.6191, + "step": 1214 + }, + { + "epoch": 1.3315068493150686, + "grad_norm": 2.637507438659668, + "learning_rate": 3.12144597887896e-06, + "loss": 1.6649, + "step": 1215 + }, + { + "epoch": 1.3326027397260274, + "grad_norm": 2.5027501583099365, + "learning_rate": 3.119415109666938e-06, + "loss": 1.6779, + "step": 1216 + }, + { + "epoch": 1.3336986301369862, + "grad_norm": 2.8430395126342773, + "learning_rate": 3.1173842404549147e-06, + "loss": 1.6428, + "step": 1217 + }, + { + "epoch": 1.3347945205479452, + "grad_norm": 2.6394317150115967, + "learning_rate": 3.1153533712428924e-06, + "loss": 1.655, + "step": 1218 + }, + { + "epoch": 1.335890410958904, + "grad_norm": 2.7121026515960693, + "learning_rate": 3.1133225020308693e-06, + "loss": 1.6798, + "step": 1219 + }, + { + "epoch": 1.336986301369863, + "grad_norm": 2.6572396755218506, + "learning_rate": 3.111291632818847e-06, + "loss": 1.6937, + "step": 1220 + }, + { + "epoch": 1.336986301369863, + "eval_loss": 1.857973575592041, + "eval_runtime": 37.0635, + "eval_samples_per_second": 26.981, + "eval_steps_per_second": 1.133, + "step": 1220 + }, + { + "epoch": 1.3380821917808219, + "grad_norm": 2.660865306854248, + "learning_rate": 3.109260763606824e-06, + "loss": 1.6454, + "step": 1221 + }, + { + "epoch": 1.339178082191781, + "grad_norm": 2.6230008602142334, + "learning_rate": 3.1072298943948015e-06, + "loss": 1.6562, + "step": 1222 + }, + { + "epoch": 1.3402739726027397, + "grad_norm": 2.60079026222229, + "learning_rate": 3.1051990251827784e-06, + "loss": 1.6311, + "step": 1223 + }, + { + "epoch": 1.3413698630136985, + "grad_norm": 2.6757969856262207, + "learning_rate": 3.1031681559707557e-06, + "loss": 1.6673, + "step": 1224 + }, + { + "epoch": 1.3424657534246576, + "grad_norm": 2.6204211711883545, + "learning_rate": 3.101137286758733e-06, + "loss": 1.665, + "step": 1225 + }, + { + "epoch": 1.3435616438356164, + "grad_norm": 2.7115979194641113, + "learning_rate": 3.09910641754671e-06, + "loss": 1.6864, + "step": 1226 + }, + { + "epoch": 1.3446575342465754, + "grad_norm": 2.604799270629883, + "learning_rate": 3.0970755483346875e-06, + "loss": 1.6364, + "step": 1227 + }, + { + "epoch": 1.3457534246575342, + "grad_norm": 2.7463159561157227, + "learning_rate": 3.0950446791226648e-06, + "loss": 1.699, + "step": 1228 + }, + { + "epoch": 1.3468493150684933, + "grad_norm": 2.744046926498413, + "learning_rate": 3.093013809910642e-06, + "loss": 1.665, + "step": 1229 + }, + { + "epoch": 1.347945205479452, + "grad_norm": 2.8082664012908936, + "learning_rate": 3.0909829406986193e-06, + "loss": 1.7027, + "step": 1230 + }, + { + "epoch": 1.347945205479452, + "eval_loss": 1.8549033403396606, + "eval_runtime": 37.1604, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 1230 + }, + { + "epoch": 1.3490410958904109, + "grad_norm": 2.6146464347839355, + "learning_rate": 3.0889520714865966e-06, + "loss": 1.6635, + "step": 1231 + }, + { + "epoch": 1.35013698630137, + "grad_norm": 2.724843978881836, + "learning_rate": 3.086921202274574e-06, + "loss": 1.64, + "step": 1232 + }, + { + "epoch": 1.3512328767123287, + "grad_norm": 2.699645757675171, + "learning_rate": 3.084890333062551e-06, + "loss": 1.5988, + "step": 1233 + }, + { + "epoch": 1.3523287671232875, + "grad_norm": 2.769742727279663, + "learning_rate": 3.0828594638505284e-06, + "loss": 1.6478, + "step": 1234 + }, + { + "epoch": 1.3534246575342466, + "grad_norm": 2.5372848510742188, + "learning_rate": 3.0808285946385057e-06, + "loss": 1.6285, + "step": 1235 + }, + { + "epoch": 1.3545205479452056, + "grad_norm": 2.6461033821105957, + "learning_rate": 3.078797725426483e-06, + "loss": 1.5669, + "step": 1236 + }, + { + "epoch": 1.3556164383561644, + "grad_norm": 2.8221795558929443, + "learning_rate": 3.07676685621446e-06, + "loss": 1.6455, + "step": 1237 + }, + { + "epoch": 1.3567123287671232, + "grad_norm": 2.6035211086273193, + "learning_rate": 3.0747359870024375e-06, + "loss": 1.6063, + "step": 1238 + }, + { + "epoch": 1.3578082191780823, + "grad_norm": 2.533945322036743, + "learning_rate": 3.0727051177904144e-06, + "loss": 1.6234, + "step": 1239 + }, + { + "epoch": 1.358904109589041, + "grad_norm": 2.627448797225952, + "learning_rate": 3.070674248578392e-06, + "loss": 1.6381, + "step": 1240 + }, + { + "epoch": 1.358904109589041, + "eval_loss": 1.8547505140304565, + "eval_runtime": 37.226, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 1240 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 2.6822423934936523, + "learning_rate": 3.068643379366369e-06, + "loss": 1.6436, + "step": 1241 + }, + { + "epoch": 1.361095890410959, + "grad_norm": 2.647813558578491, + "learning_rate": 3.0666125101543466e-06, + "loss": 1.5901, + "step": 1242 + }, + { + "epoch": 1.362191780821918, + "grad_norm": 2.619610548019409, + "learning_rate": 3.0645816409423235e-06, + "loss": 1.6559, + "step": 1243 + }, + { + "epoch": 1.3632876712328768, + "grad_norm": 2.7284183502197266, + "learning_rate": 3.062550771730301e-06, + "loss": 1.6167, + "step": 1244 + }, + { + "epoch": 1.3643835616438356, + "grad_norm": 2.585660934448242, + "learning_rate": 3.060519902518278e-06, + "loss": 1.6176, + "step": 1245 + }, + { + "epoch": 1.3654794520547946, + "grad_norm": 2.689528465270996, + "learning_rate": 3.0584890333062557e-06, + "loss": 1.5651, + "step": 1246 + }, + { + "epoch": 1.3665753424657534, + "grad_norm": 2.484328508377075, + "learning_rate": 3.0564581640942326e-06, + "loss": 1.6487, + "step": 1247 + }, + { + "epoch": 1.3676712328767122, + "grad_norm": 2.629913806915283, + "learning_rate": 3.0544272948822094e-06, + "loss": 1.6099, + "step": 1248 + }, + { + "epoch": 1.3687671232876713, + "grad_norm": 2.711730480194092, + "learning_rate": 3.052396425670187e-06, + "loss": 1.6702, + "step": 1249 + }, + { + "epoch": 1.36986301369863, + "grad_norm": 2.590545415878296, + "learning_rate": 3.050365556458164e-06, + "loss": 1.6078, + "step": 1250 + }, + { + "epoch": 1.36986301369863, + "eval_loss": 1.8545923233032227, + "eval_runtime": 37.2534, + "eval_samples_per_second": 26.843, + "eval_steps_per_second": 1.127, + "step": 1250 + }, + { + "epoch": 1.370958904109589, + "grad_norm": 2.5952389240264893, + "learning_rate": 3.0483346872461417e-06, + "loss": 1.6084, + "step": 1251 + }, + { + "epoch": 1.372054794520548, + "grad_norm": 2.739316940307617, + "learning_rate": 3.0463038180341185e-06, + "loss": 1.6427, + "step": 1252 + }, + { + "epoch": 1.373150684931507, + "grad_norm": 2.811095952987671, + "learning_rate": 3.0442729488220962e-06, + "loss": 1.6817, + "step": 1253 + }, + { + "epoch": 1.3742465753424657, + "grad_norm": 2.606715440750122, + "learning_rate": 3.042242079610073e-06, + "loss": 1.6695, + "step": 1254 + }, + { + "epoch": 1.3753424657534246, + "grad_norm": 2.481234550476074, + "learning_rate": 3.0402112103980508e-06, + "loss": 1.5762, + "step": 1255 + }, + { + "epoch": 1.3764383561643836, + "grad_norm": 2.508068084716797, + "learning_rate": 3.0381803411860276e-06, + "loss": 1.612, + "step": 1256 + }, + { + "epoch": 1.3775342465753424, + "grad_norm": 2.5694284439086914, + "learning_rate": 3.0361494719740053e-06, + "loss": 1.6457, + "step": 1257 + }, + { + "epoch": 1.3786301369863014, + "grad_norm": 2.7029778957366943, + "learning_rate": 3.034118602761982e-06, + "loss": 1.6201, + "step": 1258 + }, + { + "epoch": 1.3797260273972602, + "grad_norm": 2.578904867172241, + "learning_rate": 3.03208773354996e-06, + "loss": 1.5978, + "step": 1259 + }, + { + "epoch": 1.3808219178082193, + "grad_norm": 2.780586004257202, + "learning_rate": 3.0300568643379367e-06, + "loss": 1.6793, + "step": 1260 + }, + { + "epoch": 1.3808219178082193, + "eval_loss": 1.8522270917892456, + "eval_runtime": 37.2388, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 1260 + }, + { + "epoch": 1.381917808219178, + "grad_norm": 2.648653984069824, + "learning_rate": 3.028025995125914e-06, + "loss": 1.6626, + "step": 1261 + }, + { + "epoch": 1.383013698630137, + "grad_norm": 2.631589651107788, + "learning_rate": 3.0259951259138913e-06, + "loss": 1.5822, + "step": 1262 + }, + { + "epoch": 1.384109589041096, + "grad_norm": 2.5863327980041504, + "learning_rate": 3.0239642567018686e-06, + "loss": 1.6322, + "step": 1263 + }, + { + "epoch": 1.3852054794520547, + "grad_norm": 2.6556999683380127, + "learning_rate": 3.021933387489846e-06, + "loss": 1.5727, + "step": 1264 + }, + { + "epoch": 1.3863013698630138, + "grad_norm": 2.6391682624816895, + "learning_rate": 3.019902518277823e-06, + "loss": 1.6334, + "step": 1265 + }, + { + "epoch": 1.3873972602739726, + "grad_norm": 2.6488680839538574, + "learning_rate": 3.0178716490658004e-06, + "loss": 1.6588, + "step": 1266 + }, + { + "epoch": 1.3884931506849316, + "grad_norm": 2.743858814239502, + "learning_rate": 3.0158407798537777e-06, + "loss": 1.5963, + "step": 1267 + }, + { + "epoch": 1.3895890410958904, + "grad_norm": 2.6499438285827637, + "learning_rate": 3.013809910641755e-06, + "loss": 1.5997, + "step": 1268 + }, + { + "epoch": 1.3906849315068492, + "grad_norm": 2.5450377464294434, + "learning_rate": 3.0117790414297322e-06, + "loss": 1.6122, + "step": 1269 + }, + { + "epoch": 1.3917808219178083, + "grad_norm": 2.592026948928833, + "learning_rate": 3.0097481722177095e-06, + "loss": 1.6362, + "step": 1270 + }, + { + "epoch": 1.3917808219178083, + "eval_loss": 1.84926176071167, + "eval_runtime": 37.2501, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 1270 + }, + { + "epoch": 1.392876712328767, + "grad_norm": 2.5699622631073, + "learning_rate": 3.0077173030056868e-06, + "loss": 1.6055, + "step": 1271 + }, + { + "epoch": 1.393972602739726, + "grad_norm": 2.779695987701416, + "learning_rate": 3.005686433793664e-06, + "loss": 1.6503, + "step": 1272 + }, + { + "epoch": 1.395068493150685, + "grad_norm": 2.6080546379089355, + "learning_rate": 3.0036555645816413e-06, + "loss": 1.6121, + "step": 1273 + }, + { + "epoch": 1.396164383561644, + "grad_norm": 2.70405650138855, + "learning_rate": 3.001624695369618e-06, + "loss": 1.6774, + "step": 1274 + }, + { + "epoch": 1.3972602739726028, + "grad_norm": 2.69488525390625, + "learning_rate": 2.999593826157596e-06, + "loss": 1.6186, + "step": 1275 + }, + { + "epoch": 1.3983561643835616, + "grad_norm": 2.9266953468322754, + "learning_rate": 2.9975629569455727e-06, + "loss": 1.6485, + "step": 1276 + }, + { + "epoch": 1.3994520547945206, + "grad_norm": 2.615251302719116, + "learning_rate": 2.9955320877335504e-06, + "loss": 1.624, + "step": 1277 + }, + { + "epoch": 1.4005479452054794, + "grad_norm": 2.574878215789795, + "learning_rate": 2.9935012185215273e-06, + "loss": 1.6637, + "step": 1278 + }, + { + "epoch": 1.4016438356164382, + "grad_norm": 2.6952786445617676, + "learning_rate": 2.991470349309505e-06, + "loss": 1.627, + "step": 1279 + }, + { + "epoch": 1.4027397260273973, + "grad_norm": 2.6076977252960205, + "learning_rate": 2.989439480097482e-06, + "loss": 1.6175, + "step": 1280 + }, + { + "epoch": 1.4027397260273973, + "eval_loss": 1.8475773334503174, + "eval_runtime": 37.2328, + "eval_samples_per_second": 26.858, + "eval_steps_per_second": 1.128, + "step": 1280 + }, + { + "epoch": 1.403835616438356, + "grad_norm": 2.5707366466522217, + "learning_rate": 2.9874086108854595e-06, + "loss": 1.6332, + "step": 1281 + }, + { + "epoch": 1.4049315068493151, + "grad_norm": 2.838609457015991, + "learning_rate": 2.9853777416734364e-06, + "loss": 1.6764, + "step": 1282 + }, + { + "epoch": 1.406027397260274, + "grad_norm": 2.771082878112793, + "learning_rate": 2.983346872461414e-06, + "loss": 1.6622, + "step": 1283 + }, + { + "epoch": 1.407123287671233, + "grad_norm": 2.6113975048065186, + "learning_rate": 2.981316003249391e-06, + "loss": 1.6405, + "step": 1284 + }, + { + "epoch": 1.4082191780821918, + "grad_norm": 2.7537200450897217, + "learning_rate": 2.9792851340373678e-06, + "loss": 1.6376, + "step": 1285 + }, + { + "epoch": 1.4093150684931506, + "grad_norm": 2.683002471923828, + "learning_rate": 2.9772542648253455e-06, + "loss": 1.6679, + "step": 1286 + }, + { + "epoch": 1.4104109589041096, + "grad_norm": 2.651897430419922, + "learning_rate": 2.9752233956133223e-06, + "loss": 1.6091, + "step": 1287 + }, + { + "epoch": 1.4115068493150684, + "grad_norm": 2.6856229305267334, + "learning_rate": 2.9731925264013e-06, + "loss": 1.6223, + "step": 1288 + }, + { + "epoch": 1.4126027397260275, + "grad_norm": 2.646522045135498, + "learning_rate": 2.971161657189277e-06, + "loss": 1.6126, + "step": 1289 + }, + { + "epoch": 1.4136986301369863, + "grad_norm": 2.592247486114502, + "learning_rate": 2.9691307879772546e-06, + "loss": 1.6088, + "step": 1290 + }, + { + "epoch": 1.4136986301369863, + "eval_loss": 1.8501410484313965, + "eval_runtime": 37.225, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 1290 + }, + { + "epoch": 1.4147945205479453, + "grad_norm": 3.03291392326355, + "learning_rate": 2.9670999187652314e-06, + "loss": 1.6051, + "step": 1291 + }, + { + "epoch": 1.415890410958904, + "grad_norm": 2.903224468231201, + "learning_rate": 2.965069049553209e-06, + "loss": 1.6753, + "step": 1292 + }, + { + "epoch": 1.416986301369863, + "grad_norm": 2.5178730487823486, + "learning_rate": 2.963038180341186e-06, + "loss": 1.5894, + "step": 1293 + }, + { + "epoch": 1.418082191780822, + "grad_norm": 2.6278252601623535, + "learning_rate": 2.9610073111291637e-06, + "loss": 1.6687, + "step": 1294 + }, + { + "epoch": 1.4191780821917808, + "grad_norm": 2.8105742931365967, + "learning_rate": 2.9589764419171405e-06, + "loss": 1.6648, + "step": 1295 + }, + { + "epoch": 1.4202739726027398, + "grad_norm": 2.5897178649902344, + "learning_rate": 2.9569455727051182e-06, + "loss": 1.6012, + "step": 1296 + }, + { + "epoch": 1.4213698630136986, + "grad_norm": 2.7172093391418457, + "learning_rate": 2.954914703493095e-06, + "loss": 1.6206, + "step": 1297 + }, + { + "epoch": 1.4224657534246576, + "grad_norm": 2.6845815181732178, + "learning_rate": 2.9528838342810724e-06, + "loss": 1.6163, + "step": 1298 + }, + { + "epoch": 1.4235616438356165, + "grad_norm": 2.652451753616333, + "learning_rate": 2.9508529650690496e-06, + "loss": 1.5916, + "step": 1299 + }, + { + "epoch": 1.4246575342465753, + "grad_norm": 2.740280866622925, + "learning_rate": 2.948822095857027e-06, + "loss": 1.627, + "step": 1300 + }, + { + "epoch": 1.4246575342465753, + "eval_loss": 1.8489265441894531, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 1300 + }, + { + "epoch": 1.4257534246575343, + "grad_norm": 2.786797285079956, + "learning_rate": 2.946791226645004e-06, + "loss": 1.6005, + "step": 1301 + }, + { + "epoch": 1.426849315068493, + "grad_norm": 2.6826601028442383, + "learning_rate": 2.9447603574329815e-06, + "loss": 1.5887, + "step": 1302 + }, + { + "epoch": 1.427945205479452, + "grad_norm": 2.881887912750244, + "learning_rate": 2.9427294882209587e-06, + "loss": 1.6198, + "step": 1303 + }, + { + "epoch": 1.429041095890411, + "grad_norm": 2.86600923538208, + "learning_rate": 2.940698619008936e-06, + "loss": 1.6252, + "step": 1304 + }, + { + "epoch": 1.43013698630137, + "grad_norm": 2.7267069816589355, + "learning_rate": 2.9386677497969133e-06, + "loss": 1.6314, + "step": 1305 + }, + { + "epoch": 1.4312328767123288, + "grad_norm": 2.768399477005005, + "learning_rate": 2.9366368805848906e-06, + "loss": 1.6434, + "step": 1306 + }, + { + "epoch": 1.4323287671232876, + "grad_norm": 2.70622181892395, + "learning_rate": 2.934606011372868e-06, + "loss": 1.6412, + "step": 1307 + }, + { + "epoch": 1.4334246575342466, + "grad_norm": 2.803354501724243, + "learning_rate": 2.932575142160845e-06, + "loss": 1.6402, + "step": 1308 + }, + { + "epoch": 1.4345205479452054, + "grad_norm": 2.8491063117980957, + "learning_rate": 2.9305442729488224e-06, + "loss": 1.6244, + "step": 1309 + }, + { + "epoch": 1.4356164383561643, + "grad_norm": 2.5652339458465576, + "learning_rate": 2.9285134037367997e-06, + "loss": 1.6315, + "step": 1310 + }, + { + "epoch": 1.4356164383561643, + "eval_loss": 1.8461908102035522, + "eval_runtime": 36.9233, + "eval_samples_per_second": 27.083, + "eval_steps_per_second": 1.137, + "step": 1310 + }, + { + "epoch": 1.4367123287671233, + "grad_norm": 2.7153713703155518, + "learning_rate": 2.9264825345247765e-06, + "loss": 1.6714, + "step": 1311 + }, + { + "epoch": 1.4378082191780823, + "grad_norm": 2.779289484024048, + "learning_rate": 2.9244516653127542e-06, + "loss": 1.6355, + "step": 1312 + }, + { + "epoch": 1.4389041095890411, + "grad_norm": 2.6597917079925537, + "learning_rate": 2.922420796100731e-06, + "loss": 1.5574, + "step": 1313 + }, + { + "epoch": 1.44, + "grad_norm": 2.7135894298553467, + "learning_rate": 2.9203899268887088e-06, + "loss": 1.6721, + "step": 1314 + }, + { + "epoch": 1.441095890410959, + "grad_norm": 2.562045097351074, + "learning_rate": 2.9183590576766856e-06, + "loss": 1.5745, + "step": 1315 + }, + { + "epoch": 1.4421917808219178, + "grad_norm": 2.6727097034454346, + "learning_rate": 2.9163281884646633e-06, + "loss": 1.6324, + "step": 1316 + }, + { + "epoch": 1.4432876712328766, + "grad_norm": 2.688309907913208, + "learning_rate": 2.91429731925264e-06, + "loss": 1.6698, + "step": 1317 + }, + { + "epoch": 1.4443835616438356, + "grad_norm": 2.8429853916168213, + "learning_rate": 2.912266450040618e-06, + "loss": 1.6378, + "step": 1318 + }, + { + "epoch": 1.4454794520547944, + "grad_norm": 2.658046245574951, + "learning_rate": 2.9102355808285947e-06, + "loss": 1.6237, + "step": 1319 + }, + { + "epoch": 1.4465753424657535, + "grad_norm": 2.6601390838623047, + "learning_rate": 2.9082047116165724e-06, + "loss": 1.6163, + "step": 1320 + }, + { + "epoch": 1.4465753424657535, + "eval_loss": 1.843716025352478, + "eval_runtime": 37.2087, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 1320 + }, + { + "epoch": 1.4476712328767123, + "grad_norm": 2.949215888977051, + "learning_rate": 2.9061738424045493e-06, + "loss": 1.5883, + "step": 1321 + }, + { + "epoch": 1.4487671232876713, + "grad_norm": 2.823991298675537, + "learning_rate": 2.904142973192527e-06, + "loss": 1.6786, + "step": 1322 + }, + { + "epoch": 1.4498630136986301, + "grad_norm": 2.6177661418914795, + "learning_rate": 2.902112103980504e-06, + "loss": 1.6337, + "step": 1323 + }, + { + "epoch": 1.450958904109589, + "grad_norm": 2.493638038635254, + "learning_rate": 2.900081234768481e-06, + "loss": 1.6328, + "step": 1324 + }, + { + "epoch": 1.452054794520548, + "grad_norm": 2.657914161682129, + "learning_rate": 2.8980503655564584e-06, + "loss": 1.6068, + "step": 1325 + }, + { + "epoch": 1.4531506849315068, + "grad_norm": 2.7800004482269287, + "learning_rate": 2.8960194963444357e-06, + "loss": 1.5963, + "step": 1326 + }, + { + "epoch": 1.4542465753424658, + "grad_norm": 2.7934350967407227, + "learning_rate": 2.893988627132413e-06, + "loss": 1.611, + "step": 1327 + }, + { + "epoch": 1.4553424657534246, + "grad_norm": 2.8139140605926514, + "learning_rate": 2.8919577579203902e-06, + "loss": 1.5621, + "step": 1328 + }, + { + "epoch": 1.4564383561643837, + "grad_norm": 2.5909969806671143, + "learning_rate": 2.8899268887083675e-06, + "loss": 1.6195, + "step": 1329 + }, + { + "epoch": 1.4575342465753425, + "grad_norm": 2.631871461868286, + "learning_rate": 2.8878960194963448e-06, + "loss": 1.6146, + "step": 1330 + }, + { + "epoch": 1.4575342465753425, + "eval_loss": 1.846065878868103, + "eval_runtime": 37.2421, + "eval_samples_per_second": 26.851, + "eval_steps_per_second": 1.128, + "step": 1330 + }, + { + "epoch": 1.4586301369863013, + "grad_norm": 2.667304754257202, + "learning_rate": 2.885865150284322e-06, + "loss": 1.6482, + "step": 1331 + }, + { + "epoch": 1.4597260273972603, + "grad_norm": 2.5922656059265137, + "learning_rate": 2.8838342810722993e-06, + "loss": 1.6231, + "step": 1332 + }, + { + "epoch": 1.4608219178082191, + "grad_norm": 2.663452386856079, + "learning_rate": 2.8818034118602766e-06, + "loss": 1.5772, + "step": 1333 + }, + { + "epoch": 1.4619178082191782, + "grad_norm": 2.6463005542755127, + "learning_rate": 2.879772542648254e-06, + "loss": 1.6333, + "step": 1334 + }, + { + "epoch": 1.463013698630137, + "grad_norm": 2.5469250679016113, + "learning_rate": 2.8777416734362307e-06, + "loss": 1.5465, + "step": 1335 + }, + { + "epoch": 1.464109589041096, + "grad_norm": 2.7316017150878906, + "learning_rate": 2.8757108042242084e-06, + "loss": 1.6856, + "step": 1336 + }, + { + "epoch": 1.4652054794520548, + "grad_norm": 2.7127692699432373, + "learning_rate": 2.8736799350121853e-06, + "loss": 1.667, + "step": 1337 + }, + { + "epoch": 1.4663013698630136, + "grad_norm": 2.7006423473358154, + "learning_rate": 2.871649065800163e-06, + "loss": 1.677, + "step": 1338 + }, + { + "epoch": 1.4673972602739727, + "grad_norm": 2.788804054260254, + "learning_rate": 2.86961819658814e-06, + "loss": 1.6092, + "step": 1339 + }, + { + "epoch": 1.4684931506849315, + "grad_norm": 2.7734181880950928, + "learning_rate": 2.8675873273761175e-06, + "loss": 1.6377, + "step": 1340 + }, + { + "epoch": 1.4684931506849315, + "eval_loss": 1.8444476127624512, + "eval_runtime": 37.2371, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 1340 + }, + { + "epoch": 1.4695890410958903, + "grad_norm": 2.6005876064300537, + "learning_rate": 2.8655564581640944e-06, + "loss": 1.6072, + "step": 1341 + }, + { + "epoch": 1.4706849315068493, + "grad_norm": 2.684293270111084, + "learning_rate": 2.863525588952072e-06, + "loss": 1.6669, + "step": 1342 + }, + { + "epoch": 1.4717808219178083, + "grad_norm": 2.6160528659820557, + "learning_rate": 2.861494719740049e-06, + "loss": 1.5439, + "step": 1343 + }, + { + "epoch": 1.4728767123287672, + "grad_norm": 2.68396258354187, + "learning_rate": 2.8594638505280266e-06, + "loss": 1.5909, + "step": 1344 + }, + { + "epoch": 1.473972602739726, + "grad_norm": 2.6292917728424072, + "learning_rate": 2.8574329813160035e-06, + "loss": 1.6225, + "step": 1345 + }, + { + "epoch": 1.475068493150685, + "grad_norm": 2.702333450317383, + "learning_rate": 2.855402112103981e-06, + "loss": 1.6224, + "step": 1346 + }, + { + "epoch": 1.4761643835616438, + "grad_norm": 2.850862503051758, + "learning_rate": 2.853371242891958e-06, + "loss": 1.5995, + "step": 1347 + }, + { + "epoch": 1.4772602739726026, + "grad_norm": 2.861351490020752, + "learning_rate": 2.851340373679935e-06, + "loss": 1.6254, + "step": 1348 + }, + { + "epoch": 1.4783561643835617, + "grad_norm": 2.7731313705444336, + "learning_rate": 2.8493095044679126e-06, + "loss": 1.6555, + "step": 1349 + }, + { + "epoch": 1.4794520547945205, + "grad_norm": 2.669189691543579, + "learning_rate": 2.8472786352558894e-06, + "loss": 1.6238, + "step": 1350 + }, + { + "epoch": 1.4794520547945205, + "eval_loss": 1.845217227935791, + "eval_runtime": 37.2809, + "eval_samples_per_second": 26.823, + "eval_steps_per_second": 1.127, + "step": 1350 + }, + { + "epoch": 1.4805479452054795, + "grad_norm": 2.64485239982605, + "learning_rate": 2.845247766043867e-06, + "loss": 1.5989, + "step": 1351 + }, + { + "epoch": 1.4816438356164383, + "grad_norm": 2.7558820247650146, + "learning_rate": 2.843216896831844e-06, + "loss": 1.5615, + "step": 1352 + }, + { + "epoch": 1.4827397260273973, + "grad_norm": 2.7818500995635986, + "learning_rate": 2.8411860276198217e-06, + "loss": 1.5962, + "step": 1353 + }, + { + "epoch": 1.4838356164383562, + "grad_norm": 2.571547746658325, + "learning_rate": 2.8391551584077985e-06, + "loss": 1.6434, + "step": 1354 + }, + { + "epoch": 1.484931506849315, + "grad_norm": 2.8242406845092773, + "learning_rate": 2.8371242891957762e-06, + "loss": 1.6537, + "step": 1355 + }, + { + "epoch": 1.486027397260274, + "grad_norm": 2.5545597076416016, + "learning_rate": 2.835093419983753e-06, + "loss": 1.6158, + "step": 1356 + }, + { + "epoch": 1.4871232876712328, + "grad_norm": 2.7329461574554443, + "learning_rate": 2.833062550771731e-06, + "loss": 1.591, + "step": 1357 + }, + { + "epoch": 1.4882191780821918, + "grad_norm": 2.5414955615997314, + "learning_rate": 2.8310316815597076e-06, + "loss": 1.5727, + "step": 1358 + }, + { + "epoch": 1.4893150684931507, + "grad_norm": 2.7625439167022705, + "learning_rate": 2.8290008123476853e-06, + "loss": 1.6343, + "step": 1359 + }, + { + "epoch": 1.4904109589041097, + "grad_norm": 2.7839064598083496, + "learning_rate": 2.826969943135662e-06, + "loss": 1.6187, + "step": 1360 + }, + { + "epoch": 1.4904109589041097, + "eval_loss": 1.8442766666412354, + "eval_runtime": 37.2101, + "eval_samples_per_second": 26.874, + "eval_steps_per_second": 1.129, + "step": 1360 + }, + { + "epoch": 1.4915068493150685, + "grad_norm": 2.615706443786621, + "learning_rate": 2.8249390739236395e-06, + "loss": 1.621, + "step": 1361 + }, + { + "epoch": 1.4926027397260273, + "grad_norm": 2.6295979022979736, + "learning_rate": 2.8229082047116167e-06, + "loss": 1.5913, + "step": 1362 + }, + { + "epoch": 1.4936986301369863, + "grad_norm": 2.7428646087646484, + "learning_rate": 2.820877335499594e-06, + "loss": 1.6015, + "step": 1363 + }, + { + "epoch": 1.4947945205479451, + "grad_norm": 2.5353381633758545, + "learning_rate": 2.8188464662875713e-06, + "loss": 1.5798, + "step": 1364 + }, + { + "epoch": 1.4958904109589042, + "grad_norm": 2.6209659576416016, + "learning_rate": 2.8168155970755486e-06, + "loss": 1.5783, + "step": 1365 + }, + { + "epoch": 1.496986301369863, + "grad_norm": 2.791372299194336, + "learning_rate": 2.814784727863526e-06, + "loss": 1.6082, + "step": 1366 + }, + { + "epoch": 1.498082191780822, + "grad_norm": 2.781949520111084, + "learning_rate": 2.812753858651503e-06, + "loss": 1.6337, + "step": 1367 + }, + { + "epoch": 1.4991780821917808, + "grad_norm": 3.0437769889831543, + "learning_rate": 2.8107229894394804e-06, + "loss": 1.5969, + "step": 1368 + }, + { + "epoch": 1.5002739726027396, + "grad_norm": 2.7201249599456787, + "learning_rate": 2.8086921202274577e-06, + "loss": 1.5717, + "step": 1369 + }, + { + "epoch": 1.5013698630136987, + "grad_norm": 2.766634464263916, + "learning_rate": 2.806661251015435e-06, + "loss": 1.6654, + "step": 1370 + }, + { + "epoch": 1.5013698630136987, + "eval_loss": 1.8452297449111938, + "eval_runtime": 37.2452, + "eval_samples_per_second": 26.849, + "eval_steps_per_second": 1.128, + "step": 1370 + }, + { + "epoch": 1.5024657534246575, + "grad_norm": 2.726876735687256, + "learning_rate": 2.8046303818034122e-06, + "loss": 1.6652, + "step": 1371 + }, + { + "epoch": 1.5035616438356163, + "grad_norm": 2.775012969970703, + "learning_rate": 2.802599512591389e-06, + "loss": 1.7293, + "step": 1372 + }, + { + "epoch": 1.5046575342465753, + "grad_norm": 2.7603225708007812, + "learning_rate": 2.8005686433793668e-06, + "loss": 1.6226, + "step": 1373 + }, + { + "epoch": 1.5057534246575344, + "grad_norm": 2.723904609680176, + "learning_rate": 2.7985377741673436e-06, + "loss": 1.5663, + "step": 1374 + }, + { + "epoch": 1.5068493150684932, + "grad_norm": 2.667384624481201, + "learning_rate": 2.7965069049553213e-06, + "loss": 1.6209, + "step": 1375 + }, + { + "epoch": 1.507945205479452, + "grad_norm": 2.7580318450927734, + "learning_rate": 2.794476035743298e-06, + "loss": 1.6705, + "step": 1376 + }, + { + "epoch": 1.509041095890411, + "grad_norm": 2.9216315746307373, + "learning_rate": 2.792445166531276e-06, + "loss": 1.6361, + "step": 1377 + }, + { + "epoch": 1.5101369863013698, + "grad_norm": 2.6552748680114746, + "learning_rate": 2.7904142973192527e-06, + "loss": 1.6498, + "step": 1378 + }, + { + "epoch": 1.5112328767123286, + "grad_norm": 2.764836549758911, + "learning_rate": 2.7883834281072304e-06, + "loss": 1.6009, + "step": 1379 + }, + { + "epoch": 1.5123287671232877, + "grad_norm": 2.5660064220428467, + "learning_rate": 2.7863525588952073e-06, + "loss": 1.576, + "step": 1380 + }, + { + "epoch": 1.5123287671232877, + "eval_loss": 1.8392945528030396, + "eval_runtime": 37.2515, + "eval_samples_per_second": 26.845, + "eval_steps_per_second": 1.127, + "step": 1380 + }, + { + "epoch": 1.5134246575342467, + "grad_norm": 2.5110044479370117, + "learning_rate": 2.784321689683185e-06, + "loss": 1.5744, + "step": 1381 + }, + { + "epoch": 1.5145205479452055, + "grad_norm": 2.7352850437164307, + "learning_rate": 2.782290820471162e-06, + "loss": 1.5852, + "step": 1382 + }, + { + "epoch": 1.5156164383561643, + "grad_norm": 2.850947618484497, + "learning_rate": 2.7802599512591395e-06, + "loss": 1.6692, + "step": 1383 + }, + { + "epoch": 1.5167123287671234, + "grad_norm": 2.7107245922088623, + "learning_rate": 2.7782290820471164e-06, + "loss": 1.6427, + "step": 1384 + }, + { + "epoch": 1.5178082191780822, + "grad_norm": 2.8236117362976074, + "learning_rate": 2.7761982128350932e-06, + "loss": 1.6375, + "step": 1385 + }, + { + "epoch": 1.518904109589041, + "grad_norm": 2.763416051864624, + "learning_rate": 2.774167343623071e-06, + "loss": 1.6864, + "step": 1386 + }, + { + "epoch": 1.52, + "grad_norm": 2.7317910194396973, + "learning_rate": 2.772136474411048e-06, + "loss": 1.6015, + "step": 1387 + }, + { + "epoch": 1.521095890410959, + "grad_norm": 2.7454333305358887, + "learning_rate": 2.7701056051990255e-06, + "loss": 1.5379, + "step": 1388 + }, + { + "epoch": 1.5221917808219179, + "grad_norm": 2.6102535724639893, + "learning_rate": 2.7680747359870023e-06, + "loss": 1.6282, + "step": 1389 + }, + { + "epoch": 1.5232876712328767, + "grad_norm": 2.658440113067627, + "learning_rate": 2.76604386677498e-06, + "loss": 1.6042, + "step": 1390 + }, + { + "epoch": 1.5232876712328767, + "eval_loss": 1.8391484022140503, + "eval_runtime": 37.1746, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1390 + }, + { + "epoch": 1.5243835616438357, + "grad_norm": 2.5731043815612793, + "learning_rate": 2.764012997562957e-06, + "loss": 1.6446, + "step": 1391 + }, + { + "epoch": 1.5254794520547945, + "grad_norm": 2.680983543395996, + "learning_rate": 2.7619821283509346e-06, + "loss": 1.5663, + "step": 1392 + }, + { + "epoch": 1.5265753424657533, + "grad_norm": 2.7525506019592285, + "learning_rate": 2.7599512591389114e-06, + "loss": 1.6628, + "step": 1393 + }, + { + "epoch": 1.5276712328767124, + "grad_norm": 2.6905839443206787, + "learning_rate": 2.757920389926889e-06, + "loss": 1.6021, + "step": 1394 + }, + { + "epoch": 1.5287671232876714, + "grad_norm": 2.850900888442993, + "learning_rate": 2.755889520714866e-06, + "loss": 1.5984, + "step": 1395 + }, + { + "epoch": 1.52986301369863, + "grad_norm": 2.6612954139709473, + "learning_rate": 2.7538586515028437e-06, + "loss": 1.5627, + "step": 1396 + }, + { + "epoch": 1.530958904109589, + "grad_norm": 2.632106304168701, + "learning_rate": 2.7518277822908205e-06, + "loss": 1.6304, + "step": 1397 + }, + { + "epoch": 1.532054794520548, + "grad_norm": 2.660377025604248, + "learning_rate": 2.749796913078798e-06, + "loss": 1.6622, + "step": 1398 + }, + { + "epoch": 1.5331506849315069, + "grad_norm": 2.497938394546509, + "learning_rate": 2.747766043866775e-06, + "loss": 1.6152, + "step": 1399 + }, + { + "epoch": 1.5342465753424657, + "grad_norm": 2.655977725982666, + "learning_rate": 2.7457351746547524e-06, + "loss": 1.6703, + "step": 1400 + }, + { + "epoch": 1.5342465753424657, + "eval_loss": 1.8361152410507202, + "eval_runtime": 37.2147, + "eval_samples_per_second": 26.871, + "eval_steps_per_second": 1.129, + "step": 1400 + }, + { + "epoch": 1.5353424657534247, + "grad_norm": 2.664874792098999, + "learning_rate": 2.7437043054427297e-06, + "loss": 1.7109, + "step": 1401 + }, + { + "epoch": 1.5364383561643835, + "grad_norm": 2.5773956775665283, + "learning_rate": 2.741673436230707e-06, + "loss": 1.6176, + "step": 1402 + }, + { + "epoch": 1.5375342465753423, + "grad_norm": 2.7094814777374268, + "learning_rate": 2.739642567018684e-06, + "loss": 1.6075, + "step": 1403 + }, + { + "epoch": 1.5386301369863014, + "grad_norm": 2.652325391769409, + "learning_rate": 2.7376116978066615e-06, + "loss": 1.6189, + "step": 1404 + }, + { + "epoch": 1.5397260273972604, + "grad_norm": 2.6892807483673096, + "learning_rate": 2.7355808285946388e-06, + "loss": 1.6122, + "step": 1405 + }, + { + "epoch": 1.5408219178082192, + "grad_norm": 2.6621243953704834, + "learning_rate": 2.733549959382616e-06, + "loss": 1.6008, + "step": 1406 + }, + { + "epoch": 1.541917808219178, + "grad_norm": 2.773747444152832, + "learning_rate": 2.7315190901705933e-06, + "loss": 1.6723, + "step": 1407 + }, + { + "epoch": 1.543013698630137, + "grad_norm": 2.853525400161743, + "learning_rate": 2.7294882209585706e-06, + "loss": 1.641, + "step": 1408 + }, + { + "epoch": 1.5441095890410959, + "grad_norm": 2.6333844661712646, + "learning_rate": 2.727457351746548e-06, + "loss": 1.587, + "step": 1409 + }, + { + "epoch": 1.5452054794520547, + "grad_norm": 2.6299996376037598, + "learning_rate": 2.725426482534525e-06, + "loss": 1.5807, + "step": 1410 + }, + { + "epoch": 1.5452054794520547, + "eval_loss": 1.8377478122711182, + "eval_runtime": 36.8958, + "eval_samples_per_second": 27.103, + "eval_steps_per_second": 1.138, + "step": 1410 + }, + { + "epoch": 1.5463013698630137, + "grad_norm": 2.6222734451293945, + "learning_rate": 2.723395613322502e-06, + "loss": 1.6225, + "step": 1411 + }, + { + "epoch": 1.5473972602739727, + "grad_norm": 2.6702005863189697, + "learning_rate": 2.7213647441104797e-06, + "loss": 1.6031, + "step": 1412 + }, + { + "epoch": 1.5484931506849315, + "grad_norm": 2.6451900005340576, + "learning_rate": 2.7193338748984565e-06, + "loss": 1.5998, + "step": 1413 + }, + { + "epoch": 1.5495890410958904, + "grad_norm": 2.769784927368164, + "learning_rate": 2.7173030056864342e-06, + "loss": 1.7579, + "step": 1414 + }, + { + "epoch": 1.5506849315068494, + "grad_norm": 2.629871368408203, + "learning_rate": 2.715272136474411e-06, + "loss": 1.564, + "step": 1415 + }, + { + "epoch": 1.5517808219178082, + "grad_norm": 2.811937093734741, + "learning_rate": 2.713241267262389e-06, + "loss": 1.6236, + "step": 1416 + }, + { + "epoch": 1.552876712328767, + "grad_norm": 2.883277654647827, + "learning_rate": 2.7112103980503656e-06, + "loss": 1.5996, + "step": 1417 + }, + { + "epoch": 1.553972602739726, + "grad_norm": 2.610964059829712, + "learning_rate": 2.7091795288383433e-06, + "loss": 1.6391, + "step": 1418 + }, + { + "epoch": 1.555068493150685, + "grad_norm": 2.6637415885925293, + "learning_rate": 2.70714865962632e-06, + "loss": 1.5983, + "step": 1419 + }, + { + "epoch": 1.5561643835616439, + "grad_norm": 2.5372607707977295, + "learning_rate": 2.705117790414298e-06, + "loss": 1.6258, + "step": 1420 + }, + { + "epoch": 1.5561643835616439, + "eval_loss": 1.838557243347168, + "eval_runtime": 36.973, + "eval_samples_per_second": 27.047, + "eval_steps_per_second": 1.136, + "step": 1420 + }, + { + "epoch": 1.5572602739726027, + "grad_norm": 2.900594711303711, + "learning_rate": 2.7030869212022747e-06, + "loss": 1.67, + "step": 1421 + }, + { + "epoch": 1.5583561643835617, + "grad_norm": 2.4924063682556152, + "learning_rate": 2.7010560519902516e-06, + "loss": 1.5879, + "step": 1422 + }, + { + "epoch": 1.5594520547945205, + "grad_norm": 2.5761115550994873, + "learning_rate": 2.6990251827782293e-06, + "loss": 1.6287, + "step": 1423 + }, + { + "epoch": 1.5605479452054793, + "grad_norm": 2.645463466644287, + "learning_rate": 2.696994313566206e-06, + "loss": 1.5925, + "step": 1424 + }, + { + "epoch": 1.5616438356164384, + "grad_norm": 2.682466506958008, + "learning_rate": 2.694963444354184e-06, + "loss": 1.6723, + "step": 1425 + }, + { + "epoch": 1.5627397260273974, + "grad_norm": 2.499749183654785, + "learning_rate": 2.6929325751421607e-06, + "loss": 1.5998, + "step": 1426 + }, + { + "epoch": 1.563835616438356, + "grad_norm": 2.659956455230713, + "learning_rate": 2.6909017059301384e-06, + "loss": 1.6311, + "step": 1427 + }, + { + "epoch": 1.564931506849315, + "grad_norm": 2.6684396266937256, + "learning_rate": 2.6888708367181153e-06, + "loss": 1.5648, + "step": 1428 + }, + { + "epoch": 1.566027397260274, + "grad_norm": 2.6868791580200195, + "learning_rate": 2.686839967506093e-06, + "loss": 1.6419, + "step": 1429 + }, + { + "epoch": 1.5671232876712329, + "grad_norm": 2.6999104022979736, + "learning_rate": 2.68480909829407e-06, + "loss": 1.55, + "step": 1430 + }, + { + "epoch": 1.5671232876712329, + "eval_loss": 1.83835768699646, + "eval_runtime": 37.0619, + "eval_samples_per_second": 26.982, + "eval_steps_per_second": 1.133, + "step": 1430 + }, + { + "epoch": 1.5682191780821917, + "grad_norm": 2.7338085174560547, + "learning_rate": 2.6827782290820475e-06, + "loss": 1.5507, + "step": 1431 + }, + { + "epoch": 1.5693150684931507, + "grad_norm": 2.7358195781707764, + "learning_rate": 2.6807473598700244e-06, + "loss": 1.6659, + "step": 1432 + }, + { + "epoch": 1.5704109589041098, + "grad_norm": 2.671983242034912, + "learning_rate": 2.678716490658002e-06, + "loss": 1.6293, + "step": 1433 + }, + { + "epoch": 1.5715068493150683, + "grad_norm": 2.551203489303589, + "learning_rate": 2.676685621445979e-06, + "loss": 1.623, + "step": 1434 + }, + { + "epoch": 1.5726027397260274, + "grad_norm": 2.6485984325408936, + "learning_rate": 2.674654752233956e-06, + "loss": 1.6127, + "step": 1435 + }, + { + "epoch": 1.5736986301369864, + "grad_norm": 2.643105983734131, + "learning_rate": 2.6726238830219335e-06, + "loss": 1.6155, + "step": 1436 + }, + { + "epoch": 1.5747945205479452, + "grad_norm": 2.769458055496216, + "learning_rate": 2.6705930138099107e-06, + "loss": 1.6576, + "step": 1437 + }, + { + "epoch": 1.575890410958904, + "grad_norm": 2.3996448516845703, + "learning_rate": 2.668562144597888e-06, + "loss": 1.5276, + "step": 1438 + }, + { + "epoch": 1.576986301369863, + "grad_norm": 2.5827524662017822, + "learning_rate": 2.6665312753858653e-06, + "loss": 1.6321, + "step": 1439 + }, + { + "epoch": 1.5780821917808219, + "grad_norm": 2.728794813156128, + "learning_rate": 2.6645004061738426e-06, + "loss": 1.6177, + "step": 1440 + }, + { + "epoch": 1.5780821917808219, + "eval_loss": 1.8374123573303223, + "eval_runtime": 37.1748, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1440 + }, + { + "epoch": 1.5791780821917807, + "grad_norm": 2.7500123977661133, + "learning_rate": 2.66246953696182e-06, + "loss": 1.5679, + "step": 1441 + }, + { + "epoch": 1.5802739726027397, + "grad_norm": 2.6778831481933594, + "learning_rate": 2.660438667749797e-06, + "loss": 1.5744, + "step": 1442 + }, + { + "epoch": 1.5813698630136988, + "grad_norm": 2.933138847351074, + "learning_rate": 2.6584077985377744e-06, + "loss": 1.6238, + "step": 1443 + }, + { + "epoch": 1.5824657534246576, + "grad_norm": 2.642246961593628, + "learning_rate": 2.6563769293257517e-06, + "loss": 1.6186, + "step": 1444 + }, + { + "epoch": 1.5835616438356164, + "grad_norm": 2.5519111156463623, + "learning_rate": 2.654346060113729e-06, + "loss": 1.5798, + "step": 1445 + }, + { + "epoch": 1.5846575342465754, + "grad_norm": 2.569063186645508, + "learning_rate": 2.6523151909017062e-06, + "loss": 1.5881, + "step": 1446 + }, + { + "epoch": 1.5857534246575342, + "grad_norm": 2.501680850982666, + "learning_rate": 2.6502843216896835e-06, + "loss": 1.5768, + "step": 1447 + }, + { + "epoch": 1.586849315068493, + "grad_norm": 2.7483339309692383, + "learning_rate": 2.6482534524776603e-06, + "loss": 1.6889, + "step": 1448 + }, + { + "epoch": 1.587945205479452, + "grad_norm": 3.004084587097168, + "learning_rate": 2.646222583265638e-06, + "loss": 1.6565, + "step": 1449 + }, + { + "epoch": 1.589041095890411, + "grad_norm": 2.793058156967163, + "learning_rate": 2.644191714053615e-06, + "loss": 1.6246, + "step": 1450 + }, + { + "epoch": 1.589041095890411, + "eval_loss": 1.83695387840271, + "eval_runtime": 37.1748, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1450 + }, + { + "epoch": 1.59013698630137, + "grad_norm": 2.803190231323242, + "learning_rate": 2.6421608448415926e-06, + "loss": 1.6051, + "step": 1451 + }, + { + "epoch": 1.5912328767123287, + "grad_norm": 2.7721307277679443, + "learning_rate": 2.6401299756295694e-06, + "loss": 1.5712, + "step": 1452 + }, + { + "epoch": 1.5923287671232877, + "grad_norm": 2.9235639572143555, + "learning_rate": 2.638099106417547e-06, + "loss": 1.6197, + "step": 1453 + }, + { + "epoch": 1.5934246575342466, + "grad_norm": 2.609719753265381, + "learning_rate": 2.636068237205524e-06, + "loss": 1.5976, + "step": 1454 + }, + { + "epoch": 1.5945205479452054, + "grad_norm": 2.7927262783050537, + "learning_rate": 2.6340373679935017e-06, + "loss": 1.6844, + "step": 1455 + }, + { + "epoch": 1.5956164383561644, + "grad_norm": 2.8688440322875977, + "learning_rate": 2.6320064987814785e-06, + "loss": 1.5285, + "step": 1456 + }, + { + "epoch": 1.5967123287671234, + "grad_norm": 2.819079637527466, + "learning_rate": 2.6299756295694562e-06, + "loss": 1.6171, + "step": 1457 + }, + { + "epoch": 1.5978082191780822, + "grad_norm": 2.825784206390381, + "learning_rate": 2.627944760357433e-06, + "loss": 1.6656, + "step": 1458 + }, + { + "epoch": 1.598904109589041, + "grad_norm": 2.6807522773742676, + "learning_rate": 2.6259138911454104e-06, + "loss": 1.6839, + "step": 1459 + }, + { + "epoch": 1.6, + "grad_norm": 2.9442570209503174, + "learning_rate": 2.6238830219333877e-06, + "loss": 1.6547, + "step": 1460 + }, + { + "epoch": 1.6, + "eval_loss": 1.8358429670333862, + "eval_runtime": 37.1499, + "eval_samples_per_second": 26.918, + "eval_steps_per_second": 1.131, + "step": 1460 + }, + { + "epoch": 1.601095890410959, + "grad_norm": 3.0025434494018555, + "learning_rate": 2.621852152721365e-06, + "loss": 1.5688, + "step": 1461 + }, + { + "epoch": 1.6021917808219177, + "grad_norm": 3.0999062061309814, + "learning_rate": 2.619821283509342e-06, + "loss": 1.639, + "step": 1462 + }, + { + "epoch": 1.6032876712328767, + "grad_norm": 2.7106330394744873, + "learning_rate": 2.6177904142973195e-06, + "loss": 1.5682, + "step": 1463 + }, + { + "epoch": 1.6043835616438358, + "grad_norm": 2.968259811401367, + "learning_rate": 2.6157595450852968e-06, + "loss": 1.6421, + "step": 1464 + }, + { + "epoch": 1.6054794520547944, + "grad_norm": 2.964728593826294, + "learning_rate": 2.613728675873274e-06, + "loss": 1.6626, + "step": 1465 + }, + { + "epoch": 1.6065753424657534, + "grad_norm": 2.8934402465820312, + "learning_rate": 2.6116978066612513e-06, + "loss": 1.6742, + "step": 1466 + }, + { + "epoch": 1.6076712328767124, + "grad_norm": 2.7678956985473633, + "learning_rate": 2.6096669374492286e-06, + "loss": 1.6139, + "step": 1467 + }, + { + "epoch": 1.6087671232876712, + "grad_norm": 2.74017333984375, + "learning_rate": 2.607636068237206e-06, + "loss": 1.5726, + "step": 1468 + }, + { + "epoch": 1.60986301369863, + "grad_norm": 2.8686470985412598, + "learning_rate": 2.605605199025183e-06, + "loss": 1.638, + "step": 1469 + }, + { + "epoch": 1.610958904109589, + "grad_norm": 3.008751392364502, + "learning_rate": 2.6035743298131604e-06, + "loss": 1.6246, + "step": 1470 + }, + { + "epoch": 1.610958904109589, + "eval_loss": 1.8375154733657837, + "eval_runtime": 37.1808, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 1470 + }, + { + "epoch": 1.612054794520548, + "grad_norm": 3.030925989151001, + "learning_rate": 2.6015434606011377e-06, + "loss": 1.6558, + "step": 1471 + }, + { + "epoch": 1.6131506849315067, + "grad_norm": 3.1855311393737793, + "learning_rate": 2.5995125913891145e-06, + "loss": 1.6488, + "step": 1472 + }, + { + "epoch": 1.6142465753424657, + "grad_norm": 2.943204879760742, + "learning_rate": 2.5974817221770922e-06, + "loss": 1.6211, + "step": 1473 + }, + { + "epoch": 1.6153424657534248, + "grad_norm": 2.681720733642578, + "learning_rate": 2.595450852965069e-06, + "loss": 1.655, + "step": 1474 + }, + { + "epoch": 1.6164383561643836, + "grad_norm": 2.750523567199707, + "learning_rate": 2.593419983753047e-06, + "loss": 1.6624, + "step": 1475 + }, + { + "epoch": 1.6175342465753424, + "grad_norm": 2.8923532962799072, + "learning_rate": 2.5913891145410236e-06, + "loss": 1.6046, + "step": 1476 + }, + { + "epoch": 1.6186301369863014, + "grad_norm": 2.852468490600586, + "learning_rate": 2.5893582453290013e-06, + "loss": 1.6557, + "step": 1477 + }, + { + "epoch": 1.6197260273972602, + "grad_norm": 2.7146172523498535, + "learning_rate": 2.587327376116978e-06, + "loss": 1.5888, + "step": 1478 + }, + { + "epoch": 1.620821917808219, + "grad_norm": 2.7109804153442383, + "learning_rate": 2.585296506904956e-06, + "loss": 1.6049, + "step": 1479 + }, + { + "epoch": 1.621917808219178, + "grad_norm": 2.9429404735565186, + "learning_rate": 2.5832656376929327e-06, + "loss": 1.6046, + "step": 1480 + }, + { + "epoch": 1.621917808219178, + "eval_loss": 1.837095856666565, + "eval_runtime": 37.19, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 1480 + }, + { + "epoch": 1.6230136986301371, + "grad_norm": 2.89047908782959, + "learning_rate": 2.5812347684809104e-06, + "loss": 1.5305, + "step": 1481 + }, + { + "epoch": 1.624109589041096, + "grad_norm": 3.0587005615234375, + "learning_rate": 2.5792038992688873e-06, + "loss": 1.7029, + "step": 1482 + }, + { + "epoch": 1.6252054794520547, + "grad_norm": 2.7209622859954834, + "learning_rate": 2.577173030056865e-06, + "loss": 1.5424, + "step": 1483 + }, + { + "epoch": 1.6263013698630138, + "grad_norm": 2.8400819301605225, + "learning_rate": 2.575142160844842e-06, + "loss": 1.6411, + "step": 1484 + }, + { + "epoch": 1.6273972602739726, + "grad_norm": 2.7427124977111816, + "learning_rate": 2.5731112916328187e-06, + "loss": 1.6421, + "step": 1485 + }, + { + "epoch": 1.6284931506849314, + "grad_norm": 2.6996700763702393, + "learning_rate": 2.5710804224207964e-06, + "loss": 1.5845, + "step": 1486 + }, + { + "epoch": 1.6295890410958904, + "grad_norm": 2.540188789367676, + "learning_rate": 2.5690495532087733e-06, + "loss": 1.5472, + "step": 1487 + }, + { + "epoch": 1.6306849315068495, + "grad_norm": 2.7682793140411377, + "learning_rate": 2.567018683996751e-06, + "loss": 1.6347, + "step": 1488 + }, + { + "epoch": 1.6317808219178083, + "grad_norm": 2.760413646697998, + "learning_rate": 2.564987814784728e-06, + "loss": 1.5826, + "step": 1489 + }, + { + "epoch": 1.632876712328767, + "grad_norm": 2.9961049556732178, + "learning_rate": 2.5629569455727055e-06, + "loss": 1.6037, + "step": 1490 + }, + { + "epoch": 1.632876712328767, + "eval_loss": 1.8322792053222656, + "eval_runtime": 37.2227, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 1490 + }, + { + "epoch": 1.633972602739726, + "grad_norm": 2.6983907222747803, + "learning_rate": 2.5609260763606824e-06, + "loss": 1.6142, + "step": 1491 + }, + { + "epoch": 1.635068493150685, + "grad_norm": 2.867401599884033, + "learning_rate": 2.55889520714866e-06, + "loss": 1.7227, + "step": 1492 + }, + { + "epoch": 1.6361643835616437, + "grad_norm": 2.7735934257507324, + "learning_rate": 2.556864337936637e-06, + "loss": 1.5856, + "step": 1493 + }, + { + "epoch": 1.6372602739726028, + "grad_norm": 2.7661404609680176, + "learning_rate": 2.5548334687246146e-06, + "loss": 1.6084, + "step": 1494 + }, + { + "epoch": 1.6383561643835618, + "grad_norm": 2.6881582736968994, + "learning_rate": 2.5528025995125915e-06, + "loss": 1.659, + "step": 1495 + }, + { + "epoch": 1.6394520547945204, + "grad_norm": 2.5745224952697754, + "learning_rate": 2.550771730300569e-06, + "loss": 1.6131, + "step": 1496 + }, + { + "epoch": 1.6405479452054794, + "grad_norm": 2.736159563064575, + "learning_rate": 2.548740861088546e-06, + "loss": 1.6469, + "step": 1497 + }, + { + "epoch": 1.6416438356164385, + "grad_norm": 2.6255311965942383, + "learning_rate": 2.5467099918765233e-06, + "loss": 1.5452, + "step": 1498 + }, + { + "epoch": 1.6427397260273973, + "grad_norm": 2.6919801235198975, + "learning_rate": 2.5446791226645006e-06, + "loss": 1.5282, + "step": 1499 + }, + { + "epoch": 1.643835616438356, + "grad_norm": 2.641533136367798, + "learning_rate": 2.542648253452478e-06, + "loss": 1.6524, + "step": 1500 + }, + { + "epoch": 1.643835616438356, + "eval_loss": 1.8287466764450073, + "eval_runtime": 37.2305, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 1500 + }, + { + "epoch": 1.644931506849315, + "grad_norm": 2.7633602619171143, + "learning_rate": 2.540617384240455e-06, + "loss": 1.6199, + "step": 1501 + }, + { + "epoch": 1.6460273972602741, + "grad_norm": 2.618852376937866, + "learning_rate": 2.5385865150284324e-06, + "loss": 1.5803, + "step": 1502 + }, + { + "epoch": 1.6471232876712327, + "grad_norm": 2.857733964920044, + "learning_rate": 2.5365556458164097e-06, + "loss": 1.5968, + "step": 1503 + }, + { + "epoch": 1.6482191780821918, + "grad_norm": 2.8334357738494873, + "learning_rate": 2.534524776604387e-06, + "loss": 1.611, + "step": 1504 + }, + { + "epoch": 1.6493150684931508, + "grad_norm": 2.638340950012207, + "learning_rate": 2.5324939073923642e-06, + "loss": 1.5048, + "step": 1505 + }, + { + "epoch": 1.6504109589041096, + "grad_norm": 2.6521124839782715, + "learning_rate": 2.5304630381803415e-06, + "loss": 1.5967, + "step": 1506 + }, + { + "epoch": 1.6515068493150684, + "grad_norm": 2.639162540435791, + "learning_rate": 2.5284321689683188e-06, + "loss": 1.6084, + "step": 1507 + }, + { + "epoch": 1.6526027397260274, + "grad_norm": 2.6480295658111572, + "learning_rate": 2.526401299756296e-06, + "loss": 1.6115, + "step": 1508 + }, + { + "epoch": 1.6536986301369863, + "grad_norm": 2.7809157371520996, + "learning_rate": 2.524370430544273e-06, + "loss": 1.5909, + "step": 1509 + }, + { + "epoch": 1.654794520547945, + "grad_norm": 2.6947197914123535, + "learning_rate": 2.5223395613322506e-06, + "loss": 1.575, + "step": 1510 + }, + { + "epoch": 1.654794520547945, + "eval_loss": 1.8312958478927612, + "eval_runtime": 36.8551, + "eval_samples_per_second": 27.133, + "eval_steps_per_second": 1.14, + "step": 1510 + }, + { + "epoch": 1.655890410958904, + "grad_norm": 2.7290775775909424, + "learning_rate": 2.5203086921202274e-06, + "loss": 1.6169, + "step": 1511 + }, + { + "epoch": 1.6569863013698631, + "grad_norm": 2.7147257328033447, + "learning_rate": 2.518277822908205e-06, + "loss": 1.589, + "step": 1512 + }, + { + "epoch": 1.658082191780822, + "grad_norm": 2.648577928543091, + "learning_rate": 2.516246953696182e-06, + "loss": 1.6113, + "step": 1513 + }, + { + "epoch": 1.6591780821917808, + "grad_norm": 2.8041188716888428, + "learning_rate": 2.5142160844841597e-06, + "loss": 1.6128, + "step": 1514 + }, + { + "epoch": 1.6602739726027398, + "grad_norm": 2.7862393856048584, + "learning_rate": 2.5121852152721365e-06, + "loss": 1.6227, + "step": 1515 + }, + { + "epoch": 1.6613698630136986, + "grad_norm": 2.698709487915039, + "learning_rate": 2.5101543460601142e-06, + "loss": 1.6169, + "step": 1516 + }, + { + "epoch": 1.6624657534246574, + "grad_norm": 2.6966660022735596, + "learning_rate": 2.508123476848091e-06, + "loss": 1.5522, + "step": 1517 + }, + { + "epoch": 1.6635616438356164, + "grad_norm": 2.86776065826416, + "learning_rate": 2.506092607636069e-06, + "loss": 1.636, + "step": 1518 + }, + { + "epoch": 1.6646575342465755, + "grad_norm": 2.835958242416382, + "learning_rate": 2.5040617384240457e-06, + "loss": 1.6021, + "step": 1519 + }, + { + "epoch": 1.6657534246575343, + "grad_norm": 2.645120143890381, + "learning_rate": 2.5020308692120234e-06, + "loss": 1.5842, + "step": 1520 + }, + { + "epoch": 1.6657534246575343, + "eval_loss": 1.8320025205612183, + "eval_runtime": 37.0745, + "eval_samples_per_second": 26.973, + "eval_steps_per_second": 1.133, + "step": 1520 + }, + { + "epoch": 1.666849315068493, + "grad_norm": 2.6796860694885254, + "learning_rate": 2.5e-06, + "loss": 1.6579, + "step": 1521 + }, + { + "epoch": 1.6679452054794521, + "grad_norm": 2.824993133544922, + "learning_rate": 2.4979691307879775e-06, + "loss": 1.5842, + "step": 1522 + }, + { + "epoch": 1.669041095890411, + "grad_norm": 2.554858922958374, + "learning_rate": 2.4959382615759548e-06, + "loss": 1.5956, + "step": 1523 + }, + { + "epoch": 1.6701369863013698, + "grad_norm": 2.621208429336548, + "learning_rate": 2.493907392363932e-06, + "loss": 1.6622, + "step": 1524 + }, + { + "epoch": 1.6712328767123288, + "grad_norm": 2.711733818054199, + "learning_rate": 2.4918765231519093e-06, + "loss": 1.6059, + "step": 1525 + }, + { + "epoch": 1.6723287671232878, + "grad_norm": 2.6141200065612793, + "learning_rate": 2.4898456539398866e-06, + "loss": 1.6105, + "step": 1526 + }, + { + "epoch": 1.6734246575342466, + "grad_norm": 2.693882942199707, + "learning_rate": 2.487814784727864e-06, + "loss": 1.6675, + "step": 1527 + }, + { + "epoch": 1.6745205479452054, + "grad_norm": 2.71944260597229, + "learning_rate": 2.4857839155158407e-06, + "loss": 1.5964, + "step": 1528 + }, + { + "epoch": 1.6756164383561645, + "grad_norm": 2.6505935192108154, + "learning_rate": 2.483753046303818e-06, + "loss": 1.5966, + "step": 1529 + }, + { + "epoch": 1.6767123287671233, + "grad_norm": 2.710515022277832, + "learning_rate": 2.4817221770917953e-06, + "loss": 1.6025, + "step": 1530 + }, + { + "epoch": 1.6767123287671233, + "eval_loss": 1.8318896293640137, + "eval_runtime": 37.2071, + "eval_samples_per_second": 26.877, + "eval_steps_per_second": 1.129, + "step": 1530 + }, + { + "epoch": 1.677808219178082, + "grad_norm": 2.8350000381469727, + "learning_rate": 2.4796913078797725e-06, + "loss": 1.6357, + "step": 1531 + }, + { + "epoch": 1.6789041095890411, + "grad_norm": 2.641923427581787, + "learning_rate": 2.47766043866775e-06, + "loss": 1.6338, + "step": 1532 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 2.6234304904937744, + "learning_rate": 2.475629569455727e-06, + "loss": 1.6422, + "step": 1533 + }, + { + "epoch": 1.6810958904109587, + "grad_norm": 2.7153754234313965, + "learning_rate": 2.4735987002437044e-06, + "loss": 1.5758, + "step": 1534 + }, + { + "epoch": 1.6821917808219178, + "grad_norm": 2.6731090545654297, + "learning_rate": 2.4715678310316816e-06, + "loss": 1.5684, + "step": 1535 + }, + { + "epoch": 1.6832876712328768, + "grad_norm": 2.64077091217041, + "learning_rate": 2.469536961819659e-06, + "loss": 1.6039, + "step": 1536 + }, + { + "epoch": 1.6843835616438356, + "grad_norm": 2.522536516189575, + "learning_rate": 2.467506092607636e-06, + "loss": 1.6054, + "step": 1537 + }, + { + "epoch": 1.6854794520547944, + "grad_norm": 2.5827460289001465, + "learning_rate": 2.4654752233956135e-06, + "loss": 1.5705, + "step": 1538 + }, + { + "epoch": 1.6865753424657535, + "grad_norm": 2.7648563385009766, + "learning_rate": 2.4634443541835907e-06, + "loss": 1.6657, + "step": 1539 + }, + { + "epoch": 1.6876712328767123, + "grad_norm": 2.6262803077697754, + "learning_rate": 2.461413484971568e-06, + "loss": 1.6021, + "step": 1540 + }, + { + "epoch": 1.6876712328767123, + "eval_loss": 1.8337559700012207, + "eval_runtime": 37.2695, + "eval_samples_per_second": 26.832, + "eval_steps_per_second": 1.127, + "step": 1540 + }, + { + "epoch": 1.688767123287671, + "grad_norm": 2.735790491104126, + "learning_rate": 2.4593826157595453e-06, + "loss": 1.6836, + "step": 1541 + }, + { + "epoch": 1.6898630136986301, + "grad_norm": 2.601675271987915, + "learning_rate": 2.4573517465475226e-06, + "loss": 1.5497, + "step": 1542 + }, + { + "epoch": 1.6909589041095892, + "grad_norm": 2.7680232524871826, + "learning_rate": 2.4553208773355e-06, + "loss": 1.5773, + "step": 1543 + }, + { + "epoch": 1.692054794520548, + "grad_norm": 2.7184901237487793, + "learning_rate": 2.453290008123477e-06, + "loss": 1.61, + "step": 1544 + }, + { + "epoch": 1.6931506849315068, + "grad_norm": 2.743377685546875, + "learning_rate": 2.4512591389114544e-06, + "loss": 1.6277, + "step": 1545 + }, + { + "epoch": 1.6942465753424658, + "grad_norm": 2.686190605163574, + "learning_rate": 2.4492282696994317e-06, + "loss": 1.6108, + "step": 1546 + }, + { + "epoch": 1.6953424657534246, + "grad_norm": 2.6933157444000244, + "learning_rate": 2.447197400487409e-06, + "loss": 1.5984, + "step": 1547 + }, + { + "epoch": 1.6964383561643834, + "grad_norm": 2.764714479446411, + "learning_rate": 2.4451665312753862e-06, + "loss": 1.6154, + "step": 1548 + }, + { + "epoch": 1.6975342465753425, + "grad_norm": 2.6571526527404785, + "learning_rate": 2.4431356620633635e-06, + "loss": 1.6502, + "step": 1549 + }, + { + "epoch": 1.6986301369863015, + "grad_norm": 2.739225149154663, + "learning_rate": 2.4411047928513408e-06, + "loss": 1.6142, + "step": 1550 + }, + { + "epoch": 1.6986301369863015, + "eval_loss": 1.8304396867752075, + "eval_runtime": 37.271, + "eval_samples_per_second": 26.831, + "eval_steps_per_second": 1.127, + "step": 1550 + }, + { + "epoch": 1.6997260273972603, + "grad_norm": 2.866999864578247, + "learning_rate": 2.439073923639318e-06, + "loss": 1.6155, + "step": 1551 + }, + { + "epoch": 1.7008219178082191, + "grad_norm": 2.670107364654541, + "learning_rate": 2.4370430544272953e-06, + "loss": 1.5526, + "step": 1552 + }, + { + "epoch": 1.7019178082191782, + "grad_norm": 2.5540761947631836, + "learning_rate": 2.435012185215272e-06, + "loss": 1.5501, + "step": 1553 + }, + { + "epoch": 1.703013698630137, + "grad_norm": 2.8047935962677, + "learning_rate": 2.4329813160032495e-06, + "loss": 1.5782, + "step": 1554 + }, + { + "epoch": 1.7041095890410958, + "grad_norm": 2.717963218688965, + "learning_rate": 2.4309504467912267e-06, + "loss": 1.6043, + "step": 1555 + }, + { + "epoch": 1.7052054794520548, + "grad_norm": 2.7823145389556885, + "learning_rate": 2.428919577579204e-06, + "loss": 1.596, + "step": 1556 + }, + { + "epoch": 1.7063013698630138, + "grad_norm": 2.8534443378448486, + "learning_rate": 2.4268887083671813e-06, + "loss": 1.6807, + "step": 1557 + }, + { + "epoch": 1.7073972602739727, + "grad_norm": 2.6131157875061035, + "learning_rate": 2.4248578391551586e-06, + "loss": 1.5833, + "step": 1558 + }, + { + "epoch": 1.7084931506849315, + "grad_norm": 2.761531352996826, + "learning_rate": 2.422826969943136e-06, + "loss": 1.5799, + "step": 1559 + }, + { + "epoch": 1.7095890410958905, + "grad_norm": 2.550750970840454, + "learning_rate": 2.420796100731113e-06, + "loss": 1.5504, + "step": 1560 + }, + { + "epoch": 1.7095890410958905, + "eval_loss": 1.8282228708267212, + "eval_runtime": 37.2167, + "eval_samples_per_second": 26.87, + "eval_steps_per_second": 1.129, + "step": 1560 + }, + { + "epoch": 1.7106849315068493, + "grad_norm": 2.7341554164886475, + "learning_rate": 2.4187652315190904e-06, + "loss": 1.6141, + "step": 1561 + }, + { + "epoch": 1.7117808219178081, + "grad_norm": 2.5879876613616943, + "learning_rate": 2.4167343623070677e-06, + "loss": 1.6121, + "step": 1562 + }, + { + "epoch": 1.7128767123287671, + "grad_norm": 2.626166820526123, + "learning_rate": 2.414703493095045e-06, + "loss": 1.5993, + "step": 1563 + }, + { + "epoch": 1.7139726027397262, + "grad_norm": 2.620342254638672, + "learning_rate": 2.4126726238830222e-06, + "loss": 1.5972, + "step": 1564 + }, + { + "epoch": 1.7150684931506848, + "grad_norm": 2.596414566040039, + "learning_rate": 2.410641754670999e-06, + "loss": 1.5827, + "step": 1565 + }, + { + "epoch": 1.7161643835616438, + "grad_norm": 2.6863791942596436, + "learning_rate": 2.4086108854589763e-06, + "loss": 1.6427, + "step": 1566 + }, + { + "epoch": 1.7172602739726028, + "grad_norm": 2.741095781326294, + "learning_rate": 2.4065800162469536e-06, + "loss": 1.5882, + "step": 1567 + }, + { + "epoch": 1.7183561643835616, + "grad_norm": 2.890286922454834, + "learning_rate": 2.404549147034931e-06, + "loss": 1.6046, + "step": 1568 + }, + { + "epoch": 1.7194520547945205, + "grad_norm": 2.699467658996582, + "learning_rate": 2.402518277822908e-06, + "loss": 1.6395, + "step": 1569 + }, + { + "epoch": 1.7205479452054795, + "grad_norm": 2.6112730503082275, + "learning_rate": 2.4004874086108854e-06, + "loss": 1.5772, + "step": 1570 + }, + { + "epoch": 1.7205479452054795, + "eval_loss": 1.824203372001648, + "eval_runtime": 37.271, + "eval_samples_per_second": 26.831, + "eval_steps_per_second": 1.127, + "step": 1570 + }, + { + "epoch": 1.7216438356164385, + "grad_norm": 2.7761809825897217, + "learning_rate": 2.3984565393988627e-06, + "loss": 1.6177, + "step": 1571 + }, + { + "epoch": 1.7227397260273971, + "grad_norm": 2.691741704940796, + "learning_rate": 2.39642567018684e-06, + "loss": 1.5837, + "step": 1572 + }, + { + "epoch": 1.7238356164383561, + "grad_norm": 2.655440330505371, + "learning_rate": 2.3943948009748173e-06, + "loss": 1.5896, + "step": 1573 + }, + { + "epoch": 1.7249315068493152, + "grad_norm": 2.8032209873199463, + "learning_rate": 2.3923639317627945e-06, + "loss": 1.6267, + "step": 1574 + }, + { + "epoch": 1.726027397260274, + "grad_norm": 2.7980844974517822, + "learning_rate": 2.390333062550772e-06, + "loss": 1.5695, + "step": 1575 + }, + { + "epoch": 1.7271232876712328, + "grad_norm": 2.6170504093170166, + "learning_rate": 2.388302193338749e-06, + "loss": 1.4959, + "step": 1576 + }, + { + "epoch": 1.7282191780821918, + "grad_norm": 2.836310386657715, + "learning_rate": 2.3862713241267264e-06, + "loss": 1.6757, + "step": 1577 + }, + { + "epoch": 1.7293150684931506, + "grad_norm": 2.868952751159668, + "learning_rate": 2.3842404549147037e-06, + "loss": 1.6065, + "step": 1578 + }, + { + "epoch": 1.7304109589041095, + "grad_norm": 2.8427889347076416, + "learning_rate": 2.382209585702681e-06, + "loss": 1.5705, + "step": 1579 + }, + { + "epoch": 1.7315068493150685, + "grad_norm": 2.6732518672943115, + "learning_rate": 2.380178716490658e-06, + "loss": 1.6437, + "step": 1580 + }, + { + "epoch": 1.7315068493150685, + "eval_loss": 1.8212717771530151, + "eval_runtime": 37.3837, + "eval_samples_per_second": 26.75, + "eval_steps_per_second": 1.123, + "step": 1580 + }, + { + "epoch": 1.7326027397260275, + "grad_norm": 2.7127609252929688, + "learning_rate": 2.3781478472786355e-06, + "loss": 1.5995, + "step": 1581 + }, + { + "epoch": 1.7336986301369863, + "grad_norm": 2.822373390197754, + "learning_rate": 2.3761169780666128e-06, + "loss": 1.6416, + "step": 1582 + }, + { + "epoch": 1.7347945205479451, + "grad_norm": 2.7300949096679688, + "learning_rate": 2.37408610885459e-06, + "loss": 1.6504, + "step": 1583 + }, + { + "epoch": 1.7358904109589042, + "grad_norm": 2.6506187915802, + "learning_rate": 2.3720552396425673e-06, + "loss": 1.5873, + "step": 1584 + }, + { + "epoch": 1.736986301369863, + "grad_norm": 2.6960768699645996, + "learning_rate": 2.3700243704305446e-06, + "loss": 1.6743, + "step": 1585 + }, + { + "epoch": 1.7380821917808218, + "grad_norm": 2.640947103500366, + "learning_rate": 2.367993501218522e-06, + "loss": 1.6105, + "step": 1586 + }, + { + "epoch": 1.7391780821917808, + "grad_norm": 2.8700079917907715, + "learning_rate": 2.365962632006499e-06, + "loss": 1.5965, + "step": 1587 + }, + { + "epoch": 1.7402739726027399, + "grad_norm": 2.748927116394043, + "learning_rate": 2.3639317627944764e-06, + "loss": 1.6386, + "step": 1588 + }, + { + "epoch": 1.7413698630136987, + "grad_norm": 2.7105116844177246, + "learning_rate": 2.3619008935824537e-06, + "loss": 1.6392, + "step": 1589 + }, + { + "epoch": 1.7424657534246575, + "grad_norm": 2.6624841690063477, + "learning_rate": 2.3598700243704305e-06, + "loss": 1.6312, + "step": 1590 + }, + { + "epoch": 1.7424657534246575, + "eval_loss": 1.8209728002548218, + "eval_runtime": 37.3533, + "eval_samples_per_second": 26.771, + "eval_steps_per_second": 1.124, + "step": 1590 + }, + { + "epoch": 1.7435616438356165, + "grad_norm": 2.903249740600586, + "learning_rate": 2.357839155158408e-06, + "loss": 1.6263, + "step": 1591 + }, + { + "epoch": 1.7446575342465753, + "grad_norm": 2.792497158050537, + "learning_rate": 2.355808285946385e-06, + "loss": 1.6206, + "step": 1592 + }, + { + "epoch": 1.7457534246575341, + "grad_norm": 2.737065315246582, + "learning_rate": 2.3537774167343624e-06, + "loss": 1.6394, + "step": 1593 + }, + { + "epoch": 1.7468493150684932, + "grad_norm": 2.651978015899658, + "learning_rate": 2.3517465475223396e-06, + "loss": 1.6307, + "step": 1594 + }, + { + "epoch": 1.7479452054794522, + "grad_norm": 2.7566919326782227, + "learning_rate": 2.349715678310317e-06, + "loss": 1.6203, + "step": 1595 + }, + { + "epoch": 1.749041095890411, + "grad_norm": 2.7853057384490967, + "learning_rate": 2.347684809098294e-06, + "loss": 1.594, + "step": 1596 + }, + { + "epoch": 1.7501369863013698, + "grad_norm": 2.770509719848633, + "learning_rate": 2.3456539398862715e-06, + "loss": 1.6491, + "step": 1597 + }, + { + "epoch": 1.7512328767123289, + "grad_norm": 2.718459367752075, + "learning_rate": 2.3436230706742487e-06, + "loss": 1.5671, + "step": 1598 + }, + { + "epoch": 1.7523287671232877, + "grad_norm": 2.724194288253784, + "learning_rate": 2.341592201462226e-06, + "loss": 1.5771, + "step": 1599 + }, + { + "epoch": 1.7534246575342465, + "grad_norm": 2.691155195236206, + "learning_rate": 2.3395613322502033e-06, + "loss": 1.5889, + "step": 1600 + }, + { + "epoch": 1.7534246575342465, + "eval_loss": 1.8217312097549438, + "eval_runtime": 37.3364, + "eval_samples_per_second": 26.784, + "eval_steps_per_second": 1.125, + "step": 1600 + }, + { + "epoch": 1.7545205479452055, + "grad_norm": 2.7511825561523438, + "learning_rate": 2.3375304630381806e-06, + "loss": 1.562, + "step": 1601 + }, + { + "epoch": 1.7556164383561645, + "grad_norm": 2.5794379711151123, + "learning_rate": 2.335499593826158e-06, + "loss": 1.5685, + "step": 1602 + }, + { + "epoch": 1.7567123287671231, + "grad_norm": 2.7914280891418457, + "learning_rate": 2.333468724614135e-06, + "loss": 1.6267, + "step": 1603 + }, + { + "epoch": 1.7578082191780822, + "grad_norm": 2.678723096847534, + "learning_rate": 2.3314378554021124e-06, + "loss": 1.6281, + "step": 1604 + }, + { + "epoch": 1.7589041095890412, + "grad_norm": 2.757333993911743, + "learning_rate": 2.3294069861900897e-06, + "loss": 1.5693, + "step": 1605 + }, + { + "epoch": 1.76, + "grad_norm": 2.6844334602355957, + "learning_rate": 2.327376116978067e-06, + "loss": 1.5699, + "step": 1606 + }, + { + "epoch": 1.7610958904109588, + "grad_norm": 2.7547338008880615, + "learning_rate": 2.3253452477660442e-06, + "loss": 1.6294, + "step": 1607 + }, + { + "epoch": 1.7621917808219179, + "grad_norm": 2.864664077758789, + "learning_rate": 2.3233143785540215e-06, + "loss": 1.5747, + "step": 1608 + }, + { + "epoch": 1.7632876712328767, + "grad_norm": 2.9029629230499268, + "learning_rate": 2.3212835093419988e-06, + "loss": 1.6182, + "step": 1609 + }, + { + "epoch": 1.7643835616438355, + "grad_norm": 2.707786798477173, + "learning_rate": 2.319252640129976e-06, + "loss": 1.5702, + "step": 1610 + }, + { + "epoch": 1.7643835616438355, + "eval_loss": 1.8216935396194458, + "eval_runtime": 36.8753, + "eval_samples_per_second": 27.118, + "eval_steps_per_second": 1.139, + "step": 1610 + }, + { + "epoch": 1.7654794520547945, + "grad_norm": 2.964879035949707, + "learning_rate": 2.3172217709179533e-06, + "loss": 1.6833, + "step": 1611 + }, + { + "epoch": 1.7665753424657535, + "grad_norm": 2.728271245956421, + "learning_rate": 2.3151909017059306e-06, + "loss": 1.636, + "step": 1612 + }, + { + "epoch": 1.7676712328767124, + "grad_norm": 2.801959753036499, + "learning_rate": 2.313160032493908e-06, + "loss": 1.5545, + "step": 1613 + }, + { + "epoch": 1.7687671232876712, + "grad_norm": 2.74782395362854, + "learning_rate": 2.311129163281885e-06, + "loss": 1.6312, + "step": 1614 + }, + { + "epoch": 1.7698630136986302, + "grad_norm": 2.512051582336426, + "learning_rate": 2.309098294069862e-06, + "loss": 1.582, + "step": 1615 + }, + { + "epoch": 1.770958904109589, + "grad_norm": 2.798088788986206, + "learning_rate": 2.3070674248578393e-06, + "loss": 1.562, + "step": 1616 + }, + { + "epoch": 1.7720547945205478, + "grad_norm": 2.795612096786499, + "learning_rate": 2.3050365556458166e-06, + "loss": 1.6132, + "step": 1617 + }, + { + "epoch": 1.7731506849315068, + "grad_norm": 2.7699942588806152, + "learning_rate": 2.303005686433794e-06, + "loss": 1.6311, + "step": 1618 + }, + { + "epoch": 1.7742465753424659, + "grad_norm": 2.867352247238159, + "learning_rate": 2.300974817221771e-06, + "loss": 1.6028, + "step": 1619 + }, + { + "epoch": 1.7753424657534247, + "grad_norm": 2.5470569133758545, + "learning_rate": 2.2989439480097484e-06, + "loss": 1.562, + "step": 1620 + }, + { + "epoch": 1.7753424657534247, + "eval_loss": 1.8190903663635254, + "eval_runtime": 37.161, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 1620 + }, + { + "epoch": 1.7764383561643835, + "grad_norm": 2.6718320846557617, + "learning_rate": 2.2969130787977257e-06, + "loss": 1.6118, + "step": 1621 + }, + { + "epoch": 1.7775342465753425, + "grad_norm": 2.5534214973449707, + "learning_rate": 2.294882209585703e-06, + "loss": 1.6395, + "step": 1622 + }, + { + "epoch": 1.7786301369863013, + "grad_norm": 2.625375747680664, + "learning_rate": 2.2928513403736802e-06, + "loss": 1.5818, + "step": 1623 + }, + { + "epoch": 1.7797260273972602, + "grad_norm": 2.663601875305176, + "learning_rate": 2.2908204711616575e-06, + "loss": 1.6015, + "step": 1624 + }, + { + "epoch": 1.7808219178082192, + "grad_norm": 2.6689491271972656, + "learning_rate": 2.2887896019496348e-06, + "loss": 1.6432, + "step": 1625 + }, + { + "epoch": 1.7819178082191782, + "grad_norm": 2.503133535385132, + "learning_rate": 2.286758732737612e-06, + "loss": 1.5805, + "step": 1626 + }, + { + "epoch": 1.783013698630137, + "grad_norm": 2.806959867477417, + "learning_rate": 2.284727863525589e-06, + "loss": 1.5918, + "step": 1627 + }, + { + "epoch": 1.7841095890410958, + "grad_norm": 2.6896889209747314, + "learning_rate": 2.282696994313566e-06, + "loss": 1.6209, + "step": 1628 + }, + { + "epoch": 1.7852054794520549, + "grad_norm": 2.686661720275879, + "learning_rate": 2.2806661251015434e-06, + "loss": 1.6051, + "step": 1629 + }, + { + "epoch": 1.7863013698630137, + "grad_norm": 2.845679759979248, + "learning_rate": 2.2786352558895207e-06, + "loss": 1.5839, + "step": 1630 + }, + { + "epoch": 1.7863013698630137, + "eval_loss": 1.8239432573318481, + "eval_runtime": 37.2256, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 1630 + }, + { + "epoch": 1.7873972602739725, + "grad_norm": 2.74546480178833, + "learning_rate": 2.276604386677498e-06, + "loss": 1.5532, + "step": 1631 + }, + { + "epoch": 1.7884931506849315, + "grad_norm": 2.857858180999756, + "learning_rate": 2.2745735174654753e-06, + "loss": 1.5829, + "step": 1632 + }, + { + "epoch": 1.7895890410958906, + "grad_norm": 2.8485302925109863, + "learning_rate": 2.2725426482534526e-06, + "loss": 1.6618, + "step": 1633 + }, + { + "epoch": 1.7906849315068492, + "grad_norm": 2.7431278228759766, + "learning_rate": 2.27051177904143e-06, + "loss": 1.6025, + "step": 1634 + }, + { + "epoch": 1.7917808219178082, + "grad_norm": 2.7853291034698486, + "learning_rate": 2.268480909829407e-06, + "loss": 1.5888, + "step": 1635 + }, + { + "epoch": 1.7928767123287672, + "grad_norm": 2.7201032638549805, + "learning_rate": 2.2664500406173844e-06, + "loss": 1.6081, + "step": 1636 + }, + { + "epoch": 1.793972602739726, + "grad_norm": 2.720609426498413, + "learning_rate": 2.2644191714053617e-06, + "loss": 1.6341, + "step": 1637 + }, + { + "epoch": 1.7950684931506848, + "grad_norm": 2.9016213417053223, + "learning_rate": 2.262388302193339e-06, + "loss": 1.5967, + "step": 1638 + }, + { + "epoch": 1.7961643835616439, + "grad_norm": 2.7303457260131836, + "learning_rate": 2.260357432981316e-06, + "loss": 1.5799, + "step": 1639 + }, + { + "epoch": 1.7972602739726027, + "grad_norm": 2.7622275352478027, + "learning_rate": 2.2583265637692935e-06, + "loss": 1.6015, + "step": 1640 + }, + { + "epoch": 1.7972602739726027, + "eval_loss": 1.822721242904663, + "eval_runtime": 37.294, + "eval_samples_per_second": 26.814, + "eval_steps_per_second": 1.126, + "step": 1640 + }, + { + "epoch": 1.7983561643835615, + "grad_norm": 2.7074835300445557, + "learning_rate": 2.2562956945572708e-06, + "loss": 1.5998, + "step": 1641 + }, + { + "epoch": 1.7994520547945205, + "grad_norm": 2.8944125175476074, + "learning_rate": 2.254264825345248e-06, + "loss": 1.5977, + "step": 1642 + }, + { + "epoch": 1.8005479452054796, + "grad_norm": 2.8836803436279297, + "learning_rate": 2.2522339561332253e-06, + "loss": 1.6044, + "step": 1643 + }, + { + "epoch": 1.8016438356164384, + "grad_norm": 2.7658469676971436, + "learning_rate": 2.2502030869212026e-06, + "loss": 1.6087, + "step": 1644 + }, + { + "epoch": 1.8027397260273972, + "grad_norm": 2.8019907474517822, + "learning_rate": 2.24817221770918e-06, + "loss": 1.6437, + "step": 1645 + }, + { + "epoch": 1.8038356164383562, + "grad_norm": 2.6863455772399902, + "learning_rate": 2.246141348497157e-06, + "loss": 1.6225, + "step": 1646 + }, + { + "epoch": 1.804931506849315, + "grad_norm": 2.5988214015960693, + "learning_rate": 2.2441104792851344e-06, + "loss": 1.5706, + "step": 1647 + }, + { + "epoch": 1.8060273972602738, + "grad_norm": 2.6351659297943115, + "learning_rate": 2.2420796100731117e-06, + "loss": 1.5973, + "step": 1648 + }, + { + "epoch": 1.8071232876712329, + "grad_norm": 2.851712226867676, + "learning_rate": 2.240048740861089e-06, + "loss": 1.6184, + "step": 1649 + }, + { + "epoch": 1.808219178082192, + "grad_norm": 2.7069613933563232, + "learning_rate": 2.2380178716490662e-06, + "loss": 1.5047, + "step": 1650 + }, + { + "epoch": 1.808219178082192, + "eval_loss": 1.81782066822052, + "eval_runtime": 37.3161, + "eval_samples_per_second": 26.798, + "eval_steps_per_second": 1.126, + "step": 1650 + }, + { + "epoch": 1.8093150684931507, + "grad_norm": 2.880129337310791, + "learning_rate": 2.2359870024370435e-06, + "loss": 1.6278, + "step": 1651 + }, + { + "epoch": 1.8104109589041095, + "grad_norm": 2.7861595153808594, + "learning_rate": 2.2339561332250204e-06, + "loss": 1.6064, + "step": 1652 + }, + { + "epoch": 1.8115068493150686, + "grad_norm": 2.8799221515655518, + "learning_rate": 2.2319252640129976e-06, + "loss": 1.6034, + "step": 1653 + }, + { + "epoch": 1.8126027397260274, + "grad_norm": 2.7535576820373535, + "learning_rate": 2.229894394800975e-06, + "loss": 1.5431, + "step": 1654 + }, + { + "epoch": 1.8136986301369862, + "grad_norm": 2.7955143451690674, + "learning_rate": 2.227863525588952e-06, + "loss": 1.531, + "step": 1655 + }, + { + "epoch": 1.8147945205479452, + "grad_norm": 2.5939323902130127, + "learning_rate": 2.2258326563769295e-06, + "loss": 1.5566, + "step": 1656 + }, + { + "epoch": 1.8158904109589042, + "grad_norm": 2.676406145095825, + "learning_rate": 2.2238017871649067e-06, + "loss": 1.56, + "step": 1657 + }, + { + "epoch": 1.816986301369863, + "grad_norm": 2.7449264526367188, + "learning_rate": 2.221770917952884e-06, + "loss": 1.599, + "step": 1658 + }, + { + "epoch": 1.8180821917808219, + "grad_norm": 2.7454347610473633, + "learning_rate": 2.2197400487408613e-06, + "loss": 1.6435, + "step": 1659 + }, + { + "epoch": 1.819178082191781, + "grad_norm": 2.5657317638397217, + "learning_rate": 2.2177091795288386e-06, + "loss": 1.5638, + "step": 1660 + }, + { + "epoch": 1.819178082191781, + "eval_loss": 1.818527102470398, + "eval_runtime": 37.2477, + "eval_samples_per_second": 26.847, + "eval_steps_per_second": 1.128, + "step": 1660 + }, + { + "epoch": 1.8202739726027397, + "grad_norm": 2.6036269664764404, + "learning_rate": 2.215678310316816e-06, + "loss": 1.5713, + "step": 1661 + }, + { + "epoch": 1.8213698630136985, + "grad_norm": 2.8862414360046387, + "learning_rate": 2.213647441104793e-06, + "loss": 1.609, + "step": 1662 + }, + { + "epoch": 1.8224657534246576, + "grad_norm": 2.6588282585144043, + "learning_rate": 2.2116165718927704e-06, + "loss": 1.5795, + "step": 1663 + }, + { + "epoch": 1.8235616438356166, + "grad_norm": 2.8073630332946777, + "learning_rate": 2.2095857026807473e-06, + "loss": 1.6011, + "step": 1664 + }, + { + "epoch": 1.8246575342465754, + "grad_norm": 2.6723380088806152, + "learning_rate": 2.2075548334687245e-06, + "loss": 1.5613, + "step": 1665 + }, + { + "epoch": 1.8257534246575342, + "grad_norm": 2.8164796829223633, + "learning_rate": 2.205523964256702e-06, + "loss": 1.579, + "step": 1666 + }, + { + "epoch": 1.8268493150684932, + "grad_norm": 2.777341842651367, + "learning_rate": 2.203493095044679e-06, + "loss": 1.6111, + "step": 1667 + }, + { + "epoch": 1.827945205479452, + "grad_norm": 2.6877028942108154, + "learning_rate": 2.2014622258326564e-06, + "loss": 1.6171, + "step": 1668 + }, + { + "epoch": 1.8290410958904109, + "grad_norm": 2.638216972351074, + "learning_rate": 2.1994313566206336e-06, + "loss": 1.6363, + "step": 1669 + }, + { + "epoch": 1.83013698630137, + "grad_norm": 2.5175044536590576, + "learning_rate": 2.197400487408611e-06, + "loss": 1.5849, + "step": 1670 + }, + { + "epoch": 1.83013698630137, + "eval_loss": 1.8164860010147095, + "eval_runtime": 37.2838, + "eval_samples_per_second": 26.821, + "eval_steps_per_second": 1.126, + "step": 1670 + }, + { + "epoch": 1.831232876712329, + "grad_norm": 2.6411025524139404, + "learning_rate": 2.195369618196588e-06, + "loss": 1.5592, + "step": 1671 + }, + { + "epoch": 1.8323287671232875, + "grad_norm": 2.6564719676971436, + "learning_rate": 2.1933387489845655e-06, + "loss": 1.591, + "step": 1672 + }, + { + "epoch": 1.8334246575342465, + "grad_norm": 2.6379082202911377, + "learning_rate": 2.1913078797725427e-06, + "loss": 1.6116, + "step": 1673 + }, + { + "epoch": 1.8345205479452056, + "grad_norm": 2.6217899322509766, + "learning_rate": 2.18927701056052e-06, + "loss": 1.6729, + "step": 1674 + }, + { + "epoch": 1.8356164383561644, + "grad_norm": 2.5114355087280273, + "learning_rate": 2.1872461413484973e-06, + "loss": 1.6089, + "step": 1675 + }, + { + "epoch": 1.8367123287671232, + "grad_norm": 2.6792423725128174, + "learning_rate": 2.1852152721364746e-06, + "loss": 1.591, + "step": 1676 + }, + { + "epoch": 1.8378082191780822, + "grad_norm": 2.8523600101470947, + "learning_rate": 2.183184402924452e-06, + "loss": 1.5724, + "step": 1677 + }, + { + "epoch": 1.838904109589041, + "grad_norm": 2.732600212097168, + "learning_rate": 2.181153533712429e-06, + "loss": 1.623, + "step": 1678 + }, + { + "epoch": 1.8399999999999999, + "grad_norm": 2.71589732170105, + "learning_rate": 2.1791226645004064e-06, + "loss": 1.5931, + "step": 1679 + }, + { + "epoch": 1.841095890410959, + "grad_norm": 2.701359748840332, + "learning_rate": 2.1770917952883837e-06, + "loss": 1.5837, + "step": 1680 + }, + { + "epoch": 1.841095890410959, + "eval_loss": 1.8175450563430786, + "eval_runtime": 37.2426, + "eval_samples_per_second": 26.851, + "eval_steps_per_second": 1.128, + "step": 1680 + }, + { + "epoch": 1.842191780821918, + "grad_norm": 2.9347689151763916, + "learning_rate": 2.175060926076361e-06, + "loss": 1.6542, + "step": 1681 + }, + { + "epoch": 1.8432876712328767, + "grad_norm": 2.825683832168579, + "learning_rate": 2.1730300568643382e-06, + "loss": 1.6752, + "step": 1682 + }, + { + "epoch": 1.8443835616438355, + "grad_norm": 2.691688299179077, + "learning_rate": 2.1709991876523155e-06, + "loss": 1.5869, + "step": 1683 + }, + { + "epoch": 1.8454794520547946, + "grad_norm": 2.635225296020508, + "learning_rate": 2.1689683184402928e-06, + "loss": 1.5377, + "step": 1684 + }, + { + "epoch": 1.8465753424657534, + "grad_norm": 2.7036449909210205, + "learning_rate": 2.16693744922827e-06, + "loss": 1.6239, + "step": 1685 + }, + { + "epoch": 1.8476712328767122, + "grad_norm": 2.7112629413604736, + "learning_rate": 2.1649065800162473e-06, + "loss": 1.5818, + "step": 1686 + }, + { + "epoch": 1.8487671232876712, + "grad_norm": 2.7251856327056885, + "learning_rate": 2.1628757108042246e-06, + "loss": 1.5807, + "step": 1687 + }, + { + "epoch": 1.8498630136986303, + "grad_norm": 2.8971238136291504, + "learning_rate": 2.160844841592202e-06, + "loss": 1.5787, + "step": 1688 + }, + { + "epoch": 1.850958904109589, + "grad_norm": 2.6737334728240967, + "learning_rate": 2.1588139723801787e-06, + "loss": 1.5635, + "step": 1689 + }, + { + "epoch": 1.8520547945205479, + "grad_norm": 2.757006883621216, + "learning_rate": 2.156783103168156e-06, + "loss": 1.5779, + "step": 1690 + }, + { + "epoch": 1.8520547945205479, + "eval_loss": 1.8178988695144653, + "eval_runtime": 37.2265, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 1690 + }, + { + "epoch": 1.853150684931507, + "grad_norm": 2.6924121379852295, + "learning_rate": 2.1547522339561333e-06, + "loss": 1.5519, + "step": 1691 + }, + { + "epoch": 1.8542465753424657, + "grad_norm": 2.742126703262329, + "learning_rate": 2.1527213647441106e-06, + "loss": 1.6144, + "step": 1692 + }, + { + "epoch": 1.8553424657534245, + "grad_norm": 2.8751533031463623, + "learning_rate": 2.150690495532088e-06, + "loss": 1.6462, + "step": 1693 + }, + { + "epoch": 1.8564383561643836, + "grad_norm": 2.804011344909668, + "learning_rate": 2.148659626320065e-06, + "loss": 1.6214, + "step": 1694 + }, + { + "epoch": 1.8575342465753426, + "grad_norm": 2.7928590774536133, + "learning_rate": 2.1466287571080424e-06, + "loss": 1.6244, + "step": 1695 + }, + { + "epoch": 1.8586301369863014, + "grad_norm": 2.749242067337036, + "learning_rate": 2.1445978878960197e-06, + "loss": 1.5749, + "step": 1696 + }, + { + "epoch": 1.8597260273972602, + "grad_norm": 2.588762044906616, + "learning_rate": 2.142567018683997e-06, + "loss": 1.5898, + "step": 1697 + }, + { + "epoch": 1.8608219178082193, + "grad_norm": 2.82580828666687, + "learning_rate": 2.140536149471974e-06, + "loss": 1.6512, + "step": 1698 + }, + { + "epoch": 1.861917808219178, + "grad_norm": 2.6267337799072266, + "learning_rate": 2.1385052802599515e-06, + "loss": 1.5664, + "step": 1699 + }, + { + "epoch": 1.8630136986301369, + "grad_norm": 2.516296863555908, + "learning_rate": 2.1364744110479288e-06, + "loss": 1.5716, + "step": 1700 + }, + { + "epoch": 1.8630136986301369, + "eval_loss": 1.8167686462402344, + "eval_runtime": 37.1827, + "eval_samples_per_second": 26.894, + "eval_steps_per_second": 1.13, + "step": 1700 + }, + { + "epoch": 1.864109589041096, + "grad_norm": 2.8827500343322754, + "learning_rate": 2.134443541835906e-06, + "loss": 1.6379, + "step": 1701 + }, + { + "epoch": 1.865205479452055, + "grad_norm": 2.6723849773406982, + "learning_rate": 2.132412672623883e-06, + "loss": 1.5855, + "step": 1702 + }, + { + "epoch": 1.8663013698630135, + "grad_norm": 2.75603985786438, + "learning_rate": 2.13038180341186e-06, + "loss": 1.5692, + "step": 1703 + }, + { + "epoch": 1.8673972602739726, + "grad_norm": 2.8822832107543945, + "learning_rate": 2.1283509341998374e-06, + "loss": 1.5985, + "step": 1704 + }, + { + "epoch": 1.8684931506849316, + "grad_norm": 3.0294716358184814, + "learning_rate": 2.1263200649878147e-06, + "loss": 1.6376, + "step": 1705 + }, + { + "epoch": 1.8695890410958904, + "grad_norm": 2.8347859382629395, + "learning_rate": 2.124289195775792e-06, + "loss": 1.5807, + "step": 1706 + }, + { + "epoch": 1.8706849315068492, + "grad_norm": 2.871610641479492, + "learning_rate": 2.1222583265637693e-06, + "loss": 1.602, + "step": 1707 + }, + { + "epoch": 1.8717808219178083, + "grad_norm": 2.6605138778686523, + "learning_rate": 2.1202274573517465e-06, + "loss": 1.5133, + "step": 1708 + }, + { + "epoch": 1.872876712328767, + "grad_norm": 2.6611695289611816, + "learning_rate": 2.118196588139724e-06, + "loss": 1.5608, + "step": 1709 + }, + { + "epoch": 1.8739726027397259, + "grad_norm": 2.594904899597168, + "learning_rate": 2.116165718927701e-06, + "loss": 1.5776, + "step": 1710 + }, + { + "epoch": 1.8739726027397259, + "eval_loss": 1.8165113925933838, + "eval_runtime": 36.8295, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 1710 + }, + { + "epoch": 1.875068493150685, + "grad_norm": 2.596139669418335, + "learning_rate": 2.1141348497156784e-06, + "loss": 1.5904, + "step": 1711 + }, + { + "epoch": 1.876164383561644, + "grad_norm": 2.75966477394104, + "learning_rate": 2.1121039805036556e-06, + "loss": 1.54, + "step": 1712 + }, + { + "epoch": 1.8772602739726028, + "grad_norm": 2.653257131576538, + "learning_rate": 2.110073111291633e-06, + "loss": 1.5693, + "step": 1713 + }, + { + "epoch": 1.8783561643835616, + "grad_norm": 2.5443379878997803, + "learning_rate": 2.10804224207961e-06, + "loss": 1.588, + "step": 1714 + }, + { + "epoch": 1.8794520547945206, + "grad_norm": 2.6353726387023926, + "learning_rate": 2.1060113728675875e-06, + "loss": 1.6178, + "step": 1715 + }, + { + "epoch": 1.8805479452054794, + "grad_norm": 2.7751545906066895, + "learning_rate": 2.1039805036555647e-06, + "loss": 1.6271, + "step": 1716 + }, + { + "epoch": 1.8816438356164382, + "grad_norm": 2.7007811069488525, + "learning_rate": 2.101949634443542e-06, + "loss": 1.5014, + "step": 1717 + }, + { + "epoch": 1.8827397260273973, + "grad_norm": 2.968076229095459, + "learning_rate": 2.0999187652315193e-06, + "loss": 1.596, + "step": 1718 + }, + { + "epoch": 1.8838356164383563, + "grad_norm": 2.6997463703155518, + "learning_rate": 2.0978878960194966e-06, + "loss": 1.562, + "step": 1719 + }, + { + "epoch": 1.884931506849315, + "grad_norm": 2.7616124153137207, + "learning_rate": 2.095857026807474e-06, + "loss": 1.5971, + "step": 1720 + }, + { + "epoch": 1.884931506849315, + "eval_loss": 1.8137410879135132, + "eval_runtime": 37.0154, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 1720 + }, + { + "epoch": 1.886027397260274, + "grad_norm": 2.6898655891418457, + "learning_rate": 2.093826157595451e-06, + "loss": 1.5932, + "step": 1721 + }, + { + "epoch": 1.887123287671233, + "grad_norm": 2.7165539264678955, + "learning_rate": 2.0917952883834284e-06, + "loss": 1.5765, + "step": 1722 + }, + { + "epoch": 1.8882191780821918, + "grad_norm": 2.7609431743621826, + "learning_rate": 2.0897644191714057e-06, + "loss": 1.6038, + "step": 1723 + }, + { + "epoch": 1.8893150684931506, + "grad_norm": 2.7460267543792725, + "learning_rate": 2.087733549959383e-06, + "loss": 1.6328, + "step": 1724 + }, + { + "epoch": 1.8904109589041096, + "grad_norm": 2.649451732635498, + "learning_rate": 2.0857026807473602e-06, + "loss": 1.5544, + "step": 1725 + }, + { + "epoch": 1.8915068493150686, + "grad_norm": 2.6734838485717773, + "learning_rate": 2.0836718115353375e-06, + "loss": 1.5431, + "step": 1726 + }, + { + "epoch": 1.8926027397260274, + "grad_norm": 2.8084394931793213, + "learning_rate": 2.0816409423233144e-06, + "loss": 1.6284, + "step": 1727 + }, + { + "epoch": 1.8936986301369862, + "grad_norm": 2.5050930976867676, + "learning_rate": 2.0796100731112916e-06, + "loss": 1.5728, + "step": 1728 + }, + { + "epoch": 1.8947945205479453, + "grad_norm": 2.7016143798828125, + "learning_rate": 2.077579203899269e-06, + "loss": 1.5816, + "step": 1729 + }, + { + "epoch": 1.895890410958904, + "grad_norm": 2.58982253074646, + "learning_rate": 2.075548334687246e-06, + "loss": 1.5347, + "step": 1730 + }, + { + "epoch": 1.895890410958904, + "eval_loss": 1.8112925291061401, + "eval_runtime": 37.0911, + "eval_samples_per_second": 26.961, + "eval_steps_per_second": 1.132, + "step": 1730 + }, + { + "epoch": 1.896986301369863, + "grad_norm": 2.776846408843994, + "learning_rate": 2.0735174654752235e-06, + "loss": 1.598, + "step": 1731 + }, + { + "epoch": 1.898082191780822, + "grad_norm": 2.5896499156951904, + "learning_rate": 2.0714865962632007e-06, + "loss": 1.5263, + "step": 1732 + }, + { + "epoch": 1.899178082191781, + "grad_norm": 2.7289252281188965, + "learning_rate": 2.069455727051178e-06, + "loss": 1.5866, + "step": 1733 + }, + { + "epoch": 1.9002739726027398, + "grad_norm": 2.4769744873046875, + "learning_rate": 2.0674248578391553e-06, + "loss": 1.5463, + "step": 1734 + }, + { + "epoch": 1.9013698630136986, + "grad_norm": 2.836574077606201, + "learning_rate": 2.0653939886271326e-06, + "loss": 1.5549, + "step": 1735 + }, + { + "epoch": 1.9024657534246576, + "grad_norm": 2.8002960681915283, + "learning_rate": 2.06336311941511e-06, + "loss": 1.6378, + "step": 1736 + }, + { + "epoch": 1.9035616438356164, + "grad_norm": 2.6109306812286377, + "learning_rate": 2.061332250203087e-06, + "loss": 1.5687, + "step": 1737 + }, + { + "epoch": 1.9046575342465752, + "grad_norm": 2.7544755935668945, + "learning_rate": 2.0593013809910644e-06, + "loss": 1.5376, + "step": 1738 + }, + { + "epoch": 1.9057534246575343, + "grad_norm": 2.7387688159942627, + "learning_rate": 2.0572705117790417e-06, + "loss": 1.5937, + "step": 1739 + }, + { + "epoch": 1.9068493150684933, + "grad_norm": 2.7123818397521973, + "learning_rate": 2.055239642567019e-06, + "loss": 1.6111, + "step": 1740 + }, + { + "epoch": 1.9068493150684933, + "eval_loss": 1.814138412475586, + "eval_runtime": 37.1458, + "eval_samples_per_second": 26.921, + "eval_steps_per_second": 1.131, + "step": 1740 + }, + { + "epoch": 1.907945205479452, + "grad_norm": 2.5983684062957764, + "learning_rate": 2.0532087733549962e-06, + "loss": 1.5888, + "step": 1741 + }, + { + "epoch": 1.909041095890411, + "grad_norm": 2.668445348739624, + "learning_rate": 2.0511779041429735e-06, + "loss": 1.6256, + "step": 1742 + }, + { + "epoch": 1.91013698630137, + "grad_norm": 2.5294125080108643, + "learning_rate": 2.0491470349309508e-06, + "loss": 1.5489, + "step": 1743 + }, + { + "epoch": 1.9112328767123288, + "grad_norm": 2.815340757369995, + "learning_rate": 2.047116165718928e-06, + "loss": 1.6041, + "step": 1744 + }, + { + "epoch": 1.9123287671232876, + "grad_norm": 2.791731595993042, + "learning_rate": 2.0450852965069053e-06, + "loss": 1.5693, + "step": 1745 + }, + { + "epoch": 1.9134246575342466, + "grad_norm": 2.6540396213531494, + "learning_rate": 2.0430544272948826e-06, + "loss": 1.6218, + "step": 1746 + }, + { + "epoch": 1.9145205479452054, + "grad_norm": 2.8304696083068848, + "learning_rate": 2.04102355808286e-06, + "loss": 1.5757, + "step": 1747 + }, + { + "epoch": 1.9156164383561642, + "grad_norm": 2.590974807739258, + "learning_rate": 2.038992688870837e-06, + "loss": 1.5322, + "step": 1748 + }, + { + "epoch": 1.9167123287671233, + "grad_norm": 2.647883176803589, + "learning_rate": 2.0369618196588144e-06, + "loss": 1.566, + "step": 1749 + }, + { + "epoch": 1.9178082191780823, + "grad_norm": 2.880141019821167, + "learning_rate": 2.0349309504467917e-06, + "loss": 1.5781, + "step": 1750 + }, + { + "epoch": 1.9178082191780823, + "eval_loss": 1.8149348497390747, + "eval_runtime": 37.126, + "eval_samples_per_second": 26.935, + "eval_steps_per_second": 1.131, + "step": 1750 + }, + { + "epoch": 1.9189041095890411, + "grad_norm": 2.970879554748535, + "learning_rate": 2.032900081234769e-06, + "loss": 1.6183, + "step": 1751 + }, + { + "epoch": 1.92, + "grad_norm": 2.774184465408325, + "learning_rate": 2.030869212022746e-06, + "loss": 1.6189, + "step": 1752 + }, + { + "epoch": 1.921095890410959, + "grad_norm": 2.7263576984405518, + "learning_rate": 2.028838342810723e-06, + "loss": 1.5613, + "step": 1753 + }, + { + "epoch": 1.9221917808219178, + "grad_norm": 2.6498911380767822, + "learning_rate": 2.0268074735987004e-06, + "loss": 1.5763, + "step": 1754 + }, + { + "epoch": 1.9232876712328766, + "grad_norm": 2.6792657375335693, + "learning_rate": 2.0247766043866777e-06, + "loss": 1.5655, + "step": 1755 + }, + { + "epoch": 1.9243835616438356, + "grad_norm": 2.7419450283050537, + "learning_rate": 2.022745735174655e-06, + "loss": 1.621, + "step": 1756 + }, + { + "epoch": 1.9254794520547946, + "grad_norm": 2.5262041091918945, + "learning_rate": 2.020714865962632e-06, + "loss": 1.5942, + "step": 1757 + }, + { + "epoch": 1.9265753424657535, + "grad_norm": 2.8236467838287354, + "learning_rate": 2.0186839967506095e-06, + "loss": 1.6178, + "step": 1758 + }, + { + "epoch": 1.9276712328767123, + "grad_norm": 2.7799246311187744, + "learning_rate": 2.0166531275385868e-06, + "loss": 1.5884, + "step": 1759 + }, + { + "epoch": 1.9287671232876713, + "grad_norm": 2.9392249584198, + "learning_rate": 2.014622258326564e-06, + "loss": 1.5414, + "step": 1760 + }, + { + "epoch": 1.9287671232876713, + "eval_loss": 1.8136014938354492, + "eval_runtime": 37.1178, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 1760 + }, + { + "epoch": 1.9298630136986301, + "grad_norm": 2.7029058933258057, + "learning_rate": 2.0125913891145413e-06, + "loss": 1.6039, + "step": 1761 + }, + { + "epoch": 1.930958904109589, + "grad_norm": 3.024052619934082, + "learning_rate": 2.0105605199025186e-06, + "loss": 1.589, + "step": 1762 + }, + { + "epoch": 1.932054794520548, + "grad_norm": 2.9040815830230713, + "learning_rate": 2.008529650690496e-06, + "loss": 1.565, + "step": 1763 + }, + { + "epoch": 1.933150684931507, + "grad_norm": 2.8503386974334717, + "learning_rate": 2.0064987814784727e-06, + "loss": 1.6156, + "step": 1764 + }, + { + "epoch": 1.9342465753424658, + "grad_norm": 2.741941452026367, + "learning_rate": 2.00446791226645e-06, + "loss": 1.5651, + "step": 1765 + }, + { + "epoch": 1.9353424657534246, + "grad_norm": 2.6356847286224365, + "learning_rate": 2.0024370430544273e-06, + "loss": 1.539, + "step": 1766 + }, + { + "epoch": 1.9364383561643836, + "grad_norm": 2.6983048915863037, + "learning_rate": 2.0004061738424045e-06, + "loss": 1.5999, + "step": 1767 + }, + { + "epoch": 1.9375342465753425, + "grad_norm": 2.739579916000366, + "learning_rate": 1.998375304630382e-06, + "loss": 1.636, + "step": 1768 + }, + { + "epoch": 1.9386301369863013, + "grad_norm": 2.754481792449951, + "learning_rate": 1.996344435418359e-06, + "loss": 1.5989, + "step": 1769 + }, + { + "epoch": 1.9397260273972603, + "grad_norm": 2.7145771980285645, + "learning_rate": 1.9943135662063364e-06, + "loss": 1.5997, + "step": 1770 + }, + { + "epoch": 1.9397260273972603, + "eval_loss": 1.8135613203048706, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 1770 + }, + { + "epoch": 1.9408219178082193, + "grad_norm": 2.7553305625915527, + "learning_rate": 1.9922826969943136e-06, + "loss": 1.6134, + "step": 1771 + }, + { + "epoch": 1.941917808219178, + "grad_norm": 2.771489381790161, + "learning_rate": 1.990251827782291e-06, + "loss": 1.6248, + "step": 1772 + }, + { + "epoch": 1.943013698630137, + "grad_norm": 2.7514898777008057, + "learning_rate": 1.988220958570268e-06, + "loss": 1.6117, + "step": 1773 + }, + { + "epoch": 1.944109589041096, + "grad_norm": 2.7693169116973877, + "learning_rate": 1.9861900893582455e-06, + "loss": 1.5963, + "step": 1774 + }, + { + "epoch": 1.9452054794520548, + "grad_norm": 2.736459255218506, + "learning_rate": 1.9841592201462227e-06, + "loss": 1.5933, + "step": 1775 + }, + { + "epoch": 1.9463013698630136, + "grad_norm": 2.595576286315918, + "learning_rate": 1.9821283509342e-06, + "loss": 1.552, + "step": 1776 + }, + { + "epoch": 1.9473972602739726, + "grad_norm": 2.712052583694458, + "learning_rate": 1.9800974817221773e-06, + "loss": 1.5547, + "step": 1777 + }, + { + "epoch": 1.9484931506849315, + "grad_norm": 2.7781100273132324, + "learning_rate": 1.9780666125101546e-06, + "loss": 1.6022, + "step": 1778 + }, + { + "epoch": 1.9495890410958903, + "grad_norm": 2.6884191036224365, + "learning_rate": 1.976035743298132e-06, + "loss": 1.5726, + "step": 1779 + }, + { + "epoch": 1.9506849315068493, + "grad_norm": 2.9216461181640625, + "learning_rate": 1.974004874086109e-06, + "loss": 1.6406, + "step": 1780 + }, + { + "epoch": 1.9506849315068493, + "eval_loss": 1.8145784139633179, + "eval_runtime": 37.1768, + "eval_samples_per_second": 26.899, + "eval_steps_per_second": 1.13, + "step": 1780 + }, + { + "epoch": 1.9517808219178083, + "grad_norm": 2.6188578605651855, + "learning_rate": 1.9719740048740864e-06, + "loss": 1.5474, + "step": 1781 + }, + { + "epoch": 1.9528767123287671, + "grad_norm": 2.6220829486846924, + "learning_rate": 1.9699431356620637e-06, + "loss": 1.5831, + "step": 1782 + }, + { + "epoch": 1.953972602739726, + "grad_norm": 2.639695405960083, + "learning_rate": 1.967912266450041e-06, + "loss": 1.5621, + "step": 1783 + }, + { + "epoch": 1.955068493150685, + "grad_norm": 3.0538783073425293, + "learning_rate": 1.9658813972380182e-06, + "loss": 1.552, + "step": 1784 + }, + { + "epoch": 1.9561643835616438, + "grad_norm": 2.693957567214966, + "learning_rate": 1.9638505280259955e-06, + "loss": 1.6789, + "step": 1785 + }, + { + "epoch": 1.9572602739726026, + "grad_norm": 2.6801669597625732, + "learning_rate": 1.9618196588139728e-06, + "loss": 1.5527, + "step": 1786 + }, + { + "epoch": 1.9583561643835616, + "grad_norm": 2.819408416748047, + "learning_rate": 1.95978878960195e-06, + "loss": 1.5887, + "step": 1787 + }, + { + "epoch": 1.9594520547945207, + "grad_norm": 2.819399833679199, + "learning_rate": 1.9577579203899273e-06, + "loss": 1.6551, + "step": 1788 + }, + { + "epoch": 1.9605479452054795, + "grad_norm": 3.0374770164489746, + "learning_rate": 1.955727051177904e-06, + "loss": 1.614, + "step": 1789 + }, + { + "epoch": 1.9616438356164383, + "grad_norm": 3.0710079669952393, + "learning_rate": 1.9536961819658815e-06, + "loss": 1.5721, + "step": 1790 + }, + { + "epoch": 1.9616438356164383, + "eval_loss": 1.8166357278823853, + "eval_runtime": 37.1684, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 1790 + }, + { + "epoch": 1.9627397260273973, + "grad_norm": 2.6994147300720215, + "learning_rate": 1.9516653127538587e-06, + "loss": 1.5532, + "step": 1791 + }, + { + "epoch": 1.9638356164383561, + "grad_norm": 2.755030632019043, + "learning_rate": 1.949634443541836e-06, + "loss": 1.5565, + "step": 1792 + }, + { + "epoch": 1.964931506849315, + "grad_norm": 2.80184006690979, + "learning_rate": 1.9476035743298133e-06, + "loss": 1.545, + "step": 1793 + }, + { + "epoch": 1.966027397260274, + "grad_norm": 2.7419755458831787, + "learning_rate": 1.9455727051177906e-06, + "loss": 1.6121, + "step": 1794 + }, + { + "epoch": 1.967123287671233, + "grad_norm": 2.754608392715454, + "learning_rate": 1.943541835905768e-06, + "loss": 1.5745, + "step": 1795 + }, + { + "epoch": 1.9682191780821918, + "grad_norm": 2.7504048347473145, + "learning_rate": 1.941510966693745e-06, + "loss": 1.6303, + "step": 1796 + }, + { + "epoch": 1.9693150684931506, + "grad_norm": 2.636611223220825, + "learning_rate": 1.9394800974817224e-06, + "loss": 1.5476, + "step": 1797 + }, + { + "epoch": 1.9704109589041097, + "grad_norm": 2.887341022491455, + "learning_rate": 1.9374492282696997e-06, + "loss": 1.6227, + "step": 1798 + }, + { + "epoch": 1.9715068493150685, + "grad_norm": 2.6758203506469727, + "learning_rate": 1.935418359057677e-06, + "loss": 1.6361, + "step": 1799 + }, + { + "epoch": 1.9726027397260273, + "grad_norm": 2.5958707332611084, + "learning_rate": 1.9333874898456542e-06, + "loss": 1.6114, + "step": 1800 + }, + { + "epoch": 1.9726027397260273, + "eval_loss": 1.8168163299560547, + "eval_runtime": 37.2242, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 1800 + }, + { + "epoch": 1.9736986301369863, + "grad_norm": 2.709763288497925, + "learning_rate": 1.931356620633631e-06, + "loss": 1.6376, + "step": 1801 + }, + { + "epoch": 1.9747945205479454, + "grad_norm": 2.7863962650299072, + "learning_rate": 1.9293257514216083e-06, + "loss": 1.523, + "step": 1802 + }, + { + "epoch": 1.9758904109589042, + "grad_norm": 2.660451889038086, + "learning_rate": 1.9272948822095856e-06, + "loss": 1.6003, + "step": 1803 + }, + { + "epoch": 1.976986301369863, + "grad_norm": 2.748889684677124, + "learning_rate": 1.925264012997563e-06, + "loss": 1.5442, + "step": 1804 + }, + { + "epoch": 1.978082191780822, + "grad_norm": 2.7090959548950195, + "learning_rate": 1.92323314378554e-06, + "loss": 1.5801, + "step": 1805 + }, + { + "epoch": 1.9791780821917808, + "grad_norm": 2.864243745803833, + "learning_rate": 1.9212022745735174e-06, + "loss": 1.6183, + "step": 1806 + }, + { + "epoch": 1.9802739726027396, + "grad_norm": 2.658129930496216, + "learning_rate": 1.9191714053614947e-06, + "loss": 1.5607, + "step": 1807 + }, + { + "epoch": 1.9813698630136987, + "grad_norm": 2.7485649585723877, + "learning_rate": 1.917140536149472e-06, + "loss": 1.5552, + "step": 1808 + }, + { + "epoch": 1.9824657534246577, + "grad_norm": 2.5688955783843994, + "learning_rate": 1.9151096669374493e-06, + "loss": 1.5358, + "step": 1809 + }, + { + "epoch": 1.9835616438356163, + "grad_norm": 2.7082202434539795, + "learning_rate": 1.9130787977254266e-06, + "loss": 1.612, + "step": 1810 + }, + { + "epoch": 1.9835616438356163, + "eval_loss": 1.8135063648223877, + "eval_runtime": 36.8909, + "eval_samples_per_second": 27.107, + "eval_steps_per_second": 1.138, + "step": 1810 + }, + { + "epoch": 1.9846575342465753, + "grad_norm": 2.872128486633301, + "learning_rate": 1.911047928513404e-06, + "loss": 1.6016, + "step": 1811 + }, + { + "epoch": 1.9857534246575343, + "grad_norm": 2.664663553237915, + "learning_rate": 1.909017059301381e-06, + "loss": 1.5586, + "step": 1812 + }, + { + "epoch": 1.9868493150684932, + "grad_norm": 2.6917688846588135, + "learning_rate": 1.9069861900893586e-06, + "loss": 1.5572, + "step": 1813 + }, + { + "epoch": 1.987945205479452, + "grad_norm": 2.688720941543579, + "learning_rate": 1.9049553208773354e-06, + "loss": 1.6015, + "step": 1814 + }, + { + "epoch": 1.989041095890411, + "grad_norm": 2.5801217555999756, + "learning_rate": 1.9029244516653127e-06, + "loss": 1.5665, + "step": 1815 + }, + { + "epoch": 1.9901369863013698, + "grad_norm": 2.596057653427124, + "learning_rate": 1.90089358245329e-06, + "loss": 1.5263, + "step": 1816 + }, + { + "epoch": 1.9912328767123286, + "grad_norm": 2.671658992767334, + "learning_rate": 1.8988627132412673e-06, + "loss": 1.5633, + "step": 1817 + }, + { + "epoch": 1.9923287671232877, + "grad_norm": 2.881671667098999, + "learning_rate": 1.8968318440292445e-06, + "loss": 1.6169, + "step": 1818 + }, + { + "epoch": 1.9934246575342467, + "grad_norm": 2.7213528156280518, + "learning_rate": 1.8948009748172218e-06, + "loss": 1.5442, + "step": 1819 + }, + { + "epoch": 1.9945205479452055, + "grad_norm": 2.721205949783325, + "learning_rate": 1.892770105605199e-06, + "loss": 1.5904, + "step": 1820 + }, + { + "epoch": 1.9945205479452055, + "eval_loss": 1.809995174407959, + "eval_runtime": 37.1476, + "eval_samples_per_second": 26.92, + "eval_steps_per_second": 1.131, + "step": 1820 + }, + { + "epoch": 1.9956164383561643, + "grad_norm": 2.6542251110076904, + "learning_rate": 1.8907392363931764e-06, + "loss": 1.6187, + "step": 1821 + }, + { + "epoch": 1.9967123287671233, + "grad_norm": 2.6850414276123047, + "learning_rate": 1.8887083671811536e-06, + "loss": 1.5728, + "step": 1822 + }, + { + "epoch": 1.9978082191780822, + "grad_norm": 2.9420509338378906, + "learning_rate": 1.886677497969131e-06, + "loss": 1.6023, + "step": 1823 + }, + { + "epoch": 1.998904109589041, + "grad_norm": 2.6539506912231445, + "learning_rate": 1.8846466287571082e-06, + "loss": 1.4886, + "step": 1824 + }, + { + "epoch": 2.0, + "grad_norm": 2.7468559741973877, + "learning_rate": 1.8826157595450855e-06, + "loss": 1.5968, + "step": 1825 + }, + { + "epoch": 2.001095890410959, + "grad_norm": 2.9334170818328857, + "learning_rate": 1.8805848903330625e-06, + "loss": 1.5099, + "step": 1826 + }, + { + "epoch": 2.0021917808219176, + "grad_norm": 3.067456007003784, + "learning_rate": 1.8785540211210398e-06, + "loss": 1.6651, + "step": 1827 + }, + { + "epoch": 2.0032876712328767, + "grad_norm": 2.8721303939819336, + "learning_rate": 1.876523151909017e-06, + "loss": 1.5666, + "step": 1828 + }, + { + "epoch": 2.0043835616438357, + "grad_norm": 2.8489670753479004, + "learning_rate": 1.8744922826969944e-06, + "loss": 1.5358, + "step": 1829 + }, + { + "epoch": 2.0054794520547947, + "grad_norm": 2.8652217388153076, + "learning_rate": 1.8724614134849716e-06, + "loss": 1.5242, + "step": 1830 + }, + { + "epoch": 2.0054794520547947, + "eval_loss": 1.8187663555145264, + "eval_runtime": 37.1841, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 1830 + }, + { + "epoch": 2.0065753424657533, + "grad_norm": 3.000204086303711, + "learning_rate": 1.870430544272949e-06, + "loss": 1.546, + "step": 1831 + }, + { + "epoch": 2.0076712328767123, + "grad_norm": 3.344660758972168, + "learning_rate": 1.8683996750609262e-06, + "loss": 1.6452, + "step": 1832 + }, + { + "epoch": 2.0087671232876714, + "grad_norm": 3.3041818141937256, + "learning_rate": 1.8663688058489035e-06, + "loss": 1.5284, + "step": 1833 + }, + { + "epoch": 2.00986301369863, + "grad_norm": 2.9112915992736816, + "learning_rate": 1.8643379366368807e-06, + "loss": 1.5534, + "step": 1834 + }, + { + "epoch": 2.010958904109589, + "grad_norm": 2.9359078407287598, + "learning_rate": 1.862307067424858e-06, + "loss": 1.5534, + "step": 1835 + }, + { + "epoch": 2.012054794520548, + "grad_norm": 3.127639055252075, + "learning_rate": 1.8602761982128353e-06, + "loss": 1.5614, + "step": 1836 + }, + { + "epoch": 2.013150684931507, + "grad_norm": 2.8351926803588867, + "learning_rate": 1.8582453290008126e-06, + "loss": 1.598, + "step": 1837 + }, + { + "epoch": 2.0142465753424657, + "grad_norm": 2.81606388092041, + "learning_rate": 1.8562144597887898e-06, + "loss": 1.5262, + "step": 1838 + }, + { + "epoch": 2.0153424657534247, + "grad_norm": 2.8759658336639404, + "learning_rate": 1.854183590576767e-06, + "loss": 1.553, + "step": 1839 + }, + { + "epoch": 2.0164383561643837, + "grad_norm": 2.9519526958465576, + "learning_rate": 1.8521527213647442e-06, + "loss": 1.5998, + "step": 1840 + }, + { + "epoch": 2.0164383561643837, + "eval_loss": 1.8135932683944702, + "eval_runtime": 37.2983, + "eval_samples_per_second": 26.811, + "eval_steps_per_second": 1.126, + "step": 1840 + }, + { + "epoch": 2.0175342465753423, + "grad_norm": 2.8636465072631836, + "learning_rate": 1.8501218521527215e-06, + "loss": 1.5193, + "step": 1841 + }, + { + "epoch": 2.0186301369863013, + "grad_norm": 2.949354648590088, + "learning_rate": 1.8480909829406987e-06, + "loss": 1.4694, + "step": 1842 + }, + { + "epoch": 2.0197260273972604, + "grad_norm": 3.0157501697540283, + "learning_rate": 1.846060113728676e-06, + "loss": 1.5175, + "step": 1843 + }, + { + "epoch": 2.020821917808219, + "grad_norm": 3.026820421218872, + "learning_rate": 1.8440292445166533e-06, + "loss": 1.559, + "step": 1844 + }, + { + "epoch": 2.021917808219178, + "grad_norm": 2.980776786804199, + "learning_rate": 1.8419983753046306e-06, + "loss": 1.4992, + "step": 1845 + }, + { + "epoch": 2.023013698630137, + "grad_norm": 2.955500364303589, + "learning_rate": 1.8399675060926078e-06, + "loss": 1.5911, + "step": 1846 + }, + { + "epoch": 2.024109589041096, + "grad_norm": 3.005505323410034, + "learning_rate": 1.8379366368805851e-06, + "loss": 1.5638, + "step": 1847 + }, + { + "epoch": 2.0252054794520546, + "grad_norm": 2.8262405395507812, + "learning_rate": 1.8359057676685624e-06, + "loss": 1.5291, + "step": 1848 + }, + { + "epoch": 2.0263013698630137, + "grad_norm": 2.7804510593414307, + "learning_rate": 1.8338748984565397e-06, + "loss": 1.5218, + "step": 1849 + }, + { + "epoch": 2.0273972602739727, + "grad_norm": 3.0078423023223877, + "learning_rate": 1.831844029244517e-06, + "loss": 1.542, + "step": 1850 + }, + { + "epoch": 2.0273972602739727, + "eval_loss": 1.8123717308044434, + "eval_runtime": 37.3156, + "eval_samples_per_second": 26.798, + "eval_steps_per_second": 1.126, + "step": 1850 + }, + { + "epoch": 2.0284931506849313, + "grad_norm": 2.9625120162963867, + "learning_rate": 1.829813160032494e-06, + "loss": 1.5202, + "step": 1851 + }, + { + "epoch": 2.0295890410958903, + "grad_norm": 2.8305511474609375, + "learning_rate": 1.8277822908204713e-06, + "loss": 1.5289, + "step": 1852 + }, + { + "epoch": 2.0306849315068494, + "grad_norm": 2.9316961765289307, + "learning_rate": 1.8257514216084486e-06, + "loss": 1.5609, + "step": 1853 + }, + { + "epoch": 2.0317808219178084, + "grad_norm": 3.128195285797119, + "learning_rate": 1.8237205523964258e-06, + "loss": 1.5452, + "step": 1854 + }, + { + "epoch": 2.032876712328767, + "grad_norm": 2.9158523082733154, + "learning_rate": 1.8216896831844031e-06, + "loss": 1.552, + "step": 1855 + }, + { + "epoch": 2.033972602739726, + "grad_norm": 2.905848264694214, + "learning_rate": 1.8196588139723804e-06, + "loss": 1.5976, + "step": 1856 + }, + { + "epoch": 2.035068493150685, + "grad_norm": 2.9548468589782715, + "learning_rate": 1.8176279447603577e-06, + "loss": 1.5167, + "step": 1857 + }, + { + "epoch": 2.0361643835616436, + "grad_norm": 2.8970632553100586, + "learning_rate": 1.815597075548335e-06, + "loss": 1.5116, + "step": 1858 + }, + { + "epoch": 2.0372602739726027, + "grad_norm": 3.001655340194702, + "learning_rate": 1.8135662063363122e-06, + "loss": 1.5591, + "step": 1859 + }, + { + "epoch": 2.0383561643835617, + "grad_norm": 3.086770534515381, + "learning_rate": 1.8115353371242895e-06, + "loss": 1.5899, + "step": 1860 + }, + { + "epoch": 2.0383561643835617, + "eval_loss": 1.8146355152130127, + "eval_runtime": 37.2779, + "eval_samples_per_second": 26.826, + "eval_steps_per_second": 1.127, + "step": 1860 + }, + { + "epoch": 2.0394520547945207, + "grad_norm": 2.8270034790039062, + "learning_rate": 1.8095044679122668e-06, + "loss": 1.4858, + "step": 1861 + }, + { + "epoch": 2.0405479452054793, + "grad_norm": 2.981127977371216, + "learning_rate": 1.807473598700244e-06, + "loss": 1.5276, + "step": 1862 + }, + { + "epoch": 2.0416438356164384, + "grad_norm": 2.867921829223633, + "learning_rate": 1.805442729488221e-06, + "loss": 1.5051, + "step": 1863 + }, + { + "epoch": 2.0427397260273974, + "grad_norm": 2.7737348079681396, + "learning_rate": 1.8034118602761982e-06, + "loss": 1.5181, + "step": 1864 + }, + { + "epoch": 2.043835616438356, + "grad_norm": 2.9307985305786133, + "learning_rate": 1.8013809910641754e-06, + "loss": 1.5358, + "step": 1865 + }, + { + "epoch": 2.044931506849315, + "grad_norm": 2.9373624324798584, + "learning_rate": 1.7993501218521527e-06, + "loss": 1.5356, + "step": 1866 + }, + { + "epoch": 2.046027397260274, + "grad_norm": 2.8827672004699707, + "learning_rate": 1.79731925264013e-06, + "loss": 1.5163, + "step": 1867 + }, + { + "epoch": 2.047123287671233, + "grad_norm": 2.9692344665527344, + "learning_rate": 1.7952883834281073e-06, + "loss": 1.5645, + "step": 1868 + }, + { + "epoch": 2.0482191780821917, + "grad_norm": 2.810917377471924, + "learning_rate": 1.7932575142160846e-06, + "loss": 1.5282, + "step": 1869 + }, + { + "epoch": 2.0493150684931507, + "grad_norm": 2.85056471824646, + "learning_rate": 1.7912266450040618e-06, + "loss": 1.5598, + "step": 1870 + }, + { + "epoch": 2.0493150684931507, + "eval_loss": 1.8124264478683472, + "eval_runtime": 37.1743, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 1870 + }, + { + "epoch": 2.0504109589041097, + "grad_norm": 2.895660877227783, + "learning_rate": 1.789195775792039e-06, + "loss": 1.5634, + "step": 1871 + }, + { + "epoch": 2.0515068493150683, + "grad_norm": 2.9039483070373535, + "learning_rate": 1.7871649065800164e-06, + "loss": 1.4754, + "step": 1872 + }, + { + "epoch": 2.0526027397260274, + "grad_norm": 3.049464225769043, + "learning_rate": 1.7851340373679937e-06, + "loss": 1.6061, + "step": 1873 + }, + { + "epoch": 2.0536986301369864, + "grad_norm": 2.877824306488037, + "learning_rate": 1.783103168155971e-06, + "loss": 1.5068, + "step": 1874 + }, + { + "epoch": 2.0547945205479454, + "grad_norm": 3.3161075115203857, + "learning_rate": 1.7810722989439482e-06, + "loss": 1.5286, + "step": 1875 + }, + { + "epoch": 2.055890410958904, + "grad_norm": 2.934148073196411, + "learning_rate": 1.7790414297319253e-06, + "loss": 1.5388, + "step": 1876 + }, + { + "epoch": 2.056986301369863, + "grad_norm": 2.9859094619750977, + "learning_rate": 1.7770105605199025e-06, + "loss": 1.5157, + "step": 1877 + }, + { + "epoch": 2.058082191780822, + "grad_norm": 2.916668653488159, + "learning_rate": 1.7749796913078798e-06, + "loss": 1.5034, + "step": 1878 + }, + { + "epoch": 2.0591780821917807, + "grad_norm": 3.16752028465271, + "learning_rate": 1.772948822095857e-06, + "loss": 1.5483, + "step": 1879 + }, + { + "epoch": 2.0602739726027397, + "grad_norm": 3.0830936431884766, + "learning_rate": 1.7709179528838344e-06, + "loss": 1.5942, + "step": 1880 + }, + { + "epoch": 2.0602739726027397, + "eval_loss": 1.8144135475158691, + "eval_runtime": 37.1761, + "eval_samples_per_second": 26.899, + "eval_steps_per_second": 1.13, + "step": 1880 + }, + { + "epoch": 2.0613698630136987, + "grad_norm": 3.1621158123016357, + "learning_rate": 1.7688870836718116e-06, + "loss": 1.5735, + "step": 1881 + }, + { + "epoch": 2.0624657534246573, + "grad_norm": 3.0182595252990723, + "learning_rate": 1.766856214459789e-06, + "loss": 1.5412, + "step": 1882 + }, + { + "epoch": 2.0635616438356164, + "grad_norm": 2.912668228149414, + "learning_rate": 1.7648253452477662e-06, + "loss": 1.5328, + "step": 1883 + }, + { + "epoch": 2.0646575342465754, + "grad_norm": 3.061091423034668, + "learning_rate": 1.7627944760357435e-06, + "loss": 1.5111, + "step": 1884 + }, + { + "epoch": 2.0657534246575344, + "grad_norm": 2.9681029319763184, + "learning_rate": 1.7607636068237208e-06, + "loss": 1.5983, + "step": 1885 + }, + { + "epoch": 2.066849315068493, + "grad_norm": 3.1734211444854736, + "learning_rate": 1.758732737611698e-06, + "loss": 1.5529, + "step": 1886 + }, + { + "epoch": 2.067945205479452, + "grad_norm": 3.0264222621917725, + "learning_rate": 1.7567018683996753e-06, + "loss": 1.5488, + "step": 1887 + }, + { + "epoch": 2.069041095890411, + "grad_norm": 2.9788169860839844, + "learning_rate": 1.7546709991876524e-06, + "loss": 1.5351, + "step": 1888 + }, + { + "epoch": 2.0701369863013697, + "grad_norm": 2.9770145416259766, + "learning_rate": 1.7526401299756296e-06, + "loss": 1.5232, + "step": 1889 + }, + { + "epoch": 2.0712328767123287, + "grad_norm": 3.2160613536834717, + "learning_rate": 1.750609260763607e-06, + "loss": 1.5319, + "step": 1890 + }, + { + "epoch": 2.0712328767123287, + "eval_loss": 1.8123860359191895, + "eval_runtime": 37.2075, + "eval_samples_per_second": 26.876, + "eval_steps_per_second": 1.129, + "step": 1890 + }, + { + "epoch": 2.0723287671232877, + "grad_norm": 2.98433518409729, + "learning_rate": 1.7485783915515842e-06, + "loss": 1.541, + "step": 1891 + }, + { + "epoch": 2.0734246575342468, + "grad_norm": 3.042020082473755, + "learning_rate": 1.7465475223395615e-06, + "loss": 1.5334, + "step": 1892 + }, + { + "epoch": 2.0745205479452054, + "grad_norm": 2.878483533859253, + "learning_rate": 1.7445166531275387e-06, + "loss": 1.4854, + "step": 1893 + }, + { + "epoch": 2.0756164383561644, + "grad_norm": 3.219987154006958, + "learning_rate": 1.742485783915516e-06, + "loss": 1.6235, + "step": 1894 + }, + { + "epoch": 2.0767123287671234, + "grad_norm": 2.9204180240631104, + "learning_rate": 1.7404549147034933e-06, + "loss": 1.6166, + "step": 1895 + }, + { + "epoch": 2.077808219178082, + "grad_norm": 2.873898506164551, + "learning_rate": 1.7384240454914706e-06, + "loss": 1.4644, + "step": 1896 + }, + { + "epoch": 2.078904109589041, + "grad_norm": 3.07304048538208, + "learning_rate": 1.7363931762794478e-06, + "loss": 1.5525, + "step": 1897 + }, + { + "epoch": 2.08, + "grad_norm": 3.065629482269287, + "learning_rate": 1.7343623070674251e-06, + "loss": 1.574, + "step": 1898 + }, + { + "epoch": 2.081095890410959, + "grad_norm": 2.8665778636932373, + "learning_rate": 1.7323314378554024e-06, + "loss": 1.5296, + "step": 1899 + }, + { + "epoch": 2.0821917808219177, + "grad_norm": 2.8402345180511475, + "learning_rate": 1.7303005686433797e-06, + "loss": 1.5473, + "step": 1900 + }, + { + "epoch": 2.0821917808219177, + "eval_loss": 1.8114396333694458, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 1900 + }, + { + "epoch": 2.0832876712328767, + "grad_norm": 2.91286301612854, + "learning_rate": 1.7282696994313567e-06, + "loss": 1.5405, + "step": 1901 + }, + { + "epoch": 2.0843835616438358, + "grad_norm": 2.966437578201294, + "learning_rate": 1.726238830219334e-06, + "loss": 1.5389, + "step": 1902 + }, + { + "epoch": 2.0854794520547943, + "grad_norm": 3.0182693004608154, + "learning_rate": 1.7242079610073113e-06, + "loss": 1.532, + "step": 1903 + }, + { + "epoch": 2.0865753424657534, + "grad_norm": 3.0323562622070312, + "learning_rate": 1.7221770917952886e-06, + "loss": 1.5946, + "step": 1904 + }, + { + "epoch": 2.0876712328767124, + "grad_norm": 2.9909002780914307, + "learning_rate": 1.7201462225832658e-06, + "loss": 1.5451, + "step": 1905 + }, + { + "epoch": 2.0887671232876714, + "grad_norm": 2.872896432876587, + "learning_rate": 1.7181153533712431e-06, + "loss": 1.5606, + "step": 1906 + }, + { + "epoch": 2.08986301369863, + "grad_norm": 2.8841099739074707, + "learning_rate": 1.7160844841592204e-06, + "loss": 1.4634, + "step": 1907 + }, + { + "epoch": 2.090958904109589, + "grad_norm": 2.9734718799591064, + "learning_rate": 1.7140536149471977e-06, + "loss": 1.4674, + "step": 1908 + }, + { + "epoch": 2.092054794520548, + "grad_norm": 2.7336618900299072, + "learning_rate": 1.712022745735175e-06, + "loss": 1.5179, + "step": 1909 + }, + { + "epoch": 2.0931506849315067, + "grad_norm": 2.7525081634521484, + "learning_rate": 1.7099918765231522e-06, + "loss": 1.5298, + "step": 1910 + }, + { + "epoch": 2.0931506849315067, + "eval_loss": 1.8111896514892578, + "eval_runtime": 36.8371, + "eval_samples_per_second": 27.147, + "eval_steps_per_second": 1.14, + "step": 1910 + }, + { + "epoch": 2.0942465753424657, + "grad_norm": 2.8897578716278076, + "learning_rate": 1.7079610073111295e-06, + "loss": 1.5392, + "step": 1911 + }, + { + "epoch": 2.0953424657534248, + "grad_norm": 2.96352219581604, + "learning_rate": 1.7059301380991068e-06, + "loss": 1.4838, + "step": 1912 + }, + { + "epoch": 2.0964383561643833, + "grad_norm": 3.0692732334136963, + "learning_rate": 1.7038992688870836e-06, + "loss": 1.5391, + "step": 1913 + }, + { + "epoch": 2.0975342465753424, + "grad_norm": 2.868380308151245, + "learning_rate": 1.701868399675061e-06, + "loss": 1.5429, + "step": 1914 + }, + { + "epoch": 2.0986301369863014, + "grad_norm": 3.111435651779175, + "learning_rate": 1.6998375304630382e-06, + "loss": 1.562, + "step": 1915 + }, + { + "epoch": 2.0997260273972604, + "grad_norm": 2.9990973472595215, + "learning_rate": 1.6978066612510155e-06, + "loss": 1.6129, + "step": 1916 + }, + { + "epoch": 2.100821917808219, + "grad_norm": 2.896906614303589, + "learning_rate": 1.6957757920389927e-06, + "loss": 1.5649, + "step": 1917 + }, + { + "epoch": 2.101917808219178, + "grad_norm": 2.966764450073242, + "learning_rate": 1.69374492282697e-06, + "loss": 1.5224, + "step": 1918 + }, + { + "epoch": 2.103013698630137, + "grad_norm": 2.98482608795166, + "learning_rate": 1.6917140536149473e-06, + "loss": 1.5204, + "step": 1919 + }, + { + "epoch": 2.1041095890410957, + "grad_norm": 3.0192346572875977, + "learning_rate": 1.6896831844029246e-06, + "loss": 1.605, + "step": 1920 + }, + { + "epoch": 2.1041095890410957, + "eval_loss": 1.8114374876022339, + "eval_runtime": 37.1796, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 1920 + }, + { + "epoch": 2.1052054794520547, + "grad_norm": 2.9359049797058105, + "learning_rate": 1.6876523151909018e-06, + "loss": 1.5317, + "step": 1921 + }, + { + "epoch": 2.1063013698630138, + "grad_norm": 3.087991952896118, + "learning_rate": 1.6856214459788791e-06, + "loss": 1.5458, + "step": 1922 + }, + { + "epoch": 2.107397260273973, + "grad_norm": 3.079904794692993, + "learning_rate": 1.6835905767668564e-06, + "loss": 1.5053, + "step": 1923 + }, + { + "epoch": 2.1084931506849314, + "grad_norm": 3.2002458572387695, + "learning_rate": 1.6815597075548337e-06, + "loss": 1.5276, + "step": 1924 + }, + { + "epoch": 2.1095890410958904, + "grad_norm": 3.1295740604400635, + "learning_rate": 1.679528838342811e-06, + "loss": 1.5831, + "step": 1925 + }, + { + "epoch": 2.1106849315068494, + "grad_norm": 2.9871461391448975, + "learning_rate": 1.677497969130788e-06, + "loss": 1.5194, + "step": 1926 + }, + { + "epoch": 2.111780821917808, + "grad_norm": 2.9117379188537598, + "learning_rate": 1.6754670999187653e-06, + "loss": 1.5178, + "step": 1927 + }, + { + "epoch": 2.112876712328767, + "grad_norm": 2.9347078800201416, + "learning_rate": 1.6734362307067426e-06, + "loss": 1.5409, + "step": 1928 + }, + { + "epoch": 2.113972602739726, + "grad_norm": 2.9074113368988037, + "learning_rate": 1.6714053614947198e-06, + "loss": 1.4845, + "step": 1929 + }, + { + "epoch": 2.115068493150685, + "grad_norm": 2.993051528930664, + "learning_rate": 1.669374492282697e-06, + "loss": 1.4716, + "step": 1930 + }, + { + "epoch": 2.115068493150685, + "eval_loss": 1.8102058172225952, + "eval_runtime": 37.2473, + "eval_samples_per_second": 26.848, + "eval_steps_per_second": 1.128, + "step": 1930 + }, + { + "epoch": 2.1161643835616437, + "grad_norm": 3.0071909427642822, + "learning_rate": 1.6673436230706744e-06, + "loss": 1.5798, + "step": 1931 + }, + { + "epoch": 2.1172602739726027, + "grad_norm": 3.1663055419921875, + "learning_rate": 1.6653127538586517e-06, + "loss": 1.5565, + "step": 1932 + }, + { + "epoch": 2.118356164383562, + "grad_norm": 3.0547966957092285, + "learning_rate": 1.663281884646629e-06, + "loss": 1.537, + "step": 1933 + }, + { + "epoch": 2.1194520547945204, + "grad_norm": 3.224301815032959, + "learning_rate": 1.6612510154346062e-06, + "loss": 1.5286, + "step": 1934 + }, + { + "epoch": 2.1205479452054794, + "grad_norm": 3.0271787643432617, + "learning_rate": 1.6592201462225835e-06, + "loss": 1.5868, + "step": 1935 + }, + { + "epoch": 2.1216438356164384, + "grad_norm": 2.952291488647461, + "learning_rate": 1.6571892770105608e-06, + "loss": 1.5617, + "step": 1936 + }, + { + "epoch": 2.1227397260273975, + "grad_norm": 3.0080809593200684, + "learning_rate": 1.655158407798538e-06, + "loss": 1.5974, + "step": 1937 + }, + { + "epoch": 2.123835616438356, + "grad_norm": 3.249833106994629, + "learning_rate": 1.653127538586515e-06, + "loss": 1.5344, + "step": 1938 + }, + { + "epoch": 2.124931506849315, + "grad_norm": 3.1032025814056396, + "learning_rate": 1.6510966693744924e-06, + "loss": 1.5711, + "step": 1939 + }, + { + "epoch": 2.126027397260274, + "grad_norm": 3.1881093978881836, + "learning_rate": 1.6490658001624696e-06, + "loss": 1.5828, + "step": 1940 + }, + { + "epoch": 2.126027397260274, + "eval_loss": 1.8083537817001343, + "eval_runtime": 37.1916, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 1940 + }, + { + "epoch": 2.1271232876712327, + "grad_norm": 3.0378284454345703, + "learning_rate": 1.647034930950447e-06, + "loss": 1.5707, + "step": 1941 + }, + { + "epoch": 2.1282191780821917, + "grad_norm": 2.8167002201080322, + "learning_rate": 1.6450040617384242e-06, + "loss": 1.5072, + "step": 1942 + }, + { + "epoch": 2.1293150684931508, + "grad_norm": 2.921473264694214, + "learning_rate": 1.6429731925264015e-06, + "loss": 1.5464, + "step": 1943 + }, + { + "epoch": 2.1304109589041094, + "grad_norm": 2.9117305278778076, + "learning_rate": 1.6409423233143788e-06, + "loss": 1.5638, + "step": 1944 + }, + { + "epoch": 2.1315068493150684, + "grad_norm": 2.9102439880371094, + "learning_rate": 1.638911454102356e-06, + "loss": 1.5266, + "step": 1945 + }, + { + "epoch": 2.1326027397260274, + "grad_norm": 2.8359367847442627, + "learning_rate": 1.6368805848903333e-06, + "loss": 1.5156, + "step": 1946 + }, + { + "epoch": 2.1336986301369865, + "grad_norm": 2.884160041809082, + "learning_rate": 1.6348497156783106e-06, + "loss": 1.5749, + "step": 1947 + }, + { + "epoch": 2.134794520547945, + "grad_norm": 2.921024799346924, + "learning_rate": 1.6328188464662879e-06, + "loss": 1.516, + "step": 1948 + }, + { + "epoch": 2.135890410958904, + "grad_norm": 2.947483539581299, + "learning_rate": 1.6307879772542651e-06, + "loss": 1.5265, + "step": 1949 + }, + { + "epoch": 2.136986301369863, + "grad_norm": 2.9654922485351562, + "learning_rate": 1.628757108042242e-06, + "loss": 1.5414, + "step": 1950 + }, + { + "epoch": 2.136986301369863, + "eval_loss": 1.8112154006958008, + "eval_runtime": 37.2211, + "eval_samples_per_second": 26.866, + "eval_steps_per_second": 1.128, + "step": 1950 + }, + { + "epoch": 2.138082191780822, + "grad_norm": 2.9900968074798584, + "learning_rate": 1.6267262388302193e-06, + "loss": 1.5712, + "step": 1951 + }, + { + "epoch": 2.1391780821917807, + "grad_norm": 2.8663604259490967, + "learning_rate": 1.6246953696181965e-06, + "loss": 1.52, + "step": 1952 + }, + { + "epoch": 2.1402739726027398, + "grad_norm": 3.0792734622955322, + "learning_rate": 1.6226645004061738e-06, + "loss": 1.5461, + "step": 1953 + }, + { + "epoch": 2.141369863013699, + "grad_norm": 3.0100409984588623, + "learning_rate": 1.620633631194151e-06, + "loss": 1.6206, + "step": 1954 + }, + { + "epoch": 2.1424657534246574, + "grad_norm": 3.061314344406128, + "learning_rate": 1.6186027619821284e-06, + "loss": 1.5761, + "step": 1955 + }, + { + "epoch": 2.1435616438356164, + "grad_norm": 2.8961095809936523, + "learning_rate": 1.6165718927701056e-06, + "loss": 1.5169, + "step": 1956 + }, + { + "epoch": 2.1446575342465755, + "grad_norm": 2.965549945831299, + "learning_rate": 1.614541023558083e-06, + "loss": 1.5615, + "step": 1957 + }, + { + "epoch": 2.145753424657534, + "grad_norm": 2.9825284481048584, + "learning_rate": 1.6125101543460602e-06, + "loss": 1.551, + "step": 1958 + }, + { + "epoch": 2.146849315068493, + "grad_norm": 2.950587749481201, + "learning_rate": 1.6104792851340375e-06, + "loss": 1.5412, + "step": 1959 + }, + { + "epoch": 2.147945205479452, + "grad_norm": 2.879544496536255, + "learning_rate": 1.6084484159220147e-06, + "loss": 1.5585, + "step": 1960 + }, + { + "epoch": 2.147945205479452, + "eval_loss": 1.8093459606170654, + "eval_runtime": 37.2932, + "eval_samples_per_second": 26.815, + "eval_steps_per_second": 1.126, + "step": 1960 + }, + { + "epoch": 2.149041095890411, + "grad_norm": 3.104874849319458, + "learning_rate": 1.606417546709992e-06, + "loss": 1.5179, + "step": 1961 + }, + { + "epoch": 2.1501369863013697, + "grad_norm": 2.9908242225646973, + "learning_rate": 1.6043866774979693e-06, + "loss": 1.5385, + "step": 1962 + }, + { + "epoch": 2.1512328767123288, + "grad_norm": 2.97064471244812, + "learning_rate": 1.6023558082859464e-06, + "loss": 1.5361, + "step": 1963 + }, + { + "epoch": 2.152328767123288, + "grad_norm": 2.856841802597046, + "learning_rate": 1.6003249390739236e-06, + "loss": 1.5011, + "step": 1964 + }, + { + "epoch": 2.1534246575342464, + "grad_norm": 3.0110960006713867, + "learning_rate": 1.598294069861901e-06, + "loss": 1.5397, + "step": 1965 + }, + { + "epoch": 2.1545205479452054, + "grad_norm": 3.0178773403167725, + "learning_rate": 1.5962632006498782e-06, + "loss": 1.5169, + "step": 1966 + }, + { + "epoch": 2.1556164383561645, + "grad_norm": 3.137674570083618, + "learning_rate": 1.5942323314378555e-06, + "loss": 1.5976, + "step": 1967 + }, + { + "epoch": 2.1567123287671235, + "grad_norm": 2.86112117767334, + "learning_rate": 1.5922014622258327e-06, + "loss": 1.5135, + "step": 1968 + }, + { + "epoch": 2.157808219178082, + "grad_norm": 2.9500656127929688, + "learning_rate": 1.59017059301381e-06, + "loss": 1.5062, + "step": 1969 + }, + { + "epoch": 2.158904109589041, + "grad_norm": 2.854247570037842, + "learning_rate": 1.5881397238017873e-06, + "loss": 1.5603, + "step": 1970 + }, + { + "epoch": 2.158904109589041, + "eval_loss": 1.809329867362976, + "eval_runtime": 37.157, + "eval_samples_per_second": 26.913, + "eval_steps_per_second": 1.13, + "step": 1970 + }, + { + "epoch": 2.16, + "grad_norm": 2.9294161796569824, + "learning_rate": 1.5861088545897646e-06, + "loss": 1.5428, + "step": 1971 + }, + { + "epoch": 2.1610958904109587, + "grad_norm": 2.833270788192749, + "learning_rate": 1.5840779853777418e-06, + "loss": 1.4925, + "step": 1972 + }, + { + "epoch": 2.1621917808219178, + "grad_norm": 3.1147077083587646, + "learning_rate": 1.5820471161657191e-06, + "loss": 1.542, + "step": 1973 + }, + { + "epoch": 2.163287671232877, + "grad_norm": 3.0233561992645264, + "learning_rate": 1.5800162469536964e-06, + "loss": 1.5136, + "step": 1974 + }, + { + "epoch": 2.1643835616438354, + "grad_norm": 3.1687583923339844, + "learning_rate": 1.5779853777416735e-06, + "loss": 1.5748, + "step": 1975 + }, + { + "epoch": 2.1654794520547944, + "grad_norm": 3.095669746398926, + "learning_rate": 1.5759545085296507e-06, + "loss": 1.5683, + "step": 1976 + }, + { + "epoch": 2.1665753424657535, + "grad_norm": 2.9901936054229736, + "learning_rate": 1.573923639317628e-06, + "loss": 1.5442, + "step": 1977 + }, + { + "epoch": 2.1676712328767125, + "grad_norm": 3.156691312789917, + "learning_rate": 1.5718927701056053e-06, + "loss": 1.5544, + "step": 1978 + }, + { + "epoch": 2.168767123287671, + "grad_norm": 2.9674503803253174, + "learning_rate": 1.5698619008935826e-06, + "loss": 1.5688, + "step": 1979 + }, + { + "epoch": 2.16986301369863, + "grad_norm": 3.0276317596435547, + "learning_rate": 1.5678310316815598e-06, + "loss": 1.5645, + "step": 1980 + }, + { + "epoch": 2.16986301369863, + "eval_loss": 1.806713581085205, + "eval_runtime": 37.1928, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 1980 + }, + { + "epoch": 2.170958904109589, + "grad_norm": 2.8310959339141846, + "learning_rate": 1.5658001624695371e-06, + "loss": 1.5071, + "step": 1981 + }, + { + "epoch": 2.172054794520548, + "grad_norm": 3.0237836837768555, + "learning_rate": 1.5637692932575144e-06, + "loss": 1.5608, + "step": 1982 + }, + { + "epoch": 2.1731506849315068, + "grad_norm": 3.0911290645599365, + "learning_rate": 1.5617384240454917e-06, + "loss": 1.5246, + "step": 1983 + }, + { + "epoch": 2.174246575342466, + "grad_norm": 2.9403584003448486, + "learning_rate": 1.559707554833469e-06, + "loss": 1.5393, + "step": 1984 + }, + { + "epoch": 2.175342465753425, + "grad_norm": 3.1798179149627686, + "learning_rate": 1.5576766856214462e-06, + "loss": 1.5836, + "step": 1985 + }, + { + "epoch": 2.1764383561643834, + "grad_norm": 2.894441843032837, + "learning_rate": 1.5556458164094235e-06, + "loss": 1.4716, + "step": 1986 + }, + { + "epoch": 2.1775342465753424, + "grad_norm": 2.990337371826172, + "learning_rate": 1.5536149471974008e-06, + "loss": 1.5124, + "step": 1987 + }, + { + "epoch": 2.1786301369863015, + "grad_norm": 2.961712598800659, + "learning_rate": 1.5515840779853778e-06, + "loss": 1.4946, + "step": 1988 + }, + { + "epoch": 2.17972602739726, + "grad_norm": 2.9735326766967773, + "learning_rate": 1.549553208773355e-06, + "loss": 1.5398, + "step": 1989 + }, + { + "epoch": 2.180821917808219, + "grad_norm": 3.0465126037597656, + "learning_rate": 1.5475223395613324e-06, + "loss": 1.5162, + "step": 1990 + }, + { + "epoch": 2.180821917808219, + "eval_loss": 1.811614990234375, + "eval_runtime": 37.2251, + "eval_samples_per_second": 26.864, + "eval_steps_per_second": 1.128, + "step": 1990 + }, + { + "epoch": 2.181917808219178, + "grad_norm": 3.0636777877807617, + "learning_rate": 1.5454914703493097e-06, + "loss": 1.5193, + "step": 1991 + }, + { + "epoch": 2.183013698630137, + "grad_norm": 3.131138563156128, + "learning_rate": 1.543460601137287e-06, + "loss": 1.581, + "step": 1992 + }, + { + "epoch": 2.1841095890410958, + "grad_norm": 3.028381824493408, + "learning_rate": 1.5414297319252642e-06, + "loss": 1.539, + "step": 1993 + }, + { + "epoch": 2.185205479452055, + "grad_norm": 3.017573356628418, + "learning_rate": 1.5393988627132415e-06, + "loss": 1.5493, + "step": 1994 + }, + { + "epoch": 2.186301369863014, + "grad_norm": 2.8733723163604736, + "learning_rate": 1.5373679935012188e-06, + "loss": 1.5399, + "step": 1995 + }, + { + "epoch": 2.1873972602739724, + "grad_norm": 3.0504605770111084, + "learning_rate": 1.535337124289196e-06, + "loss": 1.4742, + "step": 1996 + }, + { + "epoch": 2.1884931506849314, + "grad_norm": 2.9904239177703857, + "learning_rate": 1.5333062550771733e-06, + "loss": 1.5289, + "step": 1997 + }, + { + "epoch": 2.1895890410958905, + "grad_norm": 2.9754273891448975, + "learning_rate": 1.5312753858651506e-06, + "loss": 1.6031, + "step": 1998 + }, + { + "epoch": 2.1906849315068495, + "grad_norm": 2.937514066696167, + "learning_rate": 1.5292445166531279e-06, + "loss": 1.5567, + "step": 1999 + }, + { + "epoch": 2.191780821917808, + "grad_norm": 3.119877815246582, + "learning_rate": 1.5272136474411047e-06, + "loss": 1.5453, + "step": 2000 + }, + { + "epoch": 2.191780821917808, + "eval_loss": 1.8095368146896362, + "eval_runtime": 37.314, + "eval_samples_per_second": 26.8, + "eval_steps_per_second": 1.126, + "step": 2000 + }, + { + "epoch": 2.192876712328767, + "grad_norm": 3.086714267730713, + "learning_rate": 1.525182778229082e-06, + "loss": 1.5792, + "step": 2001 + }, + { + "epoch": 2.193972602739726, + "grad_norm": 2.915955066680908, + "learning_rate": 1.5231519090170593e-06, + "loss": 1.5599, + "step": 2002 + }, + { + "epoch": 2.1950684931506848, + "grad_norm": 2.9115076065063477, + "learning_rate": 1.5211210398050365e-06, + "loss": 1.5286, + "step": 2003 + }, + { + "epoch": 2.196164383561644, + "grad_norm": 2.989149808883667, + "learning_rate": 1.5190901705930138e-06, + "loss": 1.5116, + "step": 2004 + }, + { + "epoch": 2.197260273972603, + "grad_norm": 3.0808184146881104, + "learning_rate": 1.517059301380991e-06, + "loss": 1.5568, + "step": 2005 + }, + { + "epoch": 2.198356164383562, + "grad_norm": 2.9627561569213867, + "learning_rate": 1.5150284321689684e-06, + "loss": 1.5237, + "step": 2006 + }, + { + "epoch": 2.1994520547945204, + "grad_norm": 2.8799078464508057, + "learning_rate": 1.5129975629569456e-06, + "loss": 1.5228, + "step": 2007 + }, + { + "epoch": 2.2005479452054795, + "grad_norm": 2.9700374603271484, + "learning_rate": 1.510966693744923e-06, + "loss": 1.5314, + "step": 2008 + }, + { + "epoch": 2.2016438356164385, + "grad_norm": 3.031604290008545, + "learning_rate": 1.5089358245329002e-06, + "loss": 1.5406, + "step": 2009 + }, + { + "epoch": 2.202739726027397, + "grad_norm": 3.101407527923584, + "learning_rate": 1.5069049553208775e-06, + "loss": 1.4971, + "step": 2010 + }, + { + "epoch": 2.202739726027397, + "eval_loss": 1.8108844757080078, + "eval_runtime": 36.8838, + "eval_samples_per_second": 27.112, + "eval_steps_per_second": 1.139, + "step": 2010 + }, + { + "epoch": 2.203835616438356, + "grad_norm": 2.955763578414917, + "learning_rate": 1.5048740861088547e-06, + "loss": 1.5488, + "step": 2011 + }, + { + "epoch": 2.204931506849315, + "grad_norm": 2.8813931941986084, + "learning_rate": 1.502843216896832e-06, + "loss": 1.5664, + "step": 2012 + }, + { + "epoch": 2.206027397260274, + "grad_norm": 3.05414080619812, + "learning_rate": 1.500812347684809e-06, + "loss": 1.4719, + "step": 2013 + }, + { + "epoch": 2.207123287671233, + "grad_norm": 2.9929933547973633, + "learning_rate": 1.4987814784727864e-06, + "loss": 1.5389, + "step": 2014 + }, + { + "epoch": 2.208219178082192, + "grad_norm": 3.0010180473327637, + "learning_rate": 1.4967506092607636e-06, + "loss": 1.5324, + "step": 2015 + }, + { + "epoch": 2.209315068493151, + "grad_norm": 2.8261168003082275, + "learning_rate": 1.494719740048741e-06, + "loss": 1.5042, + "step": 2016 + }, + { + "epoch": 2.2104109589041094, + "grad_norm": 3.0346126556396484, + "learning_rate": 1.4926888708367182e-06, + "loss": 1.5427, + "step": 2017 + }, + { + "epoch": 2.2115068493150685, + "grad_norm": 2.9425783157348633, + "learning_rate": 1.4906580016246955e-06, + "loss": 1.5695, + "step": 2018 + }, + { + "epoch": 2.2126027397260275, + "grad_norm": 3.1421637535095215, + "learning_rate": 1.4886271324126727e-06, + "loss": 1.5791, + "step": 2019 + }, + { + "epoch": 2.213698630136986, + "grad_norm": 3.0382707118988037, + "learning_rate": 1.48659626320065e-06, + "loss": 1.5787, + "step": 2020 + }, + { + "epoch": 2.213698630136986, + "eval_loss": 1.8077536821365356, + "eval_runtime": 37.1811, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 2020 + }, + { + "epoch": 2.214794520547945, + "grad_norm": 2.8756654262542725, + "learning_rate": 1.4845653939886273e-06, + "loss": 1.4515, + "step": 2021 + }, + { + "epoch": 2.215890410958904, + "grad_norm": 2.833338975906372, + "learning_rate": 1.4825345247766046e-06, + "loss": 1.463, + "step": 2022 + }, + { + "epoch": 2.216986301369863, + "grad_norm": 2.884336233139038, + "learning_rate": 1.4805036555645818e-06, + "loss": 1.4809, + "step": 2023 + }, + { + "epoch": 2.2180821917808218, + "grad_norm": 2.998903274536133, + "learning_rate": 1.4784727863525591e-06, + "loss": 1.5163, + "step": 2024 + }, + { + "epoch": 2.219178082191781, + "grad_norm": 2.7807600498199463, + "learning_rate": 1.4764419171405362e-06, + "loss": 1.4888, + "step": 2025 + }, + { + "epoch": 2.22027397260274, + "grad_norm": 3.0383498668670654, + "learning_rate": 1.4744110479285135e-06, + "loss": 1.5659, + "step": 2026 + }, + { + "epoch": 2.2213698630136984, + "grad_norm": 3.162415027618408, + "learning_rate": 1.4723801787164907e-06, + "loss": 1.5648, + "step": 2027 + }, + { + "epoch": 2.2224657534246575, + "grad_norm": 3.0969479084014893, + "learning_rate": 1.470349309504468e-06, + "loss": 1.5614, + "step": 2028 + }, + { + "epoch": 2.2235616438356165, + "grad_norm": 2.968904733657837, + "learning_rate": 1.4683184402924453e-06, + "loss": 1.5426, + "step": 2029 + }, + { + "epoch": 2.2246575342465755, + "grad_norm": 3.1130635738372803, + "learning_rate": 1.4662875710804226e-06, + "loss": 1.5695, + "step": 2030 + }, + { + "epoch": 2.2246575342465755, + "eval_loss": 1.8049863576889038, + "eval_runtime": 37.2207, + "eval_samples_per_second": 26.867, + "eval_steps_per_second": 1.128, + "step": 2030 + }, + { + "epoch": 2.225753424657534, + "grad_norm": 2.9921040534973145, + "learning_rate": 1.4642567018683998e-06, + "loss": 1.6212, + "step": 2031 + }, + { + "epoch": 2.226849315068493, + "grad_norm": 2.8416500091552734, + "learning_rate": 1.4622258326563771e-06, + "loss": 1.5199, + "step": 2032 + }, + { + "epoch": 2.227945205479452, + "grad_norm": 3.01324462890625, + "learning_rate": 1.4601949634443544e-06, + "loss": 1.5195, + "step": 2033 + }, + { + "epoch": 2.2290410958904108, + "grad_norm": 3.052415132522583, + "learning_rate": 1.4581640942323317e-06, + "loss": 1.5252, + "step": 2034 + }, + { + "epoch": 2.23013698630137, + "grad_norm": 3.011387348175049, + "learning_rate": 1.456133225020309e-06, + "loss": 1.6332, + "step": 2035 + }, + { + "epoch": 2.231232876712329, + "grad_norm": 3.0655579566955566, + "learning_rate": 1.4541023558082862e-06, + "loss": 1.5483, + "step": 2036 + }, + { + "epoch": 2.232328767123288, + "grad_norm": 2.988328695297241, + "learning_rate": 1.4520714865962635e-06, + "loss": 1.531, + "step": 2037 + }, + { + "epoch": 2.2334246575342465, + "grad_norm": 3.1820969581604004, + "learning_rate": 1.4500406173842406e-06, + "loss": 1.5699, + "step": 2038 + }, + { + "epoch": 2.2345205479452055, + "grad_norm": 3.100234270095825, + "learning_rate": 1.4480097481722178e-06, + "loss": 1.5344, + "step": 2039 + }, + { + "epoch": 2.2356164383561645, + "grad_norm": 3.056027889251709, + "learning_rate": 1.4459788789601951e-06, + "loss": 1.5451, + "step": 2040 + }, + { + "epoch": 2.2356164383561645, + "eval_loss": 1.8070225715637207, + "eval_runtime": 37.246, + "eval_samples_per_second": 26.849, + "eval_steps_per_second": 1.128, + "step": 2040 + }, + { + "epoch": 2.236712328767123, + "grad_norm": 2.988844633102417, + "learning_rate": 1.4439480097481724e-06, + "loss": 1.5011, + "step": 2041 + }, + { + "epoch": 2.237808219178082, + "grad_norm": 3.191542387008667, + "learning_rate": 1.4419171405361497e-06, + "loss": 1.4973, + "step": 2042 + }, + { + "epoch": 2.238904109589041, + "grad_norm": 3.1525838375091553, + "learning_rate": 1.439886271324127e-06, + "loss": 1.5488, + "step": 2043 + }, + { + "epoch": 2.24, + "grad_norm": 3.2659125328063965, + "learning_rate": 1.4378554021121042e-06, + "loss": 1.5416, + "step": 2044 + }, + { + "epoch": 2.241095890410959, + "grad_norm": 2.917837619781494, + "learning_rate": 1.4358245329000815e-06, + "loss": 1.5403, + "step": 2045 + }, + { + "epoch": 2.242191780821918, + "grad_norm": 3.0324769020080566, + "learning_rate": 1.4337936636880588e-06, + "loss": 1.5691, + "step": 2046 + }, + { + "epoch": 2.243287671232877, + "grad_norm": 3.0153560638427734, + "learning_rate": 1.431762794476036e-06, + "loss": 1.5501, + "step": 2047 + }, + { + "epoch": 2.2443835616438355, + "grad_norm": 2.926806926727295, + "learning_rate": 1.4297319252640133e-06, + "loss": 1.4968, + "step": 2048 + }, + { + "epoch": 2.2454794520547945, + "grad_norm": 3.12540864944458, + "learning_rate": 1.4277010560519906e-06, + "loss": 1.537, + "step": 2049 + }, + { + "epoch": 2.2465753424657535, + "grad_norm": 2.965358257293701, + "learning_rate": 1.4256701868399674e-06, + "loss": 1.5092, + "step": 2050 + }, + { + "epoch": 2.2465753424657535, + "eval_loss": 1.8053083419799805, + "eval_runtime": 37.2348, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 2050 + }, + { + "epoch": 2.247671232876712, + "grad_norm": 3.0532124042510986, + "learning_rate": 1.4236393176279447e-06, + "loss": 1.5817, + "step": 2051 + }, + { + "epoch": 2.248767123287671, + "grad_norm": 2.9400527477264404, + "learning_rate": 1.421608448415922e-06, + "loss": 1.5404, + "step": 2052 + }, + { + "epoch": 2.24986301369863, + "grad_norm": 3.175868272781372, + "learning_rate": 1.4195775792038993e-06, + "loss": 1.6286, + "step": 2053 + }, + { + "epoch": 2.250958904109589, + "grad_norm": 3.103508949279785, + "learning_rate": 1.4175467099918765e-06, + "loss": 1.5203, + "step": 2054 + }, + { + "epoch": 2.252054794520548, + "grad_norm": 2.868535041809082, + "learning_rate": 1.4155158407798538e-06, + "loss": 1.5479, + "step": 2055 + }, + { + "epoch": 2.253150684931507, + "grad_norm": 2.9070188999176025, + "learning_rate": 1.413484971567831e-06, + "loss": 1.5549, + "step": 2056 + }, + { + "epoch": 2.254246575342466, + "grad_norm": 3.1373631954193115, + "learning_rate": 1.4114541023558084e-06, + "loss": 1.4801, + "step": 2057 + }, + { + "epoch": 2.255342465753425, + "grad_norm": 2.9325647354125977, + "learning_rate": 1.4094232331437856e-06, + "loss": 1.5237, + "step": 2058 + }, + { + "epoch": 2.2564383561643835, + "grad_norm": 3.034519672393799, + "learning_rate": 1.407392363931763e-06, + "loss": 1.5659, + "step": 2059 + }, + { + "epoch": 2.2575342465753425, + "grad_norm": 3.0421204566955566, + "learning_rate": 1.4053614947197402e-06, + "loss": 1.6203, + "step": 2060 + }, + { + "epoch": 2.2575342465753425, + "eval_loss": 1.8065176010131836, + "eval_runtime": 37.2394, + "eval_samples_per_second": 26.853, + "eval_steps_per_second": 1.128, + "step": 2060 + }, + { + "epoch": 2.2586301369863016, + "grad_norm": 3.2316622734069824, + "learning_rate": 1.4033306255077175e-06, + "loss": 1.5379, + "step": 2061 + }, + { + "epoch": 2.25972602739726, + "grad_norm": 2.912513017654419, + "learning_rate": 1.4012997562956945e-06, + "loss": 1.4734, + "step": 2062 + }, + { + "epoch": 2.260821917808219, + "grad_norm": 3.00978422164917, + "learning_rate": 1.3992688870836718e-06, + "loss": 1.5865, + "step": 2063 + }, + { + "epoch": 2.261917808219178, + "grad_norm": 2.9873268604278564, + "learning_rate": 1.397238017871649e-06, + "loss": 1.5594, + "step": 2064 + }, + { + "epoch": 2.263013698630137, + "grad_norm": 3.122812509536743, + "learning_rate": 1.3952071486596264e-06, + "loss": 1.548, + "step": 2065 + }, + { + "epoch": 2.264109589041096, + "grad_norm": 3.104060411453247, + "learning_rate": 1.3931762794476036e-06, + "loss": 1.4993, + "step": 2066 + }, + { + "epoch": 2.265205479452055, + "grad_norm": 2.9612960815429688, + "learning_rate": 1.391145410235581e-06, + "loss": 1.5778, + "step": 2067 + }, + { + "epoch": 2.266301369863014, + "grad_norm": 3.1859002113342285, + "learning_rate": 1.3891145410235582e-06, + "loss": 1.5271, + "step": 2068 + }, + { + "epoch": 2.2673972602739725, + "grad_norm": 2.9132840633392334, + "learning_rate": 1.3870836718115355e-06, + "loss": 1.4565, + "step": 2069 + }, + { + "epoch": 2.2684931506849315, + "grad_norm": 2.930168867111206, + "learning_rate": 1.3850528025995127e-06, + "loss": 1.5279, + "step": 2070 + }, + { + "epoch": 2.2684931506849315, + "eval_loss": 1.8070082664489746, + "eval_runtime": 37.2365, + "eval_samples_per_second": 26.855, + "eval_steps_per_second": 1.128, + "step": 2070 + }, + { + "epoch": 2.2695890410958905, + "grad_norm": 3.0058515071868896, + "learning_rate": 1.38302193338749e-06, + "loss": 1.5264, + "step": 2071 + }, + { + "epoch": 2.270684931506849, + "grad_norm": 2.979008197784424, + "learning_rate": 1.3809910641754673e-06, + "loss": 1.4461, + "step": 2072 + }, + { + "epoch": 2.271780821917808, + "grad_norm": 3.0119616985321045, + "learning_rate": 1.3789601949634446e-06, + "loss": 1.553, + "step": 2073 + }, + { + "epoch": 2.272876712328767, + "grad_norm": 2.9711477756500244, + "learning_rate": 1.3769293257514218e-06, + "loss": 1.4921, + "step": 2074 + }, + { + "epoch": 2.2739726027397262, + "grad_norm": 2.935929536819458, + "learning_rate": 1.374898456539399e-06, + "loss": 1.5517, + "step": 2075 + }, + { + "epoch": 2.275068493150685, + "grad_norm": 2.9821221828460693, + "learning_rate": 1.3728675873273762e-06, + "loss": 1.4942, + "step": 2076 + }, + { + "epoch": 2.276164383561644, + "grad_norm": 2.9329755306243896, + "learning_rate": 1.3708367181153535e-06, + "loss": 1.5853, + "step": 2077 + }, + { + "epoch": 2.277260273972603, + "grad_norm": 3.084319829940796, + "learning_rate": 1.3688058489033307e-06, + "loss": 1.5481, + "step": 2078 + }, + { + "epoch": 2.2783561643835615, + "grad_norm": 3.2445032596588135, + "learning_rate": 1.366774979691308e-06, + "loss": 1.5568, + "step": 2079 + }, + { + "epoch": 2.2794520547945205, + "grad_norm": 3.0369651317596436, + "learning_rate": 1.3647441104792853e-06, + "loss": 1.5595, + "step": 2080 + }, + { + "epoch": 2.2794520547945205, + "eval_loss": 1.8065739870071411, + "eval_runtime": 37.2125, + "eval_samples_per_second": 26.873, + "eval_steps_per_second": 1.129, + "step": 2080 + }, + { + "epoch": 2.2805479452054795, + "grad_norm": 3.03692364692688, + "learning_rate": 1.3627132412672626e-06, + "loss": 1.5457, + "step": 2081 + }, + { + "epoch": 2.281643835616438, + "grad_norm": 2.951610803604126, + "learning_rate": 1.3606823720552398e-06, + "loss": 1.5026, + "step": 2082 + }, + { + "epoch": 2.282739726027397, + "grad_norm": 2.88623046875, + "learning_rate": 1.3586515028432171e-06, + "loss": 1.5404, + "step": 2083 + }, + { + "epoch": 2.283835616438356, + "grad_norm": 3.012883424758911, + "learning_rate": 1.3566206336311944e-06, + "loss": 1.4848, + "step": 2084 + }, + { + "epoch": 2.2849315068493152, + "grad_norm": 2.9949960708618164, + "learning_rate": 1.3545897644191717e-06, + "loss": 1.5366, + "step": 2085 + }, + { + "epoch": 2.286027397260274, + "grad_norm": 2.8936238288879395, + "learning_rate": 1.352558895207149e-06, + "loss": 1.5148, + "step": 2086 + }, + { + "epoch": 2.287123287671233, + "grad_norm": 2.967973232269287, + "learning_rate": 1.3505280259951258e-06, + "loss": 1.5742, + "step": 2087 + }, + { + "epoch": 2.288219178082192, + "grad_norm": 3.1881325244903564, + "learning_rate": 1.348497156783103e-06, + "loss": 1.5391, + "step": 2088 + }, + { + "epoch": 2.289315068493151, + "grad_norm": 2.9883408546447754, + "learning_rate": 1.3464662875710804e-06, + "loss": 1.5666, + "step": 2089 + }, + { + "epoch": 2.2904109589041095, + "grad_norm": 2.827197313308716, + "learning_rate": 1.3444354183590576e-06, + "loss": 1.4941, + "step": 2090 + }, + { + "epoch": 2.2904109589041095, + "eval_loss": 1.8056135177612305, + "eval_runtime": 37.227, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 2090 + }, + { + "epoch": 2.2915068493150685, + "grad_norm": 3.132901906967163, + "learning_rate": 1.342404549147035e-06, + "loss": 1.5699, + "step": 2091 + }, + { + "epoch": 2.2926027397260276, + "grad_norm": 3.0904805660247803, + "learning_rate": 1.3403736799350122e-06, + "loss": 1.5173, + "step": 2092 + }, + { + "epoch": 2.293698630136986, + "grad_norm": 3.1590170860290527, + "learning_rate": 1.3383428107229895e-06, + "loss": 1.5546, + "step": 2093 + }, + { + "epoch": 2.294794520547945, + "grad_norm": 3.110947370529175, + "learning_rate": 1.3363119415109667e-06, + "loss": 1.5677, + "step": 2094 + }, + { + "epoch": 2.2958904109589042, + "grad_norm": 2.9345972537994385, + "learning_rate": 1.334281072298944e-06, + "loss": 1.5682, + "step": 2095 + }, + { + "epoch": 2.296986301369863, + "grad_norm": 2.9347832202911377, + "learning_rate": 1.3322502030869213e-06, + "loss": 1.5463, + "step": 2096 + }, + { + "epoch": 2.298082191780822, + "grad_norm": 2.859834671020508, + "learning_rate": 1.3302193338748986e-06, + "loss": 1.5414, + "step": 2097 + }, + { + "epoch": 2.299178082191781, + "grad_norm": 2.980107545852661, + "learning_rate": 1.3281884646628758e-06, + "loss": 1.5704, + "step": 2098 + }, + { + "epoch": 2.30027397260274, + "grad_norm": 3.131226062774658, + "learning_rate": 1.3261575954508531e-06, + "loss": 1.5642, + "step": 2099 + }, + { + "epoch": 2.3013698630136985, + "grad_norm": 3.0369224548339844, + "learning_rate": 1.3241267262388302e-06, + "loss": 1.5461, + "step": 2100 + }, + { + "epoch": 2.3013698630136985, + "eval_loss": 1.8044562339782715, + "eval_runtime": 37.2273, + "eval_samples_per_second": 26.862, + "eval_steps_per_second": 1.128, + "step": 2100 + }, + { + "epoch": 2.3024657534246575, + "grad_norm": 2.9488580226898193, + "learning_rate": 1.3220958570268074e-06, + "loss": 1.5401, + "step": 2101 + }, + { + "epoch": 2.3035616438356166, + "grad_norm": 2.983025312423706, + "learning_rate": 1.3200649878147847e-06, + "loss": 1.4761, + "step": 2102 + }, + { + "epoch": 2.304657534246575, + "grad_norm": 3.0856857299804688, + "learning_rate": 1.318034118602762e-06, + "loss": 1.5183, + "step": 2103 + }, + { + "epoch": 2.305753424657534, + "grad_norm": 3.0773403644561768, + "learning_rate": 1.3160032493907393e-06, + "loss": 1.5612, + "step": 2104 + }, + { + "epoch": 2.3068493150684932, + "grad_norm": 3.0656943321228027, + "learning_rate": 1.3139723801787166e-06, + "loss": 1.5646, + "step": 2105 + }, + { + "epoch": 2.3079452054794523, + "grad_norm": 3.225187301635742, + "learning_rate": 1.3119415109666938e-06, + "loss": 1.5778, + "step": 2106 + }, + { + "epoch": 2.309041095890411, + "grad_norm": 3.0056018829345703, + "learning_rate": 1.309910641754671e-06, + "loss": 1.5346, + "step": 2107 + }, + { + "epoch": 2.31013698630137, + "grad_norm": 2.9136228561401367, + "learning_rate": 1.3078797725426484e-06, + "loss": 1.4936, + "step": 2108 + }, + { + "epoch": 2.311232876712329, + "grad_norm": 2.873004913330078, + "learning_rate": 1.3058489033306257e-06, + "loss": 1.5123, + "step": 2109 + }, + { + "epoch": 2.3123287671232875, + "grad_norm": 2.9853620529174805, + "learning_rate": 1.303818034118603e-06, + "loss": 1.5149, + "step": 2110 + }, + { + "epoch": 2.3123287671232875, + "eval_loss": 1.803491473197937, + "eval_runtime": 36.8542, + "eval_samples_per_second": 27.134, + "eval_steps_per_second": 1.14, + "step": 2110 + }, + { + "epoch": 2.3134246575342465, + "grad_norm": 2.9889047145843506, + "learning_rate": 1.3017871649065802e-06, + "loss": 1.5294, + "step": 2111 + }, + { + "epoch": 2.3145205479452056, + "grad_norm": 2.9619243144989014, + "learning_rate": 1.2997562956945573e-06, + "loss": 1.5282, + "step": 2112 + }, + { + "epoch": 2.315616438356164, + "grad_norm": 2.9584786891937256, + "learning_rate": 1.2977254264825345e-06, + "loss": 1.4972, + "step": 2113 + }, + { + "epoch": 2.316712328767123, + "grad_norm": 2.9817779064178467, + "learning_rate": 1.2956945572705118e-06, + "loss": 1.5167, + "step": 2114 + }, + { + "epoch": 2.317808219178082, + "grad_norm": 3.018979072570801, + "learning_rate": 1.293663688058489e-06, + "loss": 1.5381, + "step": 2115 + }, + { + "epoch": 2.3189041095890413, + "grad_norm": 2.881077289581299, + "learning_rate": 1.2916328188464664e-06, + "loss": 1.5077, + "step": 2116 + }, + { + "epoch": 2.32, + "grad_norm": 2.9734835624694824, + "learning_rate": 1.2896019496344436e-06, + "loss": 1.4644, + "step": 2117 + }, + { + "epoch": 2.321095890410959, + "grad_norm": 3.111104726791382, + "learning_rate": 1.287571080422421e-06, + "loss": 1.5056, + "step": 2118 + }, + { + "epoch": 2.322191780821918, + "grad_norm": 3.1056535243988037, + "learning_rate": 1.2855402112103982e-06, + "loss": 1.5665, + "step": 2119 + }, + { + "epoch": 2.323287671232877, + "grad_norm": 3.0189976692199707, + "learning_rate": 1.2835093419983755e-06, + "loss": 1.4849, + "step": 2120 + }, + { + "epoch": 2.323287671232877, + "eval_loss": 1.8044172525405884, + "eval_runtime": 37.0338, + "eval_samples_per_second": 27.002, + "eval_steps_per_second": 1.134, + "step": 2120 + }, + { + "epoch": 2.3243835616438355, + "grad_norm": 3.037372589111328, + "learning_rate": 1.2814784727863528e-06, + "loss": 1.4979, + "step": 2121 + }, + { + "epoch": 2.3254794520547946, + "grad_norm": 3.011446952819824, + "learning_rate": 1.27944760357433e-06, + "loss": 1.5108, + "step": 2122 + }, + { + "epoch": 2.3265753424657536, + "grad_norm": 3.189188003540039, + "learning_rate": 1.2774167343623073e-06, + "loss": 1.5814, + "step": 2123 + }, + { + "epoch": 2.327671232876712, + "grad_norm": 3.1129472255706787, + "learning_rate": 1.2753858651502846e-06, + "loss": 1.5634, + "step": 2124 + }, + { + "epoch": 2.328767123287671, + "grad_norm": 2.98749041557312, + "learning_rate": 1.2733549959382616e-06, + "loss": 1.5051, + "step": 2125 + }, + { + "epoch": 2.3298630136986302, + "grad_norm": 3.0713346004486084, + "learning_rate": 1.271324126726239e-06, + "loss": 1.5239, + "step": 2126 + }, + { + "epoch": 2.330958904109589, + "grad_norm": 3.0081963539123535, + "learning_rate": 1.2692932575142162e-06, + "loss": 1.5469, + "step": 2127 + }, + { + "epoch": 2.332054794520548, + "grad_norm": 2.9524900913238525, + "learning_rate": 1.2672623883021935e-06, + "loss": 1.4701, + "step": 2128 + }, + { + "epoch": 2.333150684931507, + "grad_norm": 2.992372751235962, + "learning_rate": 1.2652315190901707e-06, + "loss": 1.4758, + "step": 2129 + }, + { + "epoch": 2.334246575342466, + "grad_norm": 3.2761764526367188, + "learning_rate": 1.263200649878148e-06, + "loss": 1.5879, + "step": 2130 + }, + { + "epoch": 2.334246575342466, + "eval_loss": 1.8030585050582886, + "eval_runtime": 37.1052, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 2130 + }, + { + "epoch": 2.3353424657534245, + "grad_norm": 3.3670222759246826, + "learning_rate": 1.2611697806661253e-06, + "loss": 1.5279, + "step": 2131 + }, + { + "epoch": 2.3364383561643836, + "grad_norm": 3.045868158340454, + "learning_rate": 1.2591389114541026e-06, + "loss": 1.5572, + "step": 2132 + }, + { + "epoch": 2.3375342465753426, + "grad_norm": 2.9535293579101562, + "learning_rate": 1.2571080422420798e-06, + "loss": 1.5463, + "step": 2133 + }, + { + "epoch": 2.338630136986301, + "grad_norm": 2.9338130950927734, + "learning_rate": 1.2550771730300571e-06, + "loss": 1.5026, + "step": 2134 + }, + { + "epoch": 2.33972602739726, + "grad_norm": 3.0596487522125244, + "learning_rate": 1.2530463038180344e-06, + "loss": 1.584, + "step": 2135 + }, + { + "epoch": 2.3408219178082192, + "grad_norm": 2.9395554065704346, + "learning_rate": 1.2510154346060117e-06, + "loss": 1.5385, + "step": 2136 + }, + { + "epoch": 2.3419178082191783, + "grad_norm": 2.8701987266540527, + "learning_rate": 1.2489845653939887e-06, + "loss": 1.5223, + "step": 2137 + }, + { + "epoch": 2.343013698630137, + "grad_norm": 3.2599759101867676, + "learning_rate": 1.246953696181966e-06, + "loss": 1.5533, + "step": 2138 + }, + { + "epoch": 2.344109589041096, + "grad_norm": 2.990359306335449, + "learning_rate": 1.2449228269699433e-06, + "loss": 1.6067, + "step": 2139 + }, + { + "epoch": 2.345205479452055, + "grad_norm": 3.009202241897583, + "learning_rate": 1.2428919577579204e-06, + "loss": 1.5358, + "step": 2140 + }, + { + "epoch": 2.345205479452055, + "eval_loss": 1.80193293094635, + "eval_runtime": 37.189, + "eval_samples_per_second": 26.89, + "eval_steps_per_second": 1.129, + "step": 2140 + }, + { + "epoch": 2.3463013698630135, + "grad_norm": 3.050416946411133, + "learning_rate": 1.2408610885458976e-06, + "loss": 1.4996, + "step": 2141 + }, + { + "epoch": 2.3473972602739726, + "grad_norm": 2.9148621559143066, + "learning_rate": 1.238830219333875e-06, + "loss": 1.5329, + "step": 2142 + }, + { + "epoch": 2.3484931506849316, + "grad_norm": 3.0416409969329834, + "learning_rate": 1.2367993501218522e-06, + "loss": 1.4983, + "step": 2143 + }, + { + "epoch": 2.34958904109589, + "grad_norm": 3.387521266937256, + "learning_rate": 1.2347684809098295e-06, + "loss": 1.5696, + "step": 2144 + }, + { + "epoch": 2.350684931506849, + "grad_norm": 3.084455728530884, + "learning_rate": 1.2327376116978067e-06, + "loss": 1.5567, + "step": 2145 + }, + { + "epoch": 2.3517808219178082, + "grad_norm": 2.912816047668457, + "learning_rate": 1.230706742485784e-06, + "loss": 1.4776, + "step": 2146 + }, + { + "epoch": 2.3528767123287673, + "grad_norm": 3.0122692584991455, + "learning_rate": 1.2286758732737613e-06, + "loss": 1.4856, + "step": 2147 + }, + { + "epoch": 2.353972602739726, + "grad_norm": 3.1856625080108643, + "learning_rate": 1.2266450040617386e-06, + "loss": 1.6215, + "step": 2148 + }, + { + "epoch": 2.355068493150685, + "grad_norm": 3.2322850227355957, + "learning_rate": 1.2246141348497158e-06, + "loss": 1.5989, + "step": 2149 + }, + { + "epoch": 2.356164383561644, + "grad_norm": 2.9880452156066895, + "learning_rate": 1.2225832656376931e-06, + "loss": 1.5332, + "step": 2150 + }, + { + "epoch": 2.356164383561644, + "eval_loss": 1.8039098978042603, + "eval_runtime": 37.2352, + "eval_samples_per_second": 26.856, + "eval_steps_per_second": 1.128, + "step": 2150 + }, + { + "epoch": 2.357260273972603, + "grad_norm": 2.957336902618408, + "learning_rate": 1.2205523964256704e-06, + "loss": 1.548, + "step": 2151 + }, + { + "epoch": 2.3583561643835615, + "grad_norm": 3.0989115238189697, + "learning_rate": 1.2185215272136477e-06, + "loss": 1.5184, + "step": 2152 + }, + { + "epoch": 2.3594520547945206, + "grad_norm": 2.8828625679016113, + "learning_rate": 1.2164906580016247e-06, + "loss": 1.548, + "step": 2153 + }, + { + "epoch": 2.3605479452054796, + "grad_norm": 3.0119025707244873, + "learning_rate": 1.214459788789602e-06, + "loss": 1.5504, + "step": 2154 + }, + { + "epoch": 2.361643835616438, + "grad_norm": 2.974620819091797, + "learning_rate": 1.2124289195775793e-06, + "loss": 1.4489, + "step": 2155 + }, + { + "epoch": 2.3627397260273972, + "grad_norm": 3.022380828857422, + "learning_rate": 1.2103980503655566e-06, + "loss": 1.4973, + "step": 2156 + }, + { + "epoch": 2.3638356164383563, + "grad_norm": 3.0495481491088867, + "learning_rate": 1.2083671811535338e-06, + "loss": 1.5131, + "step": 2157 + }, + { + "epoch": 2.364931506849315, + "grad_norm": 3.146232843399048, + "learning_rate": 1.2063363119415111e-06, + "loss": 1.5287, + "step": 2158 + }, + { + "epoch": 2.366027397260274, + "grad_norm": 2.92659592628479, + "learning_rate": 1.2043054427294882e-06, + "loss": 1.5128, + "step": 2159 + }, + { + "epoch": 2.367123287671233, + "grad_norm": 3.1443896293640137, + "learning_rate": 1.2022745735174654e-06, + "loss": 1.5145, + "step": 2160 + }, + { + "epoch": 2.367123287671233, + "eval_loss": 1.8029189109802246, + "eval_runtime": 37.2059, + "eval_samples_per_second": 26.877, + "eval_steps_per_second": 1.129, + "step": 2160 + }, + { + "epoch": 2.368219178082192, + "grad_norm": 3.0999011993408203, + "learning_rate": 1.2002437043054427e-06, + "loss": 1.5418, + "step": 2161 + }, + { + "epoch": 2.3693150684931505, + "grad_norm": 3.05961012840271, + "learning_rate": 1.19821283509342e-06, + "loss": 1.5809, + "step": 2162 + }, + { + "epoch": 2.3704109589041096, + "grad_norm": 3.056363582611084, + "learning_rate": 1.1961819658813973e-06, + "loss": 1.5744, + "step": 2163 + }, + { + "epoch": 2.3715068493150686, + "grad_norm": 3.012352705001831, + "learning_rate": 1.1941510966693746e-06, + "loss": 1.5419, + "step": 2164 + }, + { + "epoch": 2.3726027397260276, + "grad_norm": 3.0131568908691406, + "learning_rate": 1.1921202274573518e-06, + "loss": 1.5101, + "step": 2165 + }, + { + "epoch": 2.3736986301369862, + "grad_norm": 2.9452834129333496, + "learning_rate": 1.190089358245329e-06, + "loss": 1.5324, + "step": 2166 + }, + { + "epoch": 2.3747945205479453, + "grad_norm": 3.2359182834625244, + "learning_rate": 1.1880584890333064e-06, + "loss": 1.5823, + "step": 2167 + }, + { + "epoch": 2.3758904109589043, + "grad_norm": 3.136212110519409, + "learning_rate": 1.1860276198212837e-06, + "loss": 1.5105, + "step": 2168 + }, + { + "epoch": 2.376986301369863, + "grad_norm": 3.0032570362091064, + "learning_rate": 1.183996750609261e-06, + "loss": 1.5518, + "step": 2169 + }, + { + "epoch": 2.378082191780822, + "grad_norm": 3.157111644744873, + "learning_rate": 1.1819658813972382e-06, + "loss": 1.5394, + "step": 2170 + }, + { + "epoch": 2.378082191780822, + "eval_loss": 1.8035898208618164, + "eval_runtime": 37.2489, + "eval_samples_per_second": 26.846, + "eval_steps_per_second": 1.128, + "step": 2170 + }, + { + "epoch": 2.379178082191781, + "grad_norm": 2.9560418128967285, + "learning_rate": 1.1799350121852153e-06, + "loss": 1.5269, + "step": 2171 + }, + { + "epoch": 2.3802739726027395, + "grad_norm": 3.0187556743621826, + "learning_rate": 1.1779041429731925e-06, + "loss": 1.5683, + "step": 2172 + }, + { + "epoch": 2.3813698630136986, + "grad_norm": 3.0405445098876953, + "learning_rate": 1.1758732737611698e-06, + "loss": 1.5676, + "step": 2173 + }, + { + "epoch": 2.3824657534246576, + "grad_norm": 3.1029672622680664, + "learning_rate": 1.173842404549147e-06, + "loss": 1.487, + "step": 2174 + }, + { + "epoch": 2.383561643835616, + "grad_norm": 3.0403008460998535, + "learning_rate": 1.1718115353371244e-06, + "loss": 1.5848, + "step": 2175 + }, + { + "epoch": 2.3846575342465752, + "grad_norm": 2.9963865280151367, + "learning_rate": 1.1697806661251016e-06, + "loss": 1.5152, + "step": 2176 + }, + { + "epoch": 2.3857534246575343, + "grad_norm": 3.085374116897583, + "learning_rate": 1.167749796913079e-06, + "loss": 1.5846, + "step": 2177 + }, + { + "epoch": 2.3868493150684933, + "grad_norm": 3.160015821456909, + "learning_rate": 1.1657189277010562e-06, + "loss": 1.5619, + "step": 2178 + }, + { + "epoch": 2.387945205479452, + "grad_norm": 3.0886569023132324, + "learning_rate": 1.1636880584890335e-06, + "loss": 1.5453, + "step": 2179 + }, + { + "epoch": 2.389041095890411, + "grad_norm": 3.148318290710449, + "learning_rate": 1.1616571892770108e-06, + "loss": 1.5343, + "step": 2180 + }, + { + "epoch": 2.389041095890411, + "eval_loss": 1.801662564277649, + "eval_runtime": 37.2754, + "eval_samples_per_second": 26.827, + "eval_steps_per_second": 1.127, + "step": 2180 + }, + { + "epoch": 2.39013698630137, + "grad_norm": 2.9711012840270996, + "learning_rate": 1.159626320064988e-06, + "loss": 1.5594, + "step": 2181 + }, + { + "epoch": 2.391232876712329, + "grad_norm": 3.1002538204193115, + "learning_rate": 1.1575954508529653e-06, + "loss": 1.5396, + "step": 2182 + }, + { + "epoch": 2.3923287671232876, + "grad_norm": 3.06974458694458, + "learning_rate": 1.1555645816409426e-06, + "loss": 1.5554, + "step": 2183 + }, + { + "epoch": 2.3934246575342466, + "grad_norm": 2.862666606903076, + "learning_rate": 1.1535337124289196e-06, + "loss": 1.5543, + "step": 2184 + }, + { + "epoch": 2.3945205479452056, + "grad_norm": 3.0684075355529785, + "learning_rate": 1.151502843216897e-06, + "loss": 1.5086, + "step": 2185 + }, + { + "epoch": 2.3956164383561642, + "grad_norm": 2.930772066116333, + "learning_rate": 1.1494719740048742e-06, + "loss": 1.4769, + "step": 2186 + }, + { + "epoch": 2.3967123287671233, + "grad_norm": 3.1012887954711914, + "learning_rate": 1.1474411047928515e-06, + "loss": 1.5583, + "step": 2187 + }, + { + "epoch": 2.3978082191780823, + "grad_norm": 2.835838556289673, + "learning_rate": 1.1454102355808287e-06, + "loss": 1.5325, + "step": 2188 + }, + { + "epoch": 2.398904109589041, + "grad_norm": 3.1359124183654785, + "learning_rate": 1.143379366368806e-06, + "loss": 1.506, + "step": 2189 + }, + { + "epoch": 2.4, + "grad_norm": 2.848830223083496, + "learning_rate": 1.141348497156783e-06, + "loss": 1.4973, + "step": 2190 + }, + { + "epoch": 2.4, + "eval_loss": 1.803225040435791, + "eval_runtime": 37.2288, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 2190 + }, + { + "epoch": 2.401095890410959, + "grad_norm": 3.146909475326538, + "learning_rate": 1.1393176279447604e-06, + "loss": 1.6196, + "step": 2191 + }, + { + "epoch": 2.402191780821918, + "grad_norm": 3.1211538314819336, + "learning_rate": 1.1372867587327376e-06, + "loss": 1.5903, + "step": 2192 + }, + { + "epoch": 2.4032876712328766, + "grad_norm": 2.8264310359954834, + "learning_rate": 1.135255889520715e-06, + "loss": 1.4559, + "step": 2193 + }, + { + "epoch": 2.4043835616438356, + "grad_norm": 3.1073036193847656, + "learning_rate": 1.1332250203086922e-06, + "loss": 1.5535, + "step": 2194 + }, + { + "epoch": 2.4054794520547946, + "grad_norm": 2.904238700866699, + "learning_rate": 1.1311941510966695e-06, + "loss": 1.5432, + "step": 2195 + }, + { + "epoch": 2.4065753424657537, + "grad_norm": 3.2940893173217773, + "learning_rate": 1.1291632818846467e-06, + "loss": 1.5287, + "step": 2196 + }, + { + "epoch": 2.4076712328767123, + "grad_norm": 3.031548500061035, + "learning_rate": 1.127132412672624e-06, + "loss": 1.4965, + "step": 2197 + }, + { + "epoch": 2.4087671232876713, + "grad_norm": 2.885241746902466, + "learning_rate": 1.1251015434606013e-06, + "loss": 1.5078, + "step": 2198 + }, + { + "epoch": 2.4098630136986303, + "grad_norm": 3.005876064300537, + "learning_rate": 1.1230706742485786e-06, + "loss": 1.5526, + "step": 2199 + }, + { + "epoch": 2.410958904109589, + "grad_norm": 3.000643253326416, + "learning_rate": 1.1210398050365558e-06, + "loss": 1.574, + "step": 2200 + }, + { + "epoch": 2.410958904109589, + "eval_loss": 1.801488995552063, + "eval_runtime": 37.2077, + "eval_samples_per_second": 26.876, + "eval_steps_per_second": 1.129, + "step": 2200 + }, + { + "epoch": 2.412054794520548, + "grad_norm": 2.9339702129364014, + "learning_rate": 1.1190089358245331e-06, + "loss": 1.502, + "step": 2201 + }, + { + "epoch": 2.413150684931507, + "grad_norm": 2.943730592727661, + "learning_rate": 1.1169780666125102e-06, + "loss": 1.4887, + "step": 2202 + }, + { + "epoch": 2.4142465753424656, + "grad_norm": 3.039198637008667, + "learning_rate": 1.1149471974004875e-06, + "loss": 1.5452, + "step": 2203 + }, + { + "epoch": 2.4153424657534246, + "grad_norm": 3.050334930419922, + "learning_rate": 1.1129163281884647e-06, + "loss": 1.4998, + "step": 2204 + }, + { + "epoch": 2.4164383561643836, + "grad_norm": 3.1679272651672363, + "learning_rate": 1.110885458976442e-06, + "loss": 1.59, + "step": 2205 + }, + { + "epoch": 2.417534246575342, + "grad_norm": 2.9258804321289062, + "learning_rate": 1.1088545897644193e-06, + "loss": 1.4986, + "step": 2206 + }, + { + "epoch": 2.4186301369863012, + "grad_norm": 3.094996213912964, + "learning_rate": 1.1068237205523966e-06, + "loss": 1.5702, + "step": 2207 + }, + { + "epoch": 2.4197260273972603, + "grad_norm": 3.027369260787964, + "learning_rate": 1.1047928513403736e-06, + "loss": 1.5621, + "step": 2208 + }, + { + "epoch": 2.4208219178082193, + "grad_norm": 3.138309955596924, + "learning_rate": 1.102761982128351e-06, + "loss": 1.5515, + "step": 2209 + }, + { + "epoch": 2.421917808219178, + "grad_norm": 2.9625754356384277, + "learning_rate": 1.1007311129163282e-06, + "loss": 1.4852, + "step": 2210 + }, + { + "epoch": 2.421917808219178, + "eval_loss": 1.8001623153686523, + "eval_runtime": 36.8373, + "eval_samples_per_second": 27.146, + "eval_steps_per_second": 1.14, + "step": 2210 + }, + { + "epoch": 2.423013698630137, + "grad_norm": 3.0380866527557373, + "learning_rate": 1.0987002437043055e-06, + "loss": 1.5184, + "step": 2211 + }, + { + "epoch": 2.424109589041096, + "grad_norm": 3.355539321899414, + "learning_rate": 1.0966693744922827e-06, + "loss": 1.5231, + "step": 2212 + }, + { + "epoch": 2.425205479452055, + "grad_norm": 2.871112108230591, + "learning_rate": 1.09463850528026e-06, + "loss": 1.4767, + "step": 2213 + }, + { + "epoch": 2.4263013698630136, + "grad_norm": 3.020303964614868, + "learning_rate": 1.0926076360682373e-06, + "loss": 1.509, + "step": 2214 + }, + { + "epoch": 2.4273972602739726, + "grad_norm": 3.1246845722198486, + "learning_rate": 1.0905767668562146e-06, + "loss": 1.5051, + "step": 2215 + }, + { + "epoch": 2.4284931506849317, + "grad_norm": 2.9896199703216553, + "learning_rate": 1.0885458976441918e-06, + "loss": 1.5094, + "step": 2216 + }, + { + "epoch": 2.4295890410958902, + "grad_norm": 3.0782384872436523, + "learning_rate": 1.0865150284321691e-06, + "loss": 1.6112, + "step": 2217 + }, + { + "epoch": 2.4306849315068493, + "grad_norm": 2.959040880203247, + "learning_rate": 1.0844841592201464e-06, + "loss": 1.5529, + "step": 2218 + }, + { + "epoch": 2.4317808219178083, + "grad_norm": 2.9968371391296387, + "learning_rate": 1.0824532900081237e-06, + "loss": 1.5692, + "step": 2219 + }, + { + "epoch": 2.432876712328767, + "grad_norm": 2.980635404586792, + "learning_rate": 1.080422420796101e-06, + "loss": 1.5445, + "step": 2220 + }, + { + "epoch": 2.432876712328767, + "eval_loss": 1.8024731874465942, + "eval_runtime": 37.0848, + "eval_samples_per_second": 26.965, + "eval_steps_per_second": 1.133, + "step": 2220 + }, + { + "epoch": 2.433972602739726, + "grad_norm": 3.1620373725891113, + "learning_rate": 1.078391551584078e-06, + "loss": 1.5357, + "step": 2221 + }, + { + "epoch": 2.435068493150685, + "grad_norm": 3.0638134479522705, + "learning_rate": 1.0763606823720553e-06, + "loss": 1.5717, + "step": 2222 + }, + { + "epoch": 2.436164383561644, + "grad_norm": 3.0402369499206543, + "learning_rate": 1.0743298131600326e-06, + "loss": 1.6096, + "step": 2223 + }, + { + "epoch": 2.4372602739726026, + "grad_norm": 3.1715261936187744, + "learning_rate": 1.0722989439480098e-06, + "loss": 1.5642, + "step": 2224 + }, + { + "epoch": 2.4383561643835616, + "grad_norm": 3.181551218032837, + "learning_rate": 1.070268074735987e-06, + "loss": 1.5177, + "step": 2225 + }, + { + "epoch": 2.4394520547945207, + "grad_norm": 3.1069509983062744, + "learning_rate": 1.0682372055239644e-06, + "loss": 1.5162, + "step": 2226 + }, + { + "epoch": 2.4405479452054797, + "grad_norm": 2.949441432952881, + "learning_rate": 1.0662063363119414e-06, + "loss": 1.5187, + "step": 2227 + }, + { + "epoch": 2.4416438356164383, + "grad_norm": 3.099307060241699, + "learning_rate": 1.0641754670999187e-06, + "loss": 1.5095, + "step": 2228 + }, + { + "epoch": 2.4427397260273973, + "grad_norm": 2.895397186279297, + "learning_rate": 1.062144597887896e-06, + "loss": 1.5468, + "step": 2229 + }, + { + "epoch": 2.4438356164383563, + "grad_norm": 3.1592397689819336, + "learning_rate": 1.0601137286758733e-06, + "loss": 1.5393, + "step": 2230 + }, + { + "epoch": 2.4438356164383563, + "eval_loss": 1.8015625476837158, + "eval_runtime": 37.166, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 2230 + }, + { + "epoch": 2.444931506849315, + "grad_norm": 3.1496217250823975, + "learning_rate": 1.0580828594638505e-06, + "loss": 1.5844, + "step": 2231 + }, + { + "epoch": 2.446027397260274, + "grad_norm": 2.8571274280548096, + "learning_rate": 1.0560519902518278e-06, + "loss": 1.466, + "step": 2232 + }, + { + "epoch": 2.447123287671233, + "grad_norm": 3.0587098598480225, + "learning_rate": 1.054021121039805e-06, + "loss": 1.5529, + "step": 2233 + }, + { + "epoch": 2.4482191780821916, + "grad_norm": 2.857961654663086, + "learning_rate": 1.0519902518277824e-06, + "loss": 1.4644, + "step": 2234 + }, + { + "epoch": 2.4493150684931506, + "grad_norm": 3.008579969406128, + "learning_rate": 1.0499593826157596e-06, + "loss": 1.4671, + "step": 2235 + }, + { + "epoch": 2.4504109589041096, + "grad_norm": 3.0578463077545166, + "learning_rate": 1.047928513403737e-06, + "loss": 1.5448, + "step": 2236 + }, + { + "epoch": 2.4515068493150687, + "grad_norm": 3.058122158050537, + "learning_rate": 1.0458976441917142e-06, + "loss": 1.529, + "step": 2237 + }, + { + "epoch": 2.4526027397260273, + "grad_norm": 2.922083854675293, + "learning_rate": 1.0438667749796915e-06, + "loss": 1.4541, + "step": 2238 + }, + { + "epoch": 2.4536986301369863, + "grad_norm": 2.946068525314331, + "learning_rate": 1.0418359057676688e-06, + "loss": 1.4616, + "step": 2239 + }, + { + "epoch": 2.4547945205479453, + "grad_norm": 2.8997962474823, + "learning_rate": 1.0398050365556458e-06, + "loss": 1.5101, + "step": 2240 + }, + { + "epoch": 2.4547945205479453, + "eval_loss": 1.8022980690002441, + "eval_runtime": 37.1978, + "eval_samples_per_second": 26.883, + "eval_steps_per_second": 1.129, + "step": 2240 + }, + { + "epoch": 2.455890410958904, + "grad_norm": 3.0214385986328125, + "learning_rate": 1.037774167343623e-06, + "loss": 1.5636, + "step": 2241 + }, + { + "epoch": 2.456986301369863, + "grad_norm": 2.946601152420044, + "learning_rate": 1.0357432981316004e-06, + "loss": 1.5143, + "step": 2242 + }, + { + "epoch": 2.458082191780822, + "grad_norm": 3.0568253993988037, + "learning_rate": 1.0337124289195776e-06, + "loss": 1.5399, + "step": 2243 + }, + { + "epoch": 2.459178082191781, + "grad_norm": 3.1164939403533936, + "learning_rate": 1.031681559707555e-06, + "loss": 1.5628, + "step": 2244 + }, + { + "epoch": 2.4602739726027396, + "grad_norm": 3.0754446983337402, + "learning_rate": 1.0296506904955322e-06, + "loss": 1.5634, + "step": 2245 + }, + { + "epoch": 2.4613698630136986, + "grad_norm": 3.028993606567383, + "learning_rate": 1.0276198212835095e-06, + "loss": 1.5104, + "step": 2246 + }, + { + "epoch": 2.4624657534246577, + "grad_norm": 3.14074444770813, + "learning_rate": 1.0255889520714867e-06, + "loss": 1.5277, + "step": 2247 + }, + { + "epoch": 2.4635616438356163, + "grad_norm": 3.0798747539520264, + "learning_rate": 1.023558082859464e-06, + "loss": 1.5152, + "step": 2248 + }, + { + "epoch": 2.4646575342465753, + "grad_norm": 2.967341899871826, + "learning_rate": 1.0215272136474413e-06, + "loss": 1.4692, + "step": 2249 + }, + { + "epoch": 2.4657534246575343, + "grad_norm": 3.1009557247161865, + "learning_rate": 1.0194963444354186e-06, + "loss": 1.5714, + "step": 2250 + }, + { + "epoch": 2.4657534246575343, + "eval_loss": 1.801601529121399, + "eval_runtime": 37.2124, + "eval_samples_per_second": 26.873, + "eval_steps_per_second": 1.129, + "step": 2250 + }, + { + "epoch": 2.466849315068493, + "grad_norm": 3.005615472793579, + "learning_rate": 1.0174654752233958e-06, + "loss": 1.4728, + "step": 2251 + }, + { + "epoch": 2.467945205479452, + "grad_norm": 3.348233699798584, + "learning_rate": 1.015434606011373e-06, + "loss": 1.5494, + "step": 2252 + }, + { + "epoch": 2.469041095890411, + "grad_norm": 3.094768762588501, + "learning_rate": 1.0134037367993502e-06, + "loss": 1.5157, + "step": 2253 + }, + { + "epoch": 2.47013698630137, + "grad_norm": 2.989004373550415, + "learning_rate": 1.0113728675873275e-06, + "loss": 1.5347, + "step": 2254 + }, + { + "epoch": 2.4712328767123286, + "grad_norm": 3.0253705978393555, + "learning_rate": 1.0093419983753047e-06, + "loss": 1.5208, + "step": 2255 + }, + { + "epoch": 2.4723287671232876, + "grad_norm": 3.2195019721984863, + "learning_rate": 1.007311129163282e-06, + "loss": 1.5438, + "step": 2256 + }, + { + "epoch": 2.4734246575342467, + "grad_norm": 2.8774361610412598, + "learning_rate": 1.0052802599512593e-06, + "loss": 1.4805, + "step": 2257 + }, + { + "epoch": 2.4745205479452057, + "grad_norm": 2.9766504764556885, + "learning_rate": 1.0032493907392364e-06, + "loss": 1.4953, + "step": 2258 + }, + { + "epoch": 2.4756164383561643, + "grad_norm": 3.0610811710357666, + "learning_rate": 1.0012185215272136e-06, + "loss": 1.531, + "step": 2259 + }, + { + "epoch": 2.4767123287671233, + "grad_norm": 2.8687760829925537, + "learning_rate": 9.99187652315191e-07, + "loss": 1.4797, + "step": 2260 + }, + { + "epoch": 2.4767123287671233, + "eval_loss": 1.800689697265625, + "eval_runtime": 37.2172, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.129, + "step": 2260 + }, + { + "epoch": 2.4778082191780824, + "grad_norm": 3.1072497367858887, + "learning_rate": 9.971567831031682e-07, + "loss": 1.577, + "step": 2261 + }, + { + "epoch": 2.478904109589041, + "grad_norm": 3.032942056655884, + "learning_rate": 9.951259138911455e-07, + "loss": 1.5316, + "step": 2262 + }, + { + "epoch": 2.48, + "grad_norm": 2.9982450008392334, + "learning_rate": 9.930950446791227e-07, + "loss": 1.4703, + "step": 2263 + }, + { + "epoch": 2.481095890410959, + "grad_norm": 3.2703280448913574, + "learning_rate": 9.910641754671e-07, + "loss": 1.5701, + "step": 2264 + }, + { + "epoch": 2.4821917808219176, + "grad_norm": 3.2202248573303223, + "learning_rate": 9.890333062550773e-07, + "loss": 1.6089, + "step": 2265 + }, + { + "epoch": 2.4832876712328766, + "grad_norm": 3.120250940322876, + "learning_rate": 9.870024370430546e-07, + "loss": 1.5333, + "step": 2266 + }, + { + "epoch": 2.4843835616438357, + "grad_norm": 3.1418509483337402, + "learning_rate": 9.849715678310318e-07, + "loss": 1.5408, + "step": 2267 + }, + { + "epoch": 2.4854794520547947, + "grad_norm": 3.055638313293457, + "learning_rate": 9.829406986190091e-07, + "loss": 1.5287, + "step": 2268 + }, + { + "epoch": 2.4865753424657533, + "grad_norm": 3.1318788528442383, + "learning_rate": 9.809098294069864e-07, + "loss": 1.5405, + "step": 2269 + }, + { + "epoch": 2.4876712328767123, + "grad_norm": 2.9671645164489746, + "learning_rate": 9.788789601949637e-07, + "loss": 1.4672, + "step": 2270 + }, + { + "epoch": 2.4876712328767123, + "eval_loss": 1.7990403175354004, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 2270 + }, + { + "epoch": 2.4887671232876714, + "grad_norm": 2.922785758972168, + "learning_rate": 9.768480909829407e-07, + "loss": 1.5212, + "step": 2271 + }, + { + "epoch": 2.48986301369863, + "grad_norm": 3.1482293605804443, + "learning_rate": 9.74817221770918e-07, + "loss": 1.5881, + "step": 2272 + }, + { + "epoch": 2.490958904109589, + "grad_norm": 3.051828145980835, + "learning_rate": 9.727863525588953e-07, + "loss": 1.5162, + "step": 2273 + }, + { + "epoch": 2.492054794520548, + "grad_norm": 3.0629403591156006, + "learning_rate": 9.707554833468726e-07, + "loss": 1.5262, + "step": 2274 + }, + { + "epoch": 2.493150684931507, + "grad_norm": 3.111855983734131, + "learning_rate": 9.687246141348498e-07, + "loss": 1.4958, + "step": 2275 + }, + { + "epoch": 2.4942465753424656, + "grad_norm": 3.0235540866851807, + "learning_rate": 9.666937449228271e-07, + "loss": 1.502, + "step": 2276 + }, + { + "epoch": 2.4953424657534247, + "grad_norm": 3.1183176040649414, + "learning_rate": 9.646628757108042e-07, + "loss": 1.5247, + "step": 2277 + }, + { + "epoch": 2.4964383561643837, + "grad_norm": 3.0862016677856445, + "learning_rate": 9.626320064987814e-07, + "loss": 1.4932, + "step": 2278 + }, + { + "epoch": 2.4975342465753423, + "grad_norm": 2.9760987758636475, + "learning_rate": 9.606011372867587e-07, + "loss": 1.5418, + "step": 2279 + }, + { + "epoch": 2.4986301369863013, + "grad_norm": 3.1599209308624268, + "learning_rate": 9.58570268074736e-07, + "loss": 1.5052, + "step": 2280 + }, + { + "epoch": 2.4986301369863013, + "eval_loss": 1.8005189895629883, + "eval_runtime": 37.2162, + "eval_samples_per_second": 26.87, + "eval_steps_per_second": 1.129, + "step": 2280 + }, + { + "epoch": 2.4997260273972604, + "grad_norm": 3.2794387340545654, + "learning_rate": 9.565393988627133e-07, + "loss": 1.5253, + "step": 2281 + }, + { + "epoch": 2.500821917808219, + "grad_norm": 3.001866340637207, + "learning_rate": 9.545085296506906e-07, + "loss": 1.5206, + "step": 2282 + }, + { + "epoch": 2.501917808219178, + "grad_norm": 3.244438886642456, + "learning_rate": 9.524776604386677e-07, + "loss": 1.547, + "step": 2283 + }, + { + "epoch": 2.503013698630137, + "grad_norm": 2.9997920989990234, + "learning_rate": 9.50446791226645e-07, + "loss": 1.5292, + "step": 2284 + }, + { + "epoch": 2.504109589041096, + "grad_norm": 3.0492494106292725, + "learning_rate": 9.484159220146223e-07, + "loss": 1.469, + "step": 2285 + }, + { + "epoch": 2.5052054794520546, + "grad_norm": 3.0012149810791016, + "learning_rate": 9.463850528025995e-07, + "loss": 1.5772, + "step": 2286 + }, + { + "epoch": 2.5063013698630137, + "grad_norm": 3.0322391986846924, + "learning_rate": 9.443541835905768e-07, + "loss": 1.5383, + "step": 2287 + }, + { + "epoch": 2.5073972602739727, + "grad_norm": 3.083080768585205, + "learning_rate": 9.423233143785541e-07, + "loss": 1.6006, + "step": 2288 + }, + { + "epoch": 2.5084931506849317, + "grad_norm": 3.260403871536255, + "learning_rate": 9.402924451665313e-07, + "loss": 1.5997, + "step": 2289 + }, + { + "epoch": 2.5095890410958903, + "grad_norm": 2.9607059955596924, + "learning_rate": 9.382615759545085e-07, + "loss": 1.4991, + "step": 2290 + }, + { + "epoch": 2.5095890410958903, + "eval_loss": 1.7979460954666138, + "eval_runtime": 37.2022, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 2290 + }, + { + "epoch": 2.5106849315068493, + "grad_norm": 3.1235525608062744, + "learning_rate": 9.362307067424858e-07, + "loss": 1.5455, + "step": 2291 + }, + { + "epoch": 2.5117808219178084, + "grad_norm": 3.1077470779418945, + "learning_rate": 9.341998375304631e-07, + "loss": 1.5144, + "step": 2292 + }, + { + "epoch": 2.512876712328767, + "grad_norm": 3.116363525390625, + "learning_rate": 9.321689683184404e-07, + "loss": 1.5675, + "step": 2293 + }, + { + "epoch": 2.513972602739726, + "grad_norm": 3.051565647125244, + "learning_rate": 9.301380991064176e-07, + "loss": 1.49, + "step": 2294 + }, + { + "epoch": 2.515068493150685, + "grad_norm": 3.14814829826355, + "learning_rate": 9.281072298943949e-07, + "loss": 1.4752, + "step": 2295 + }, + { + "epoch": 2.5161643835616436, + "grad_norm": 3.072158098220825, + "learning_rate": 9.260763606823721e-07, + "loss": 1.5109, + "step": 2296 + }, + { + "epoch": 2.5172602739726027, + "grad_norm": 3.3150458335876465, + "learning_rate": 9.240454914703494e-07, + "loss": 1.5393, + "step": 2297 + }, + { + "epoch": 2.5183561643835617, + "grad_norm": 3.0790274143218994, + "learning_rate": 9.220146222583266e-07, + "loss": 1.5439, + "step": 2298 + }, + { + "epoch": 2.5194520547945203, + "grad_norm": 3.0580363273620605, + "learning_rate": 9.199837530463039e-07, + "loss": 1.4959, + "step": 2299 + }, + { + "epoch": 2.5205479452054793, + "grad_norm": 3.077010154724121, + "learning_rate": 9.179528838342812e-07, + "loss": 1.5255, + "step": 2300 + }, + { + "epoch": 2.5205479452054793, + "eval_loss": 1.7971885204315186, + "eval_runtime": 37.2976, + "eval_samples_per_second": 26.811, + "eval_steps_per_second": 1.126, + "step": 2300 + }, + { + "epoch": 2.5216438356164383, + "grad_norm": 3.12821102142334, + "learning_rate": 9.159220146222585e-07, + "loss": 1.5193, + "step": 2301 + }, + { + "epoch": 2.5227397260273974, + "grad_norm": 3.1390695571899414, + "learning_rate": 9.138911454102356e-07, + "loss": 1.5382, + "step": 2302 + }, + { + "epoch": 2.5238356164383564, + "grad_norm": 2.8846688270568848, + "learning_rate": 9.118602761982129e-07, + "loss": 1.4563, + "step": 2303 + }, + { + "epoch": 2.524931506849315, + "grad_norm": 3.1843085289001465, + "learning_rate": 9.098294069861902e-07, + "loss": 1.5269, + "step": 2304 + }, + { + "epoch": 2.526027397260274, + "grad_norm": 3.0500407218933105, + "learning_rate": 9.077985377741675e-07, + "loss": 1.5271, + "step": 2305 + }, + { + "epoch": 2.527123287671233, + "grad_norm": 3.0183236598968506, + "learning_rate": 9.057676685621447e-07, + "loss": 1.5016, + "step": 2306 + }, + { + "epoch": 2.5282191780821917, + "grad_norm": 2.9937262535095215, + "learning_rate": 9.03736799350122e-07, + "loss": 1.459, + "step": 2307 + }, + { + "epoch": 2.5293150684931507, + "grad_norm": 3.206115245819092, + "learning_rate": 9.017059301380991e-07, + "loss": 1.5792, + "step": 2308 + }, + { + "epoch": 2.5304109589041097, + "grad_norm": 3.0888969898223877, + "learning_rate": 8.996750609260764e-07, + "loss": 1.5516, + "step": 2309 + }, + { + "epoch": 2.5315068493150683, + "grad_norm": 3.1038601398468018, + "learning_rate": 8.976441917140536e-07, + "loss": 1.5866, + "step": 2310 + }, + { + "epoch": 2.5315068493150683, + "eval_loss": 1.7977495193481445, + "eval_runtime": 36.8684, + "eval_samples_per_second": 27.124, + "eval_steps_per_second": 1.139, + "step": 2310 + }, + { + "epoch": 2.5326027397260273, + "grad_norm": 3.1794545650482178, + "learning_rate": 8.956133225020309e-07, + "loss": 1.5253, + "step": 2311 + }, + { + "epoch": 2.5336986301369864, + "grad_norm": 3.0712389945983887, + "learning_rate": 8.935824532900082e-07, + "loss": 1.5987, + "step": 2312 + }, + { + "epoch": 2.534794520547945, + "grad_norm": 3.003493309020996, + "learning_rate": 8.915515840779855e-07, + "loss": 1.5095, + "step": 2313 + }, + { + "epoch": 2.535890410958904, + "grad_norm": 2.973238468170166, + "learning_rate": 8.895207148659626e-07, + "loss": 1.4938, + "step": 2314 + }, + { + "epoch": 2.536986301369863, + "grad_norm": 3.0642077922821045, + "learning_rate": 8.874898456539399e-07, + "loss": 1.5608, + "step": 2315 + }, + { + "epoch": 2.538082191780822, + "grad_norm": 2.9557077884674072, + "learning_rate": 8.854589764419172e-07, + "loss": 1.5371, + "step": 2316 + }, + { + "epoch": 2.539178082191781, + "grad_norm": 3.1857285499572754, + "learning_rate": 8.834281072298945e-07, + "loss": 1.5329, + "step": 2317 + }, + { + "epoch": 2.5402739726027397, + "grad_norm": 2.9202258586883545, + "learning_rate": 8.813972380178717e-07, + "loss": 1.4842, + "step": 2318 + }, + { + "epoch": 2.5413698630136987, + "grad_norm": 3.0559780597686768, + "learning_rate": 8.79366368805849e-07, + "loss": 1.5195, + "step": 2319 + }, + { + "epoch": 2.5424657534246577, + "grad_norm": 3.0907952785491943, + "learning_rate": 8.773354995938262e-07, + "loss": 1.5533, + "step": 2320 + }, + { + "epoch": 2.5424657534246577, + "eval_loss": 1.7972254753112793, + "eval_runtime": 37.0715, + "eval_samples_per_second": 26.975, + "eval_steps_per_second": 1.133, + "step": 2320 + }, + { + "epoch": 2.5435616438356163, + "grad_norm": 2.9118940830230713, + "learning_rate": 8.753046303818035e-07, + "loss": 1.5019, + "step": 2321 + }, + { + "epoch": 2.5446575342465754, + "grad_norm": 2.923349618911743, + "learning_rate": 8.732737611697807e-07, + "loss": 1.4992, + "step": 2322 + }, + { + "epoch": 2.5457534246575344, + "grad_norm": 3.0067269802093506, + "learning_rate": 8.71242891957758e-07, + "loss": 1.5017, + "step": 2323 + }, + { + "epoch": 2.546849315068493, + "grad_norm": 3.0100181102752686, + "learning_rate": 8.692120227457353e-07, + "loss": 1.5653, + "step": 2324 + }, + { + "epoch": 2.547945205479452, + "grad_norm": 3.1054582595825195, + "learning_rate": 8.671811535337126e-07, + "loss": 1.5613, + "step": 2325 + }, + { + "epoch": 2.549041095890411, + "grad_norm": 3.315070867538452, + "learning_rate": 8.651502843216898e-07, + "loss": 1.5248, + "step": 2326 + }, + { + "epoch": 2.5501369863013696, + "grad_norm": 3.3228695392608643, + "learning_rate": 8.63119415109667e-07, + "loss": 1.5944, + "step": 2327 + }, + { + "epoch": 2.5512328767123287, + "grad_norm": 3.1381492614746094, + "learning_rate": 8.610885458976443e-07, + "loss": 1.5479, + "step": 2328 + }, + { + "epoch": 2.5523287671232877, + "grad_norm": 3.0747134685516357, + "learning_rate": 8.590576766856216e-07, + "loss": 1.549, + "step": 2329 + }, + { + "epoch": 2.5534246575342463, + "grad_norm": 2.9555068016052246, + "learning_rate": 8.570268074735988e-07, + "loss": 1.4895, + "step": 2330 + }, + { + "epoch": 2.5534246575342463, + "eval_loss": 1.794063687324524, + "eval_runtime": 37.1608, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 2330 + }, + { + "epoch": 2.5545205479452053, + "grad_norm": 2.8772523403167725, + "learning_rate": 8.549959382615761e-07, + "loss": 1.4958, + "step": 2331 + }, + { + "epoch": 2.5556164383561644, + "grad_norm": 3.0156941413879395, + "learning_rate": 8.529650690495534e-07, + "loss": 1.4979, + "step": 2332 + }, + { + "epoch": 2.5567123287671234, + "grad_norm": 2.9901013374328613, + "learning_rate": 8.509341998375305e-07, + "loss": 1.5308, + "step": 2333 + }, + { + "epoch": 2.5578082191780824, + "grad_norm": 2.9730477333068848, + "learning_rate": 8.489033306255077e-07, + "loss": 1.4726, + "step": 2334 + }, + { + "epoch": 2.558904109589041, + "grad_norm": 2.947180986404419, + "learning_rate": 8.46872461413485e-07, + "loss": 1.4537, + "step": 2335 + }, + { + "epoch": 2.56, + "grad_norm": 3.084651231765747, + "learning_rate": 8.448415922014623e-07, + "loss": 1.4904, + "step": 2336 + }, + { + "epoch": 2.561095890410959, + "grad_norm": 2.9790945053100586, + "learning_rate": 8.428107229894396e-07, + "loss": 1.5085, + "step": 2337 + }, + { + "epoch": 2.5621917808219177, + "grad_norm": 3.0045242309570312, + "learning_rate": 8.407798537774168e-07, + "loss": 1.5358, + "step": 2338 + }, + { + "epoch": 2.5632876712328767, + "grad_norm": 3.1456592082977295, + "learning_rate": 8.38748984565394e-07, + "loss": 1.5304, + "step": 2339 + }, + { + "epoch": 2.5643835616438357, + "grad_norm": 3.109750747680664, + "learning_rate": 8.367181153533713e-07, + "loss": 1.5096, + "step": 2340 + }, + { + "epoch": 2.5643835616438357, + "eval_loss": 1.7974557876586914, + "eval_runtime": 37.2006, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 2340 + }, + { + "epoch": 2.5654794520547943, + "grad_norm": 3.319150924682617, + "learning_rate": 8.346872461413486e-07, + "loss": 1.5134, + "step": 2341 + }, + { + "epoch": 2.5665753424657534, + "grad_norm": 3.2575652599334717, + "learning_rate": 8.326563769293258e-07, + "loss": 1.528, + "step": 2342 + }, + { + "epoch": 2.5676712328767124, + "grad_norm": 3.140742063522339, + "learning_rate": 8.306255077173031e-07, + "loss": 1.5189, + "step": 2343 + }, + { + "epoch": 2.568767123287671, + "grad_norm": 3.076082229614258, + "learning_rate": 8.285946385052804e-07, + "loss": 1.5953, + "step": 2344 + }, + { + "epoch": 2.56986301369863, + "grad_norm": 3.036564350128174, + "learning_rate": 8.265637692932575e-07, + "loss": 1.471, + "step": 2345 + }, + { + "epoch": 2.570958904109589, + "grad_norm": 2.9877052307128906, + "learning_rate": 8.245329000812348e-07, + "loss": 1.5544, + "step": 2346 + }, + { + "epoch": 2.572054794520548, + "grad_norm": 3.0643317699432373, + "learning_rate": 8.225020308692121e-07, + "loss": 1.5504, + "step": 2347 + }, + { + "epoch": 2.573150684931507, + "grad_norm": 2.9769644737243652, + "learning_rate": 8.204711616571894e-07, + "loss": 1.4828, + "step": 2348 + }, + { + "epoch": 2.5742465753424657, + "grad_norm": 3.032136917114258, + "learning_rate": 8.184402924451667e-07, + "loss": 1.5235, + "step": 2349 + }, + { + "epoch": 2.5753424657534247, + "grad_norm": 3.0864198207855225, + "learning_rate": 8.164094232331439e-07, + "loss": 1.4741, + "step": 2350 + }, + { + "epoch": 2.5753424657534247, + "eval_loss": 1.7938393354415894, + "eval_runtime": 37.1923, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 2350 + }, + { + "epoch": 2.5764383561643838, + "grad_norm": 3.471435546875, + "learning_rate": 8.14378554021121e-07, + "loss": 1.5469, + "step": 2351 + }, + { + "epoch": 2.5775342465753424, + "grad_norm": 3.0380778312683105, + "learning_rate": 8.123476848090983e-07, + "loss": 1.5283, + "step": 2352 + }, + { + "epoch": 2.5786301369863014, + "grad_norm": 2.977381706237793, + "learning_rate": 8.103168155970755e-07, + "loss": 1.4938, + "step": 2353 + }, + { + "epoch": 2.5797260273972604, + "grad_norm": 3.1362359523773193, + "learning_rate": 8.082859463850528e-07, + "loss": 1.4999, + "step": 2354 + }, + { + "epoch": 2.580821917808219, + "grad_norm": 3.0398125648498535, + "learning_rate": 8.062550771730301e-07, + "loss": 1.5279, + "step": 2355 + }, + { + "epoch": 2.581917808219178, + "grad_norm": 3.082155466079712, + "learning_rate": 8.042242079610074e-07, + "loss": 1.5236, + "step": 2356 + }, + { + "epoch": 2.583013698630137, + "grad_norm": 3.1600003242492676, + "learning_rate": 8.021933387489846e-07, + "loss": 1.5148, + "step": 2357 + }, + { + "epoch": 2.5841095890410957, + "grad_norm": 3.16538143157959, + "learning_rate": 8.001624695369618e-07, + "loss": 1.5069, + "step": 2358 + }, + { + "epoch": 2.5852054794520547, + "grad_norm": 3.0698554515838623, + "learning_rate": 7.981316003249391e-07, + "loss": 1.5225, + "step": 2359 + }, + { + "epoch": 2.5863013698630137, + "grad_norm": 3.2182974815368652, + "learning_rate": 7.961007311129164e-07, + "loss": 1.5523, + "step": 2360 + }, + { + "epoch": 2.5863013698630137, + "eval_loss": 1.798264980316162, + "eval_runtime": 37.2022, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 2360 + }, + { + "epoch": 2.5873972602739728, + "grad_norm": 2.8935117721557617, + "learning_rate": 7.940698619008936e-07, + "loss": 1.475, + "step": 2361 + }, + { + "epoch": 2.5884931506849314, + "grad_norm": 3.1140449047088623, + "learning_rate": 7.920389926888709e-07, + "loss": 1.5181, + "step": 2362 + }, + { + "epoch": 2.5895890410958904, + "grad_norm": 3.0791847705841064, + "learning_rate": 7.900081234768482e-07, + "loss": 1.621, + "step": 2363 + }, + { + "epoch": 2.5906849315068494, + "grad_norm": 2.9496588706970215, + "learning_rate": 7.879772542648254e-07, + "loss": 1.4655, + "step": 2364 + }, + { + "epoch": 2.5917808219178085, + "grad_norm": 3.1934399604797363, + "learning_rate": 7.859463850528026e-07, + "loss": 1.5467, + "step": 2365 + }, + { + "epoch": 2.592876712328767, + "grad_norm": 2.9768447875976562, + "learning_rate": 7.839155158407799e-07, + "loss": 1.5299, + "step": 2366 + }, + { + "epoch": 2.593972602739726, + "grad_norm": 3.0377297401428223, + "learning_rate": 7.818846466287572e-07, + "loss": 1.5182, + "step": 2367 + }, + { + "epoch": 2.595068493150685, + "grad_norm": 3.1019999980926514, + "learning_rate": 7.798537774167345e-07, + "loss": 1.5407, + "step": 2368 + }, + { + "epoch": 2.5961643835616437, + "grad_norm": 3.2140676975250244, + "learning_rate": 7.778229082047117e-07, + "loss": 1.539, + "step": 2369 + }, + { + "epoch": 2.5972602739726027, + "grad_norm": 3.0868372917175293, + "learning_rate": 7.757920389926889e-07, + "loss": 1.4917, + "step": 2370 + }, + { + "epoch": 2.5972602739726027, + "eval_loss": 1.7955214977264404, + "eval_runtime": 37.1895, + "eval_samples_per_second": 26.889, + "eval_steps_per_second": 1.129, + "step": 2370 + }, + { + "epoch": 2.5983561643835618, + "grad_norm": 3.048440456390381, + "learning_rate": 7.737611697806662e-07, + "loss": 1.5443, + "step": 2371 + }, + { + "epoch": 2.5994520547945204, + "grad_norm": 3.092072010040283, + "learning_rate": 7.717303005686435e-07, + "loss": 1.4901, + "step": 2372 + }, + { + "epoch": 2.6005479452054794, + "grad_norm": 2.969160795211792, + "learning_rate": 7.696994313566207e-07, + "loss": 1.5504, + "step": 2373 + }, + { + "epoch": 2.6016438356164384, + "grad_norm": 3.172166109085083, + "learning_rate": 7.67668562144598e-07, + "loss": 1.6071, + "step": 2374 + }, + { + "epoch": 2.602739726027397, + "grad_norm": 3.038698434829712, + "learning_rate": 7.656376929325753e-07, + "loss": 1.4694, + "step": 2375 + }, + { + "epoch": 2.603835616438356, + "grad_norm": 3.0223464965820312, + "learning_rate": 7.636068237205524e-07, + "loss": 1.4732, + "step": 2376 + }, + { + "epoch": 2.604931506849315, + "grad_norm": 3.101727247238159, + "learning_rate": 7.615759545085296e-07, + "loss": 1.5483, + "step": 2377 + }, + { + "epoch": 2.606027397260274, + "grad_norm": 3.0078635215759277, + "learning_rate": 7.595450852965069e-07, + "loss": 1.496, + "step": 2378 + }, + { + "epoch": 2.607123287671233, + "grad_norm": 3.0310235023498535, + "learning_rate": 7.575142160844842e-07, + "loss": 1.5061, + "step": 2379 + }, + { + "epoch": 2.6082191780821917, + "grad_norm": 3.0811500549316406, + "learning_rate": 7.554833468724615e-07, + "loss": 1.5496, + "step": 2380 + }, + { + "epoch": 2.6082191780821917, + "eval_loss": 1.7973451614379883, + "eval_runtime": 37.2306, + "eval_samples_per_second": 26.86, + "eval_steps_per_second": 1.128, + "step": 2380 + }, + { + "epoch": 2.6093150684931508, + "grad_norm": 3.055795192718506, + "learning_rate": 7.534524776604387e-07, + "loss": 1.4962, + "step": 2381 + }, + { + "epoch": 2.61041095890411, + "grad_norm": 3.017982244491577, + "learning_rate": 7.51421608448416e-07, + "loss": 1.485, + "step": 2382 + }, + { + "epoch": 2.6115068493150684, + "grad_norm": 3.167466402053833, + "learning_rate": 7.493907392363932e-07, + "loss": 1.5319, + "step": 2383 + }, + { + "epoch": 2.6126027397260274, + "grad_norm": 2.9665164947509766, + "learning_rate": 7.473598700243705e-07, + "loss": 1.4988, + "step": 2384 + }, + { + "epoch": 2.6136986301369864, + "grad_norm": 3.1092183589935303, + "learning_rate": 7.453290008123477e-07, + "loss": 1.4631, + "step": 2385 + }, + { + "epoch": 2.614794520547945, + "grad_norm": 3.131699800491333, + "learning_rate": 7.43298131600325e-07, + "loss": 1.5174, + "step": 2386 + }, + { + "epoch": 2.615890410958904, + "grad_norm": 3.0466747283935547, + "learning_rate": 7.412672623883023e-07, + "loss": 1.526, + "step": 2387 + }, + { + "epoch": 2.616986301369863, + "grad_norm": 3.0042684078216553, + "learning_rate": 7.392363931762796e-07, + "loss": 1.5067, + "step": 2388 + }, + { + "epoch": 2.6180821917808217, + "grad_norm": 2.98500919342041, + "learning_rate": 7.372055239642567e-07, + "loss": 1.4675, + "step": 2389 + }, + { + "epoch": 2.6191780821917807, + "grad_norm": 3.1981570720672607, + "learning_rate": 7.35174654752234e-07, + "loss": 1.5101, + "step": 2390 + }, + { + "epoch": 2.6191780821917807, + "eval_loss": 1.7951487302780151, + "eval_runtime": 37.16, + "eval_samples_per_second": 26.911, + "eval_steps_per_second": 1.13, + "step": 2390 + }, + { + "epoch": 2.6202739726027398, + "grad_norm": 3.1528236865997314, + "learning_rate": 7.331437855402113e-07, + "loss": 1.5426, + "step": 2391 + }, + { + "epoch": 2.621369863013699, + "grad_norm": 3.1162030696868896, + "learning_rate": 7.311129163281886e-07, + "loss": 1.5578, + "step": 2392 + }, + { + "epoch": 2.6224657534246574, + "grad_norm": 3.001851797103882, + "learning_rate": 7.290820471161658e-07, + "loss": 1.4719, + "step": 2393 + }, + { + "epoch": 2.6235616438356164, + "grad_norm": 3.12337064743042, + "learning_rate": 7.270511779041431e-07, + "loss": 1.5517, + "step": 2394 + }, + { + "epoch": 2.6246575342465754, + "grad_norm": 3.021188259124756, + "learning_rate": 7.250203086921203e-07, + "loss": 1.5562, + "step": 2395 + }, + { + "epoch": 2.6257534246575345, + "grad_norm": 2.927506923675537, + "learning_rate": 7.229894394800976e-07, + "loss": 1.4793, + "step": 2396 + }, + { + "epoch": 2.626849315068493, + "grad_norm": 2.9731264114379883, + "learning_rate": 7.209585702680748e-07, + "loss": 1.511, + "step": 2397 + }, + { + "epoch": 2.627945205479452, + "grad_norm": 3.02195405960083, + "learning_rate": 7.189277010560521e-07, + "loss": 1.5471, + "step": 2398 + }, + { + "epoch": 2.629041095890411, + "grad_norm": 2.9983696937561035, + "learning_rate": 7.168968318440294e-07, + "loss": 1.4707, + "step": 2399 + }, + { + "epoch": 2.6301369863013697, + "grad_norm": 2.978363513946533, + "learning_rate": 7.148659626320067e-07, + "loss": 1.5097, + "step": 2400 + }, + { + "epoch": 2.6301369863013697, + "eval_loss": 1.7952522039413452, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 2400 + }, + { + "epoch": 2.6312328767123287, + "grad_norm": 2.9679200649261475, + "learning_rate": 7.128350934199837e-07, + "loss": 1.4826, + "step": 2401 + }, + { + "epoch": 2.632328767123288, + "grad_norm": 3.200427293777466, + "learning_rate": 7.10804224207961e-07, + "loss": 1.5683, + "step": 2402 + }, + { + "epoch": 2.6334246575342464, + "grad_norm": 3.23142409324646, + "learning_rate": 7.087733549959383e-07, + "loss": 1.5547, + "step": 2403 + }, + { + "epoch": 2.6345205479452054, + "grad_norm": 2.960176706314087, + "learning_rate": 7.067424857839155e-07, + "loss": 1.5101, + "step": 2404 + }, + { + "epoch": 2.6356164383561644, + "grad_norm": 3.000246286392212, + "learning_rate": 7.047116165718928e-07, + "loss": 1.5053, + "step": 2405 + }, + { + "epoch": 2.636712328767123, + "grad_norm": 3.0891222953796387, + "learning_rate": 7.026807473598701e-07, + "loss": 1.523, + "step": 2406 + }, + { + "epoch": 2.637808219178082, + "grad_norm": 3.108332633972168, + "learning_rate": 7.006498781478473e-07, + "loss": 1.5048, + "step": 2407 + }, + { + "epoch": 2.638904109589041, + "grad_norm": 3.2009339332580566, + "learning_rate": 6.986190089358245e-07, + "loss": 1.5752, + "step": 2408 + }, + { + "epoch": 2.64, + "grad_norm": 3.0235002040863037, + "learning_rate": 6.965881397238018e-07, + "loss": 1.4905, + "step": 2409 + }, + { + "epoch": 2.641095890410959, + "grad_norm": 2.922283411026001, + "learning_rate": 6.945572705117791e-07, + "loss": 1.5213, + "step": 2410 + }, + { + "epoch": 2.641095890410959, + "eval_loss": 1.793642520904541, + "eval_runtime": 36.7927, + "eval_samples_per_second": 27.179, + "eval_steps_per_second": 1.142, + "step": 2410 + }, + { + "epoch": 2.6421917808219177, + "grad_norm": 2.9516892433166504, + "learning_rate": 6.925264012997564e-07, + "loss": 1.5529, + "step": 2411 + }, + { + "epoch": 2.643287671232877, + "grad_norm": 2.9784276485443115, + "learning_rate": 6.904955320877336e-07, + "loss": 1.5354, + "step": 2412 + }, + { + "epoch": 2.644383561643836, + "grad_norm": 3.0090866088867188, + "learning_rate": 6.884646628757109e-07, + "loss": 1.536, + "step": 2413 + }, + { + "epoch": 2.6454794520547944, + "grad_norm": 3.0705292224884033, + "learning_rate": 6.864337936636881e-07, + "loss": 1.533, + "step": 2414 + }, + { + "epoch": 2.6465753424657534, + "grad_norm": 3.071178436279297, + "learning_rate": 6.844029244516654e-07, + "loss": 1.5829, + "step": 2415 + }, + { + "epoch": 2.6476712328767125, + "grad_norm": 3.083031177520752, + "learning_rate": 6.823720552396426e-07, + "loss": 1.5843, + "step": 2416 + }, + { + "epoch": 2.648767123287671, + "grad_norm": 2.956490993499756, + "learning_rate": 6.803411860276199e-07, + "loss": 1.4679, + "step": 2417 + }, + { + "epoch": 2.64986301369863, + "grad_norm": 2.9852354526519775, + "learning_rate": 6.783103168155972e-07, + "loss": 1.551, + "step": 2418 + }, + { + "epoch": 2.650958904109589, + "grad_norm": 2.925212860107422, + "learning_rate": 6.762794476035745e-07, + "loss": 1.5336, + "step": 2419 + }, + { + "epoch": 2.6520547945205477, + "grad_norm": 3.0673046112060547, + "learning_rate": 6.742485783915515e-07, + "loss": 1.5216, + "step": 2420 + }, + { + "epoch": 2.6520547945205477, + "eval_loss": 1.7941372394561768, + "eval_runtime": 37.0302, + "eval_samples_per_second": 27.005, + "eval_steps_per_second": 1.134, + "step": 2420 + }, + { + "epoch": 2.6531506849315067, + "grad_norm": 3.1326067447662354, + "learning_rate": 6.722177091795288e-07, + "loss": 1.5622, + "step": 2421 + }, + { + "epoch": 2.6542465753424658, + "grad_norm": 3.1275596618652344, + "learning_rate": 6.701868399675061e-07, + "loss": 1.5486, + "step": 2422 + }, + { + "epoch": 2.655342465753425, + "grad_norm": 3.1078732013702393, + "learning_rate": 6.681559707554834e-07, + "loss": 1.5447, + "step": 2423 + }, + { + "epoch": 2.6564383561643834, + "grad_norm": 2.9929449558258057, + "learning_rate": 6.661251015434606e-07, + "loss": 1.5025, + "step": 2424 + }, + { + "epoch": 2.6575342465753424, + "grad_norm": 3.0392446517944336, + "learning_rate": 6.640942323314379e-07, + "loss": 1.4899, + "step": 2425 + }, + { + "epoch": 2.6586301369863015, + "grad_norm": 3.0346617698669434, + "learning_rate": 6.620633631194151e-07, + "loss": 1.4928, + "step": 2426 + }, + { + "epoch": 2.6597260273972605, + "grad_norm": 2.8798530101776123, + "learning_rate": 6.600324939073924e-07, + "loss": 1.5303, + "step": 2427 + }, + { + "epoch": 2.660821917808219, + "grad_norm": 2.9803898334503174, + "learning_rate": 6.580016246953696e-07, + "loss": 1.539, + "step": 2428 + }, + { + "epoch": 2.661917808219178, + "grad_norm": 3.0963869094848633, + "learning_rate": 6.559707554833469e-07, + "loss": 1.4965, + "step": 2429 + }, + { + "epoch": 2.663013698630137, + "grad_norm": 3.211057424545288, + "learning_rate": 6.539398862713242e-07, + "loss": 1.5838, + "step": 2430 + }, + { + "epoch": 2.663013698630137, + "eval_loss": 1.7935749292373657, + "eval_runtime": 37.1224, + "eval_samples_per_second": 26.938, + "eval_steps_per_second": 1.131, + "step": 2430 + }, + { + "epoch": 2.6641095890410957, + "grad_norm": 3.0385782718658447, + "learning_rate": 6.519090170593015e-07, + "loss": 1.5031, + "step": 2431 + }, + { + "epoch": 2.6652054794520548, + "grad_norm": 3.094062328338623, + "learning_rate": 6.498781478472786e-07, + "loss": 1.5756, + "step": 2432 + }, + { + "epoch": 2.666301369863014, + "grad_norm": 3.1397554874420166, + "learning_rate": 6.478472786352559e-07, + "loss": 1.5406, + "step": 2433 + }, + { + "epoch": 2.6673972602739724, + "grad_norm": 3.0731546878814697, + "learning_rate": 6.458164094232332e-07, + "loss": 1.557, + "step": 2434 + }, + { + "epoch": 2.6684931506849314, + "grad_norm": 3.0219526290893555, + "learning_rate": 6.437855402112105e-07, + "loss": 1.52, + "step": 2435 + }, + { + "epoch": 2.6695890410958905, + "grad_norm": 3.072672128677368, + "learning_rate": 6.417546709991877e-07, + "loss": 1.469, + "step": 2436 + }, + { + "epoch": 2.670684931506849, + "grad_norm": 3.137406826019287, + "learning_rate": 6.39723801787165e-07, + "loss": 1.4968, + "step": 2437 + }, + { + "epoch": 2.671780821917808, + "grad_norm": 3.0841257572174072, + "learning_rate": 6.376929325751423e-07, + "loss": 1.5499, + "step": 2438 + }, + { + "epoch": 2.672876712328767, + "grad_norm": 3.013788938522339, + "learning_rate": 6.356620633631195e-07, + "loss": 1.5216, + "step": 2439 + }, + { + "epoch": 2.673972602739726, + "grad_norm": 2.9357147216796875, + "learning_rate": 6.336311941510967e-07, + "loss": 1.5075, + "step": 2440 + }, + { + "epoch": 2.673972602739726, + "eval_loss": 1.7945020198822021, + "eval_runtime": 37.1916, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 2440 + }, + { + "epoch": 2.675068493150685, + "grad_norm": 3.0886213779449463, + "learning_rate": 6.31600324939074e-07, + "loss": 1.4917, + "step": 2441 + }, + { + "epoch": 2.6761643835616438, + "grad_norm": 3.053446054458618, + "learning_rate": 6.295694557270513e-07, + "loss": 1.5294, + "step": 2442 + }, + { + "epoch": 2.677260273972603, + "grad_norm": 3.089285135269165, + "learning_rate": 6.275385865150286e-07, + "loss": 1.5642, + "step": 2443 + }, + { + "epoch": 2.678356164383562, + "grad_norm": 3.050245523452759, + "learning_rate": 6.255077173030058e-07, + "loss": 1.5349, + "step": 2444 + }, + { + "epoch": 2.6794520547945204, + "grad_norm": 3.0645339488983154, + "learning_rate": 6.23476848090983e-07, + "loss": 1.5199, + "step": 2445 + }, + { + "epoch": 2.6805479452054795, + "grad_norm": 3.1500461101531982, + "learning_rate": 6.214459788789602e-07, + "loss": 1.5207, + "step": 2446 + }, + { + "epoch": 2.6816438356164385, + "grad_norm": 3.0756518840789795, + "learning_rate": 6.194151096669375e-07, + "loss": 1.5196, + "step": 2447 + }, + { + "epoch": 2.682739726027397, + "grad_norm": 3.110304594039917, + "learning_rate": 6.173842404549147e-07, + "loss": 1.5655, + "step": 2448 + }, + { + "epoch": 2.683835616438356, + "grad_norm": 3.282001495361328, + "learning_rate": 6.15353371242892e-07, + "loss": 1.5035, + "step": 2449 + }, + { + "epoch": 2.684931506849315, + "grad_norm": 3.086653709411621, + "learning_rate": 6.133225020308693e-07, + "loss": 1.5169, + "step": 2450 + }, + { + "epoch": 2.684931506849315, + "eval_loss": 1.7946088314056396, + "eval_runtime": 37.1196, + "eval_samples_per_second": 26.94, + "eval_steps_per_second": 1.131, + "step": 2450 + }, + { + "epoch": 2.6860273972602737, + "grad_norm": 3.133683443069458, + "learning_rate": 6.112916328188466e-07, + "loss": 1.4536, + "step": 2451 + }, + { + "epoch": 2.6871232876712328, + "grad_norm": 3.0638301372528076, + "learning_rate": 6.092607636068238e-07, + "loss": 1.5105, + "step": 2452 + }, + { + "epoch": 2.688219178082192, + "grad_norm": 3.1290526390075684, + "learning_rate": 6.07229894394801e-07, + "loss": 1.5884, + "step": 2453 + }, + { + "epoch": 2.689315068493151, + "grad_norm": 2.979188919067383, + "learning_rate": 6.051990251827783e-07, + "loss": 1.4983, + "step": 2454 + }, + { + "epoch": 2.69041095890411, + "grad_norm": 2.994475841522217, + "learning_rate": 6.031681559707556e-07, + "loss": 1.5049, + "step": 2455 + }, + { + "epoch": 2.6915068493150685, + "grad_norm": 2.9304025173187256, + "learning_rate": 6.011372867587327e-07, + "loss": 1.5474, + "step": 2456 + }, + { + "epoch": 2.6926027397260275, + "grad_norm": 3.293808698654175, + "learning_rate": 5.9910641754671e-07, + "loss": 1.5259, + "step": 2457 + }, + { + "epoch": 2.6936986301369865, + "grad_norm": 2.9812941551208496, + "learning_rate": 5.970755483346873e-07, + "loss": 1.5044, + "step": 2458 + }, + { + "epoch": 2.694794520547945, + "grad_norm": 3.121180295944214, + "learning_rate": 5.950446791226646e-07, + "loss": 1.5145, + "step": 2459 + }, + { + "epoch": 2.695890410958904, + "grad_norm": 3.028796434402466, + "learning_rate": 5.930138099106418e-07, + "loss": 1.5004, + "step": 2460 + }, + { + "epoch": 2.695890410958904, + "eval_loss": 1.794010043144226, + "eval_runtime": 37.1506, + "eval_samples_per_second": 26.917, + "eval_steps_per_second": 1.131, + "step": 2460 + }, + { + "epoch": 2.696986301369863, + "grad_norm": 3.0366451740264893, + "learning_rate": 5.909829406986191e-07, + "loss": 1.506, + "step": 2461 + }, + { + "epoch": 2.6980821917808218, + "grad_norm": 2.9850924015045166, + "learning_rate": 5.889520714865963e-07, + "loss": 1.5112, + "step": 2462 + }, + { + "epoch": 2.699178082191781, + "grad_norm": 3.078281879425049, + "learning_rate": 5.869212022745735e-07, + "loss": 1.5259, + "step": 2463 + }, + { + "epoch": 2.70027397260274, + "grad_norm": 2.9461851119995117, + "learning_rate": 5.848903330625508e-07, + "loss": 1.4813, + "step": 2464 + }, + { + "epoch": 2.7013698630136984, + "grad_norm": 3.152405261993408, + "learning_rate": 5.828594638505281e-07, + "loss": 1.5301, + "step": 2465 + }, + { + "epoch": 2.7024657534246574, + "grad_norm": 2.9270081520080566, + "learning_rate": 5.808285946385054e-07, + "loss": 1.5131, + "step": 2466 + }, + { + "epoch": 2.7035616438356165, + "grad_norm": 2.9786133766174316, + "learning_rate": 5.787977254264827e-07, + "loss": 1.5078, + "step": 2467 + }, + { + "epoch": 2.704657534246575, + "grad_norm": 3.0664288997650146, + "learning_rate": 5.767668562144598e-07, + "loss": 1.5181, + "step": 2468 + }, + { + "epoch": 2.705753424657534, + "grad_norm": 3.100987195968628, + "learning_rate": 5.747359870024371e-07, + "loss": 1.5724, + "step": 2469 + }, + { + "epoch": 2.706849315068493, + "grad_norm": 3.1864078044891357, + "learning_rate": 5.727051177904144e-07, + "loss": 1.5061, + "step": 2470 + }, + { + "epoch": 2.706849315068493, + "eval_loss": 1.794380784034729, + "eval_runtime": 37.1631, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 2470 + }, + { + "epoch": 2.707945205479452, + "grad_norm": 3.060481071472168, + "learning_rate": 5.706742485783915e-07, + "loss": 1.5203, + "step": 2471 + }, + { + "epoch": 2.709041095890411, + "grad_norm": 3.142359972000122, + "learning_rate": 5.686433793663688e-07, + "loss": 1.5542, + "step": 2472 + }, + { + "epoch": 2.71013698630137, + "grad_norm": 3.0684356689453125, + "learning_rate": 5.666125101543461e-07, + "loss": 1.4767, + "step": 2473 + }, + { + "epoch": 2.711232876712329, + "grad_norm": 2.994058609008789, + "learning_rate": 5.645816409423234e-07, + "loss": 1.4633, + "step": 2474 + }, + { + "epoch": 2.712328767123288, + "grad_norm": 3.107590436935425, + "learning_rate": 5.625507717303006e-07, + "loss": 1.5778, + "step": 2475 + }, + { + "epoch": 2.7134246575342464, + "grad_norm": 3.068570613861084, + "learning_rate": 5.605199025182779e-07, + "loss": 1.5193, + "step": 2476 + }, + { + "epoch": 2.7145205479452055, + "grad_norm": 2.893538236618042, + "learning_rate": 5.584890333062551e-07, + "loss": 1.5034, + "step": 2477 + }, + { + "epoch": 2.7156164383561645, + "grad_norm": 2.9960875511169434, + "learning_rate": 5.564581640942324e-07, + "loss": 1.4942, + "step": 2478 + }, + { + "epoch": 2.716712328767123, + "grad_norm": 3.087170124053955, + "learning_rate": 5.544272948822096e-07, + "loss": 1.563, + "step": 2479 + }, + { + "epoch": 2.717808219178082, + "grad_norm": 3.0415587425231934, + "learning_rate": 5.523964256701868e-07, + "loss": 1.5014, + "step": 2480 + }, + { + "epoch": 2.717808219178082, + "eval_loss": 1.7932865619659424, + "eval_runtime": 37.1873, + "eval_samples_per_second": 26.891, + "eval_steps_per_second": 1.129, + "step": 2480 + }, + { + "epoch": 2.718904109589041, + "grad_norm": 2.9672157764434814, + "learning_rate": 5.503655564581641e-07, + "loss": 1.5224, + "step": 2481 + }, + { + "epoch": 2.7199999999999998, + "grad_norm": 3.0993940830230713, + "learning_rate": 5.483346872461414e-07, + "loss": 1.5473, + "step": 2482 + }, + { + "epoch": 2.721095890410959, + "grad_norm": 2.986877918243408, + "learning_rate": 5.463038180341186e-07, + "loss": 1.5016, + "step": 2483 + }, + { + "epoch": 2.722191780821918, + "grad_norm": 3.0450809001922607, + "learning_rate": 5.442729488220959e-07, + "loss": 1.5269, + "step": 2484 + }, + { + "epoch": 2.723287671232877, + "grad_norm": 3.173659563064575, + "learning_rate": 5.422420796100732e-07, + "loss": 1.4728, + "step": 2485 + }, + { + "epoch": 2.724383561643836, + "grad_norm": 3.1881349086761475, + "learning_rate": 5.402112103980505e-07, + "loss": 1.5194, + "step": 2486 + }, + { + "epoch": 2.7254794520547945, + "grad_norm": 3.026097536087036, + "learning_rate": 5.381803411860276e-07, + "loss": 1.5116, + "step": 2487 + }, + { + "epoch": 2.7265753424657535, + "grad_norm": 3.087247371673584, + "learning_rate": 5.361494719740049e-07, + "loss": 1.5529, + "step": 2488 + }, + { + "epoch": 2.7276712328767125, + "grad_norm": 3.2861785888671875, + "learning_rate": 5.341186027619822e-07, + "loss": 1.5826, + "step": 2489 + }, + { + "epoch": 2.728767123287671, + "grad_norm": 2.9725217819213867, + "learning_rate": 5.320877335499594e-07, + "loss": 1.4895, + "step": 2490 + }, + { + "epoch": 2.728767123287671, + "eval_loss": 1.7935879230499268, + "eval_runtime": 37.1175, + "eval_samples_per_second": 26.941, + "eval_steps_per_second": 1.132, + "step": 2490 + }, + { + "epoch": 2.72986301369863, + "grad_norm": 3.196855068206787, + "learning_rate": 5.300568643379366e-07, + "loss": 1.5027, + "step": 2491 + }, + { + "epoch": 2.730958904109589, + "grad_norm": 3.107867956161499, + "learning_rate": 5.280259951259139e-07, + "loss": 1.5775, + "step": 2492 + }, + { + "epoch": 2.732054794520548, + "grad_norm": 3.1798512935638428, + "learning_rate": 5.259951259138912e-07, + "loss": 1.5307, + "step": 2493 + }, + { + "epoch": 2.733150684931507, + "grad_norm": 3.231285572052002, + "learning_rate": 5.239642567018685e-07, + "loss": 1.497, + "step": 2494 + }, + { + "epoch": 2.734246575342466, + "grad_norm": 3.0765933990478516, + "learning_rate": 5.219333874898457e-07, + "loss": 1.4904, + "step": 2495 + }, + { + "epoch": 2.7353424657534244, + "grad_norm": 3.0214293003082275, + "learning_rate": 5.199025182778229e-07, + "loss": 1.5045, + "step": 2496 + }, + { + "epoch": 2.7364383561643835, + "grad_norm": 3.1160240173339844, + "learning_rate": 5.178716490658002e-07, + "loss": 1.5285, + "step": 2497 + }, + { + "epoch": 2.7375342465753425, + "grad_norm": 3.013947010040283, + "learning_rate": 5.158407798537775e-07, + "loss": 1.5078, + "step": 2498 + }, + { + "epoch": 2.7386301369863015, + "grad_norm": 3.0118439197540283, + "learning_rate": 5.138099106417547e-07, + "loss": 1.4794, + "step": 2499 + }, + { + "epoch": 2.73972602739726, + "grad_norm": 2.976755380630493, + "learning_rate": 5.11779041429732e-07, + "loss": 1.4494, + "step": 2500 + }, + { + "epoch": 2.73972602739726, + "eval_loss": 1.791616439819336, + "eval_runtime": 37.1554, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 2500 + }, + { + "epoch": 2.740821917808219, + "grad_norm": 3.0231242179870605, + "learning_rate": 5.097481722177093e-07, + "loss": 1.5071, + "step": 2501 + }, + { + "epoch": 2.741917808219178, + "grad_norm": 3.108156442642212, + "learning_rate": 5.077173030056865e-07, + "loss": 1.4962, + "step": 2502 + }, + { + "epoch": 2.743013698630137, + "grad_norm": 3.2055060863494873, + "learning_rate": 5.056864337936637e-07, + "loss": 1.5675, + "step": 2503 + }, + { + "epoch": 2.744109589041096, + "grad_norm": 3.0649781227111816, + "learning_rate": 5.03655564581641e-07, + "loss": 1.5521, + "step": 2504 + }, + { + "epoch": 2.745205479452055, + "grad_norm": 3.1816694736480713, + "learning_rate": 5.016246953696182e-07, + "loss": 1.543, + "step": 2505 + }, + { + "epoch": 2.746301369863014, + "grad_norm": 3.316678285598755, + "learning_rate": 4.995938261575955e-07, + "loss": 1.6335, + "step": 2506 + }, + { + "epoch": 2.7473972602739725, + "grad_norm": 3.123028516769409, + "learning_rate": 4.975629569455727e-07, + "loss": 1.4836, + "step": 2507 + }, + { + "epoch": 2.7484931506849315, + "grad_norm": 3.0578713417053223, + "learning_rate": 4.9553208773355e-07, + "loss": 1.5301, + "step": 2508 + }, + { + "epoch": 2.7495890410958905, + "grad_norm": 3.0232229232788086, + "learning_rate": 4.935012185215273e-07, + "loss": 1.5856, + "step": 2509 + }, + { + "epoch": 2.750684931506849, + "grad_norm": 2.906506061553955, + "learning_rate": 4.914703493095046e-07, + "loss": 1.5181, + "step": 2510 + }, + { + "epoch": 2.750684931506849, + "eval_loss": 1.7911875247955322, + "eval_runtime": 36.8288, + "eval_samples_per_second": 27.153, + "eval_steps_per_second": 1.14, + "step": 2510 + }, + { + "epoch": 2.751780821917808, + "grad_norm": 3.1703567504882812, + "learning_rate": 4.894394800974818e-07, + "loss": 1.48, + "step": 2511 + }, + { + "epoch": 2.752876712328767, + "grad_norm": 3.089897394180298, + "learning_rate": 4.87408610885459e-07, + "loss": 1.4843, + "step": 2512 + }, + { + "epoch": 2.7539726027397258, + "grad_norm": 3.0825729370117188, + "learning_rate": 4.853777416734363e-07, + "loss": 1.6004, + "step": 2513 + }, + { + "epoch": 2.755068493150685, + "grad_norm": 3.0502867698669434, + "learning_rate": 4.833468724614136e-07, + "loss": 1.5501, + "step": 2514 + }, + { + "epoch": 2.756164383561644, + "grad_norm": 3.188497543334961, + "learning_rate": 4.813160032493907e-07, + "loss": 1.5212, + "step": 2515 + }, + { + "epoch": 2.757260273972603, + "grad_norm": 3.026137351989746, + "learning_rate": 4.79285134037368e-07, + "loss": 1.4996, + "step": 2516 + }, + { + "epoch": 2.758356164383562, + "grad_norm": 3.0630528926849365, + "learning_rate": 4.772542648253453e-07, + "loss": 1.4752, + "step": 2517 + }, + { + "epoch": 2.7594520547945205, + "grad_norm": 3.0613367557525635, + "learning_rate": 4.752233956133225e-07, + "loss": 1.4897, + "step": 2518 + }, + { + "epoch": 2.7605479452054795, + "grad_norm": 3.2524094581604004, + "learning_rate": 4.731925264012998e-07, + "loss": 1.5155, + "step": 2519 + }, + { + "epoch": 2.7616438356164386, + "grad_norm": 3.218522310256958, + "learning_rate": 4.7116165718927705e-07, + "loss": 1.572, + "step": 2520 + }, + { + "epoch": 2.7616438356164386, + "eval_loss": 1.7935888767242432, + "eval_runtime": 37.0328, + "eval_samples_per_second": 27.003, + "eval_steps_per_second": 1.134, + "step": 2520 + }, + { + "epoch": 2.762739726027397, + "grad_norm": 3.080104112625122, + "learning_rate": 4.6913078797725427e-07, + "loss": 1.5323, + "step": 2521 + }, + { + "epoch": 2.763835616438356, + "grad_norm": 3.2323362827301025, + "learning_rate": 4.6709991876523155e-07, + "loss": 1.5327, + "step": 2522 + }, + { + "epoch": 2.764931506849315, + "grad_norm": 3.128486394882202, + "learning_rate": 4.650690495532088e-07, + "loss": 1.5108, + "step": 2523 + }, + { + "epoch": 2.766027397260274, + "grad_norm": 3.2815701961517334, + "learning_rate": 4.6303818034118605e-07, + "loss": 1.5923, + "step": 2524 + }, + { + "epoch": 2.767123287671233, + "grad_norm": 3.1217830181121826, + "learning_rate": 4.610073111291633e-07, + "loss": 1.527, + "step": 2525 + }, + { + "epoch": 2.768219178082192, + "grad_norm": 2.9235475063323975, + "learning_rate": 4.589764419171406e-07, + "loss": 1.4587, + "step": 2526 + }, + { + "epoch": 2.7693150684931505, + "grad_norm": 3.0745975971221924, + "learning_rate": 4.569455727051178e-07, + "loss": 1.5, + "step": 2527 + }, + { + "epoch": 2.7704109589041095, + "grad_norm": 3.066887140274048, + "learning_rate": 4.549147034930951e-07, + "loss": 1.4934, + "step": 2528 + }, + { + "epoch": 2.7715068493150685, + "grad_norm": 3.014899969100952, + "learning_rate": 4.5288383428107237e-07, + "loss": 1.5524, + "step": 2529 + }, + { + "epoch": 2.7726027397260276, + "grad_norm": 2.960284948348999, + "learning_rate": 4.5085296506904954e-07, + "loss": 1.5275, + "step": 2530 + }, + { + "epoch": 2.7726027397260276, + "eval_loss": 1.7911875247955322, + "eval_runtime": 37.1342, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 2530 + }, + { + "epoch": 2.773698630136986, + "grad_norm": 3.0453267097473145, + "learning_rate": 4.488220958570268e-07, + "loss": 1.5016, + "step": 2531 + }, + { + "epoch": 2.774794520547945, + "grad_norm": 2.960782766342163, + "learning_rate": 4.467912266450041e-07, + "loss": 1.4835, + "step": 2532 + }, + { + "epoch": 2.775890410958904, + "grad_norm": 3.017273187637329, + "learning_rate": 4.447603574329813e-07, + "loss": 1.4803, + "step": 2533 + }, + { + "epoch": 2.7769863013698632, + "grad_norm": 3.0041613578796387, + "learning_rate": 4.427294882209586e-07, + "loss": 1.5319, + "step": 2534 + }, + { + "epoch": 2.778082191780822, + "grad_norm": 3.0769989490509033, + "learning_rate": 4.4069861900893587e-07, + "loss": 1.4896, + "step": 2535 + }, + { + "epoch": 2.779178082191781, + "grad_norm": 3.212188243865967, + "learning_rate": 4.386677497969131e-07, + "loss": 1.5266, + "step": 2536 + }, + { + "epoch": 2.78027397260274, + "grad_norm": 3.116922378540039, + "learning_rate": 4.3663688058489037e-07, + "loss": 1.6035, + "step": 2537 + }, + { + "epoch": 2.7813698630136985, + "grad_norm": 3.0492711067199707, + "learning_rate": 4.3460601137286764e-07, + "loss": 1.533, + "step": 2538 + }, + { + "epoch": 2.7824657534246575, + "grad_norm": 3.095778703689575, + "learning_rate": 4.325751421608449e-07, + "loss": 1.5406, + "step": 2539 + }, + { + "epoch": 2.7835616438356166, + "grad_norm": 3.1188480854034424, + "learning_rate": 4.3054427294882214e-07, + "loss": 1.5195, + "step": 2540 + }, + { + "epoch": 2.7835616438356166, + "eval_loss": 1.7925901412963867, + "eval_runtime": 37.142, + "eval_samples_per_second": 26.924, + "eval_steps_per_second": 1.131, + "step": 2540 + }, + { + "epoch": 2.784657534246575, + "grad_norm": 3.128167152404785, + "learning_rate": 4.285134037367994e-07, + "loss": 1.5596, + "step": 2541 + }, + { + "epoch": 2.785753424657534, + "grad_norm": 3.066786766052246, + "learning_rate": 4.264825345247767e-07, + "loss": 1.5076, + "step": 2542 + }, + { + "epoch": 2.786849315068493, + "grad_norm": 2.929079532623291, + "learning_rate": 4.2445166531275386e-07, + "loss": 1.455, + "step": 2543 + }, + { + "epoch": 2.787945205479452, + "grad_norm": 2.9974899291992188, + "learning_rate": 4.2242079610073114e-07, + "loss": 1.4632, + "step": 2544 + }, + { + "epoch": 2.789041095890411, + "grad_norm": 2.837831974029541, + "learning_rate": 4.203899268887084e-07, + "loss": 1.4636, + "step": 2545 + }, + { + "epoch": 2.79013698630137, + "grad_norm": 3.127777338027954, + "learning_rate": 4.1835905767668564e-07, + "loss": 1.5818, + "step": 2546 + }, + { + "epoch": 2.791232876712329, + "grad_norm": 3.158036231994629, + "learning_rate": 4.163281884646629e-07, + "loss": 1.5878, + "step": 2547 + }, + { + "epoch": 2.792328767123288, + "grad_norm": 2.9551289081573486, + "learning_rate": 4.142973192526402e-07, + "loss": 1.5212, + "step": 2548 + }, + { + "epoch": 2.7934246575342465, + "grad_norm": 3.065821886062622, + "learning_rate": 4.122664500406174e-07, + "loss": 1.5345, + "step": 2549 + }, + { + "epoch": 2.7945205479452055, + "grad_norm": 3.1351091861724854, + "learning_rate": 4.102355808285947e-07, + "loss": 1.5261, + "step": 2550 + }, + { + "epoch": 2.7945205479452055, + "eval_loss": 1.7908694744110107, + "eval_runtime": 37.1662, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 2550 + }, + { + "epoch": 2.7956164383561646, + "grad_norm": 3.016324996948242, + "learning_rate": 4.0820471161657196e-07, + "loss": 1.4744, + "step": 2551 + }, + { + "epoch": 2.796712328767123, + "grad_norm": 3.124586343765259, + "learning_rate": 4.0617384240454913e-07, + "loss": 1.5254, + "step": 2552 + }, + { + "epoch": 2.797808219178082, + "grad_norm": 3.0363805294036865, + "learning_rate": 4.041429731925264e-07, + "loss": 1.5493, + "step": 2553 + }, + { + "epoch": 2.7989041095890412, + "grad_norm": 2.9451563358306885, + "learning_rate": 4.021121039805037e-07, + "loss": 1.4651, + "step": 2554 + }, + { + "epoch": 2.8, + "grad_norm": 3.054746150970459, + "learning_rate": 4.000812347684809e-07, + "loss": 1.4819, + "step": 2555 + }, + { + "epoch": 2.801095890410959, + "grad_norm": 3.0478272438049316, + "learning_rate": 3.980503655564582e-07, + "loss": 1.5043, + "step": 2556 + }, + { + "epoch": 2.802191780821918, + "grad_norm": 3.0467426776885986, + "learning_rate": 3.9601949634443546e-07, + "loss": 1.5102, + "step": 2557 + }, + { + "epoch": 2.8032876712328765, + "grad_norm": 3.003763437271118, + "learning_rate": 3.939886271324127e-07, + "loss": 1.544, + "step": 2558 + }, + { + "epoch": 2.8043835616438355, + "grad_norm": 3.1108367443084717, + "learning_rate": 3.9195775792038996e-07, + "loss": 1.4968, + "step": 2559 + }, + { + "epoch": 2.8054794520547945, + "grad_norm": 3.112025737762451, + "learning_rate": 3.8992688870836723e-07, + "loss": 1.5567, + "step": 2560 + }, + { + "epoch": 2.8054794520547945, + "eval_loss": 1.7925978899002075, + "eval_runtime": 37.1639, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 2560 + }, + { + "epoch": 2.8065753424657536, + "grad_norm": 3.211944818496704, + "learning_rate": 3.8789601949634446e-07, + "loss": 1.5392, + "step": 2561 + }, + { + "epoch": 2.807671232876712, + "grad_norm": 3.160818099975586, + "learning_rate": 3.8586515028432173e-07, + "loss": 1.515, + "step": 2562 + }, + { + "epoch": 2.808767123287671, + "grad_norm": 3.0587360858917236, + "learning_rate": 3.83834281072299e-07, + "loss": 1.4292, + "step": 2563 + }, + { + "epoch": 2.8098630136986302, + "grad_norm": 3.024881601333618, + "learning_rate": 3.818034118602762e-07, + "loss": 1.5086, + "step": 2564 + }, + { + "epoch": 2.8109589041095893, + "grad_norm": 3.0528507232666016, + "learning_rate": 3.7977254264825345e-07, + "loss": 1.4853, + "step": 2565 + }, + { + "epoch": 2.812054794520548, + "grad_norm": 3.2083396911621094, + "learning_rate": 3.7774167343623073e-07, + "loss": 1.5042, + "step": 2566 + }, + { + "epoch": 2.813150684931507, + "grad_norm": 3.041503429412842, + "learning_rate": 3.75710804224208e-07, + "loss": 1.5243, + "step": 2567 + }, + { + "epoch": 2.814246575342466, + "grad_norm": 2.916165351867676, + "learning_rate": 3.7367993501218523e-07, + "loss": 1.5611, + "step": 2568 + }, + { + "epoch": 2.8153424657534245, + "grad_norm": 3.1837453842163086, + "learning_rate": 3.716490658001625e-07, + "loss": 1.4937, + "step": 2569 + }, + { + "epoch": 2.8164383561643835, + "grad_norm": 2.9144062995910645, + "learning_rate": 3.696181965881398e-07, + "loss": 1.5002, + "step": 2570 + }, + { + "epoch": 2.8164383561643835, + "eval_loss": 1.7910205125808716, + "eval_runtime": 37.1572, + "eval_samples_per_second": 26.913, + "eval_steps_per_second": 1.13, + "step": 2570 + }, + { + "epoch": 2.8175342465753426, + "grad_norm": 2.890509843826294, + "learning_rate": 3.67587327376117e-07, + "loss": 1.4698, + "step": 2571 + }, + { + "epoch": 2.818630136986301, + "grad_norm": 3.0228216648101807, + "learning_rate": 3.655564581640943e-07, + "loss": 1.524, + "step": 2572 + }, + { + "epoch": 2.81972602739726, + "grad_norm": 3.0013716220855713, + "learning_rate": 3.6352558895207155e-07, + "loss": 1.5113, + "step": 2573 + }, + { + "epoch": 2.8208219178082192, + "grad_norm": 3.0757668018341064, + "learning_rate": 3.614947197400488e-07, + "loss": 1.4841, + "step": 2574 + }, + { + "epoch": 2.821917808219178, + "grad_norm": 3.0872232913970947, + "learning_rate": 3.5946385052802605e-07, + "loss": 1.5261, + "step": 2575 + }, + { + "epoch": 2.823013698630137, + "grad_norm": 3.1712300777435303, + "learning_rate": 3.5743298131600333e-07, + "loss": 1.5361, + "step": 2576 + }, + { + "epoch": 2.824109589041096, + "grad_norm": 3.080634593963623, + "learning_rate": 3.554021121039805e-07, + "loss": 1.5013, + "step": 2577 + }, + { + "epoch": 2.825205479452055, + "grad_norm": 3.1428229808807373, + "learning_rate": 3.533712428919578e-07, + "loss": 1.5609, + "step": 2578 + }, + { + "epoch": 2.826301369863014, + "grad_norm": 3.0235114097595215, + "learning_rate": 3.5134037367993505e-07, + "loss": 1.5111, + "step": 2579 + }, + { + "epoch": 2.8273972602739725, + "grad_norm": 3.019428253173828, + "learning_rate": 3.4930950446791227e-07, + "loss": 1.5164, + "step": 2580 + }, + { + "epoch": 2.8273972602739725, + "eval_loss": 1.7917933464050293, + "eval_runtime": 37.1652, + "eval_samples_per_second": 26.907, + "eval_steps_per_second": 1.13, + "step": 2580 + }, + { + "epoch": 2.8284931506849316, + "grad_norm": 2.9536592960357666, + "learning_rate": 3.4727863525588955e-07, + "loss": 1.4759, + "step": 2581 + }, + { + "epoch": 2.8295890410958906, + "grad_norm": 3.0202090740203857, + "learning_rate": 3.452477660438668e-07, + "loss": 1.5058, + "step": 2582 + }, + { + "epoch": 2.830684931506849, + "grad_norm": 3.128509759902954, + "learning_rate": 3.4321689683184405e-07, + "loss": 1.5609, + "step": 2583 + }, + { + "epoch": 2.831780821917808, + "grad_norm": 3.0520644187927246, + "learning_rate": 3.411860276198213e-07, + "loss": 1.5206, + "step": 2584 + }, + { + "epoch": 2.8328767123287673, + "grad_norm": 3.075646162033081, + "learning_rate": 3.391551584077986e-07, + "loss": 1.4684, + "step": 2585 + }, + { + "epoch": 2.833972602739726, + "grad_norm": 3.0036284923553467, + "learning_rate": 3.3712428919577577e-07, + "loss": 1.522, + "step": 2586 + }, + { + "epoch": 2.835068493150685, + "grad_norm": 3.076817512512207, + "learning_rate": 3.3509341998375304e-07, + "loss": 1.5147, + "step": 2587 + }, + { + "epoch": 2.836164383561644, + "grad_norm": 3.0329620838165283, + "learning_rate": 3.330625507717303e-07, + "loss": 1.4853, + "step": 2588 + }, + { + "epoch": 2.8372602739726025, + "grad_norm": 3.1748528480529785, + "learning_rate": 3.3103168155970754e-07, + "loss": 1.5139, + "step": 2589 + }, + { + "epoch": 2.8383561643835615, + "grad_norm": 3.1956450939178467, + "learning_rate": 3.290008123476848e-07, + "loss": 1.5641, + "step": 2590 + }, + { + "epoch": 2.8383561643835615, + "eval_loss": 1.7919085025787354, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 2590 + }, + { + "epoch": 2.8394520547945206, + "grad_norm": 3.2937188148498535, + "learning_rate": 3.269699431356621e-07, + "loss": 1.5882, + "step": 2591 + }, + { + "epoch": 2.8405479452054796, + "grad_norm": 3.1744468212127686, + "learning_rate": 3.249390739236393e-07, + "loss": 1.5296, + "step": 2592 + }, + { + "epoch": 2.8416438356164386, + "grad_norm": 3.053405284881592, + "learning_rate": 3.229082047116166e-07, + "loss": 1.4622, + "step": 2593 + }, + { + "epoch": 2.842739726027397, + "grad_norm": 3.268449544906616, + "learning_rate": 3.2087733549959387e-07, + "loss": 1.5696, + "step": 2594 + }, + { + "epoch": 2.8438356164383563, + "grad_norm": 3.144965887069702, + "learning_rate": 3.1884646628757114e-07, + "loss": 1.5724, + "step": 2595 + }, + { + "epoch": 2.8449315068493153, + "grad_norm": 2.952338457107544, + "learning_rate": 3.1681559707554837e-07, + "loss": 1.5392, + "step": 2596 + }, + { + "epoch": 2.846027397260274, + "grad_norm": 3.4757699966430664, + "learning_rate": 3.1478472786352564e-07, + "loss": 1.552, + "step": 2597 + }, + { + "epoch": 2.847123287671233, + "grad_norm": 2.991934061050415, + "learning_rate": 3.127538586515029e-07, + "loss": 1.4903, + "step": 2598 + }, + { + "epoch": 2.848219178082192, + "grad_norm": 3.0654208660125732, + "learning_rate": 3.107229894394801e-07, + "loss": 1.5247, + "step": 2599 + }, + { + "epoch": 2.8493150684931505, + "grad_norm": 3.049363374710083, + "learning_rate": 3.0869212022745736e-07, + "loss": 1.5061, + "step": 2600 + }, + { + "epoch": 2.8493150684931505, + "eval_loss": 1.7906988859176636, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 2600 + }, + { + "epoch": 2.8504109589041096, + "grad_norm": 3.009659767150879, + "learning_rate": 3.0666125101543464e-07, + "loss": 1.4634, + "step": 2601 + }, + { + "epoch": 2.8515068493150686, + "grad_norm": 3.0317633152008057, + "learning_rate": 3.046303818034119e-07, + "loss": 1.5197, + "step": 2602 + }, + { + "epoch": 2.852602739726027, + "grad_norm": 3.139240264892578, + "learning_rate": 3.0259951259138914e-07, + "loss": 1.5799, + "step": 2603 + }, + { + "epoch": 2.853698630136986, + "grad_norm": 3.1429147720336914, + "learning_rate": 3.0056864337936636e-07, + "loss": 1.5231, + "step": 2604 + }, + { + "epoch": 2.8547945205479452, + "grad_norm": 2.8958003520965576, + "learning_rate": 2.9853777416734364e-07, + "loss": 1.4091, + "step": 2605 + }, + { + "epoch": 2.855890410958904, + "grad_norm": 3.0315115451812744, + "learning_rate": 2.965069049553209e-07, + "loss": 1.4946, + "step": 2606 + }, + { + "epoch": 2.856986301369863, + "grad_norm": 2.97930645942688, + "learning_rate": 2.9447603574329814e-07, + "loss": 1.5335, + "step": 2607 + }, + { + "epoch": 2.858082191780822, + "grad_norm": 2.9786713123321533, + "learning_rate": 2.924451665312754e-07, + "loss": 1.4754, + "step": 2608 + }, + { + "epoch": 2.859178082191781, + "grad_norm": 2.867501735687256, + "learning_rate": 2.904142973192527e-07, + "loss": 1.4864, + "step": 2609 + }, + { + "epoch": 2.86027397260274, + "grad_norm": 2.9504475593566895, + "learning_rate": 2.883834281072299e-07, + "loss": 1.4219, + "step": 2610 + }, + { + "epoch": 2.86027397260274, + "eval_loss": 1.791516900062561, + "eval_runtime": 36.775, + "eval_samples_per_second": 27.192, + "eval_steps_per_second": 1.142, + "step": 2610 + }, + { + "epoch": 2.8613698630136986, + "grad_norm": 3.0030910968780518, + "learning_rate": 2.863525588952072e-07, + "loss": 1.497, + "step": 2611 + }, + { + "epoch": 2.8624657534246576, + "grad_norm": 3.0113725662231445, + "learning_rate": 2.843216896831844e-07, + "loss": 1.4988, + "step": 2612 + }, + { + "epoch": 2.8635616438356166, + "grad_norm": 3.03090763092041, + "learning_rate": 2.822908204711617e-07, + "loss": 1.4446, + "step": 2613 + }, + { + "epoch": 2.864657534246575, + "grad_norm": 3.162860631942749, + "learning_rate": 2.8025995125913896e-07, + "loss": 1.5284, + "step": 2614 + }, + { + "epoch": 2.8657534246575342, + "grad_norm": 3.04756236076355, + "learning_rate": 2.782290820471162e-07, + "loss": 1.4982, + "step": 2615 + }, + { + "epoch": 2.8668493150684933, + "grad_norm": 3.243168830871582, + "learning_rate": 2.761982128350934e-07, + "loss": 1.5532, + "step": 2616 + }, + { + "epoch": 2.867945205479452, + "grad_norm": 3.218254327774048, + "learning_rate": 2.741673436230707e-07, + "loss": 1.5133, + "step": 2617 + }, + { + "epoch": 2.869041095890411, + "grad_norm": 3.4805715084075928, + "learning_rate": 2.7213647441104796e-07, + "loss": 1.5435, + "step": 2618 + }, + { + "epoch": 2.87013698630137, + "grad_norm": 3.14292311668396, + "learning_rate": 2.7010560519902523e-07, + "loss": 1.5228, + "step": 2619 + }, + { + "epoch": 2.8712328767123285, + "grad_norm": 2.990907907485962, + "learning_rate": 2.6807473598700246e-07, + "loss": 1.4448, + "step": 2620 + }, + { + "epoch": 2.8712328767123285, + "eval_loss": 1.791623830795288, + "eval_runtime": 37.0152, + "eval_samples_per_second": 27.016, + "eval_steps_per_second": 1.135, + "step": 2620 + }, + { + "epoch": 2.8723287671232876, + "grad_norm": 3.1701385974884033, + "learning_rate": 2.660438667749797e-07, + "loss": 1.5142, + "step": 2621 + }, + { + "epoch": 2.8734246575342466, + "grad_norm": 2.9661409854888916, + "learning_rate": 2.6401299756295696e-07, + "loss": 1.4705, + "step": 2622 + }, + { + "epoch": 2.8745205479452056, + "grad_norm": 3.066345453262329, + "learning_rate": 2.6198212835093423e-07, + "loss": 1.503, + "step": 2623 + }, + { + "epoch": 2.8756164383561647, + "grad_norm": 3.003727912902832, + "learning_rate": 2.5995125913891145e-07, + "loss": 1.5003, + "step": 2624 + }, + { + "epoch": 2.8767123287671232, + "grad_norm": 3.1008756160736084, + "learning_rate": 2.5792038992688873e-07, + "loss": 1.4562, + "step": 2625 + }, + { + "epoch": 2.8778082191780823, + "grad_norm": 2.978329658508301, + "learning_rate": 2.55889520714866e-07, + "loss": 1.4794, + "step": 2626 + }, + { + "epoch": 2.8789041095890413, + "grad_norm": 3.0397987365722656, + "learning_rate": 2.5385865150284323e-07, + "loss": 1.5283, + "step": 2627 + }, + { + "epoch": 2.88, + "grad_norm": 3.0798213481903076, + "learning_rate": 2.518277822908205e-07, + "loss": 1.5293, + "step": 2628 + }, + { + "epoch": 2.881095890410959, + "grad_norm": 3.052346706390381, + "learning_rate": 2.4979691307879773e-07, + "loss": 1.5243, + "step": 2629 + }, + { + "epoch": 2.882191780821918, + "grad_norm": 3.0079402923583984, + "learning_rate": 2.47766043866775e-07, + "loss": 1.5363, + "step": 2630 + }, + { + "epoch": 2.882191780821918, + "eval_loss": 1.7894965410232544, + "eval_runtime": 37.135, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 2630 + }, + { + "epoch": 2.8832876712328765, + "grad_norm": 2.92486572265625, + "learning_rate": 2.457351746547523e-07, + "loss": 1.4625, + "step": 2631 + }, + { + "epoch": 2.8843835616438356, + "grad_norm": 2.9858713150024414, + "learning_rate": 2.437043054427295e-07, + "loss": 1.4548, + "step": 2632 + }, + { + "epoch": 2.8854794520547946, + "grad_norm": 3.1012473106384277, + "learning_rate": 2.416734362307068e-07, + "loss": 1.5375, + "step": 2633 + }, + { + "epoch": 2.886575342465753, + "grad_norm": 3.0200417041778564, + "learning_rate": 2.39642567018684e-07, + "loss": 1.52, + "step": 2634 + }, + { + "epoch": 2.8876712328767122, + "grad_norm": 3.1536433696746826, + "learning_rate": 2.3761169780666125e-07, + "loss": 1.4482, + "step": 2635 + }, + { + "epoch": 2.8887671232876713, + "grad_norm": 3.0201385021209717, + "learning_rate": 2.3558082859463852e-07, + "loss": 1.5278, + "step": 2636 + }, + { + "epoch": 2.8898630136986303, + "grad_norm": 2.9468398094177246, + "learning_rate": 2.3354995938261577e-07, + "loss": 1.488, + "step": 2637 + }, + { + "epoch": 2.890958904109589, + "grad_norm": 3.0052077770233154, + "learning_rate": 2.3151909017059302e-07, + "loss": 1.4361, + "step": 2638 + }, + { + "epoch": 2.892054794520548, + "grad_norm": 3.0999598503112793, + "learning_rate": 2.294882209585703e-07, + "loss": 1.5549, + "step": 2639 + }, + { + "epoch": 2.893150684931507, + "grad_norm": 3.097472667694092, + "learning_rate": 2.2745735174654755e-07, + "loss": 1.5241, + "step": 2640 + }, + { + "epoch": 2.893150684931507, + "eval_loss": 1.7891727685928345, + "eval_runtime": 37.1783, + "eval_samples_per_second": 26.897, + "eval_steps_per_second": 1.13, + "step": 2640 + }, + { + "epoch": 2.894246575342466, + "grad_norm": 3.031191110610962, + "learning_rate": 2.2542648253452477e-07, + "loss": 1.4707, + "step": 2641 + }, + { + "epoch": 2.8953424657534246, + "grad_norm": 2.9850590229034424, + "learning_rate": 2.2339561332250205e-07, + "loss": 1.5065, + "step": 2642 + }, + { + "epoch": 2.8964383561643836, + "grad_norm": 3.2836239337921143, + "learning_rate": 2.213647441104793e-07, + "loss": 1.5513, + "step": 2643 + }, + { + "epoch": 2.8975342465753426, + "grad_norm": 2.9327352046966553, + "learning_rate": 2.1933387489845655e-07, + "loss": 1.5419, + "step": 2644 + }, + { + "epoch": 2.8986301369863012, + "grad_norm": 2.9825339317321777, + "learning_rate": 2.1730300568643382e-07, + "loss": 1.5028, + "step": 2645 + }, + { + "epoch": 2.8997260273972603, + "grad_norm": 3.1225833892822266, + "learning_rate": 2.1527213647441107e-07, + "loss": 1.4981, + "step": 2646 + }, + { + "epoch": 2.9008219178082193, + "grad_norm": 2.9758846759796143, + "learning_rate": 2.1324126726238835e-07, + "loss": 1.531, + "step": 2647 + }, + { + "epoch": 2.901917808219178, + "grad_norm": 3.020021915435791, + "learning_rate": 2.1121039805036557e-07, + "loss": 1.5065, + "step": 2648 + }, + { + "epoch": 2.903013698630137, + "grad_norm": 3.11961030960083, + "learning_rate": 2.0917952883834282e-07, + "loss": 1.5208, + "step": 2649 + }, + { + "epoch": 2.904109589041096, + "grad_norm": 3.110372304916382, + "learning_rate": 2.071486596263201e-07, + "loss": 1.5579, + "step": 2650 + }, + { + "epoch": 2.904109589041096, + "eval_loss": 1.789390206336975, + "eval_runtime": 37.1767, + "eval_samples_per_second": 26.899, + "eval_steps_per_second": 1.13, + "step": 2650 + }, + { + "epoch": 2.9052054794520545, + "grad_norm": 2.935211658477783, + "learning_rate": 2.0511779041429734e-07, + "loss": 1.4802, + "step": 2651 + }, + { + "epoch": 2.9063013698630136, + "grad_norm": 3.1486573219299316, + "learning_rate": 2.0308692120227457e-07, + "loss": 1.4791, + "step": 2652 + }, + { + "epoch": 2.9073972602739726, + "grad_norm": 2.9627110958099365, + "learning_rate": 2.0105605199025184e-07, + "loss": 1.5436, + "step": 2653 + }, + { + "epoch": 2.9084931506849316, + "grad_norm": 3.106480836868286, + "learning_rate": 1.990251827782291e-07, + "loss": 1.5475, + "step": 2654 + }, + { + "epoch": 2.9095890410958907, + "grad_norm": 2.9908578395843506, + "learning_rate": 1.9699431356620634e-07, + "loss": 1.5735, + "step": 2655 + }, + { + "epoch": 2.9106849315068493, + "grad_norm": 3.0215280055999756, + "learning_rate": 1.9496344435418362e-07, + "loss": 1.4571, + "step": 2656 + }, + { + "epoch": 2.9117808219178083, + "grad_norm": 3.046536684036255, + "learning_rate": 1.9293257514216087e-07, + "loss": 1.5105, + "step": 2657 + }, + { + "epoch": 2.9128767123287673, + "grad_norm": 3.1733031272888184, + "learning_rate": 1.909017059301381e-07, + "loss": 1.4974, + "step": 2658 + }, + { + "epoch": 2.913972602739726, + "grad_norm": 3.0829505920410156, + "learning_rate": 1.8887083671811536e-07, + "loss": 1.4964, + "step": 2659 + }, + { + "epoch": 2.915068493150685, + "grad_norm": 2.9070191383361816, + "learning_rate": 1.8683996750609261e-07, + "loss": 1.4919, + "step": 2660 + }, + { + "epoch": 2.915068493150685, + "eval_loss": 1.7900432348251343, + "eval_runtime": 37.172, + "eval_samples_per_second": 26.902, + "eval_steps_per_second": 1.13, + "step": 2660 + }, + { + "epoch": 2.916164383561644, + "grad_norm": 3.207252264022827, + "learning_rate": 1.848090982940699e-07, + "loss": 1.5279, + "step": 2661 + }, + { + "epoch": 2.9172602739726026, + "grad_norm": 3.038796901702881, + "learning_rate": 1.8277822908204714e-07, + "loss": 1.5855, + "step": 2662 + }, + { + "epoch": 2.9183561643835616, + "grad_norm": 3.0879733562469482, + "learning_rate": 1.807473598700244e-07, + "loss": 1.4994, + "step": 2663 + }, + { + "epoch": 2.9194520547945206, + "grad_norm": 3.070927143096924, + "learning_rate": 1.7871649065800166e-07, + "loss": 1.5297, + "step": 2664 + }, + { + "epoch": 2.9205479452054792, + "grad_norm": 3.2004566192626953, + "learning_rate": 1.766856214459789e-07, + "loss": 1.6152, + "step": 2665 + }, + { + "epoch": 2.9216438356164383, + "grad_norm": 3.0775976181030273, + "learning_rate": 1.7465475223395614e-07, + "loss": 1.531, + "step": 2666 + }, + { + "epoch": 2.9227397260273973, + "grad_norm": 3.0248641967773438, + "learning_rate": 1.726238830219334e-07, + "loss": 1.5075, + "step": 2667 + }, + { + "epoch": 2.9238356164383563, + "grad_norm": 2.9836952686309814, + "learning_rate": 1.7059301380991066e-07, + "loss": 1.5113, + "step": 2668 + }, + { + "epoch": 2.924931506849315, + "grad_norm": 3.0869314670562744, + "learning_rate": 1.6856214459788788e-07, + "loss": 1.4939, + "step": 2669 + }, + { + "epoch": 2.926027397260274, + "grad_norm": 3.080989122390747, + "learning_rate": 1.6653127538586516e-07, + "loss": 1.4619, + "step": 2670 + }, + { + "epoch": 2.926027397260274, + "eval_loss": 1.7903239727020264, + "eval_runtime": 37.1883, + "eval_samples_per_second": 26.89, + "eval_steps_per_second": 1.129, + "step": 2670 + }, + { + "epoch": 2.927123287671233, + "grad_norm": 3.0589723587036133, + "learning_rate": 1.645004061738424e-07, + "loss": 1.5439, + "step": 2671 + }, + { + "epoch": 2.928219178082192, + "grad_norm": 3.051922082901001, + "learning_rate": 1.6246953696181966e-07, + "loss": 1.4704, + "step": 2672 + }, + { + "epoch": 2.9293150684931506, + "grad_norm": 2.993330478668213, + "learning_rate": 1.6043866774979693e-07, + "loss": 1.5143, + "step": 2673 + }, + { + "epoch": 2.9304109589041096, + "grad_norm": 3.135922908782959, + "learning_rate": 1.5840779853777418e-07, + "loss": 1.5586, + "step": 2674 + }, + { + "epoch": 2.9315068493150687, + "grad_norm": 3.115062713623047, + "learning_rate": 1.5637692932575146e-07, + "loss": 1.5121, + "step": 2675 + }, + { + "epoch": 2.9326027397260273, + "grad_norm": 3.1141114234924316, + "learning_rate": 1.5434606011372868e-07, + "loss": 1.495, + "step": 2676 + }, + { + "epoch": 2.9336986301369863, + "grad_norm": 3.0367932319641113, + "learning_rate": 1.5231519090170596e-07, + "loss": 1.5532, + "step": 2677 + }, + { + "epoch": 2.9347945205479453, + "grad_norm": 3.059889316558838, + "learning_rate": 1.5028432168968318e-07, + "loss": 1.5488, + "step": 2678 + }, + { + "epoch": 2.935890410958904, + "grad_norm": 3.003568649291992, + "learning_rate": 1.4825345247766046e-07, + "loss": 1.5268, + "step": 2679 + }, + { + "epoch": 2.936986301369863, + "grad_norm": 3.139242649078369, + "learning_rate": 1.462225832656377e-07, + "loss": 1.5565, + "step": 2680 + }, + { + "epoch": 2.936986301369863, + "eval_loss": 1.7899202108383179, + "eval_runtime": 37.2012, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 2680 + }, + { + "epoch": 2.938082191780822, + "grad_norm": 3.0351994037628174, + "learning_rate": 1.4419171405361496e-07, + "loss": 1.4995, + "step": 2681 + }, + { + "epoch": 2.9391780821917806, + "grad_norm": 3.088860034942627, + "learning_rate": 1.421608448415922e-07, + "loss": 1.4923, + "step": 2682 + }, + { + "epoch": 2.9402739726027396, + "grad_norm": 3.014510154724121, + "learning_rate": 1.4012997562956948e-07, + "loss": 1.538, + "step": 2683 + }, + { + "epoch": 2.9413698630136986, + "grad_norm": 3.2081291675567627, + "learning_rate": 1.380991064175467e-07, + "loss": 1.5398, + "step": 2684 + }, + { + "epoch": 2.9424657534246577, + "grad_norm": 3.107496500015259, + "learning_rate": 1.3606823720552398e-07, + "loss": 1.5569, + "step": 2685 + }, + { + "epoch": 2.9435616438356167, + "grad_norm": 3.365082263946533, + "learning_rate": 1.3403736799350123e-07, + "loss": 1.5471, + "step": 2686 + }, + { + "epoch": 2.9446575342465753, + "grad_norm": 3.104659080505371, + "learning_rate": 1.3200649878147848e-07, + "loss": 1.5063, + "step": 2687 + }, + { + "epoch": 2.9457534246575343, + "grad_norm": 3.0596365928649902, + "learning_rate": 1.2997562956945573e-07, + "loss": 1.5267, + "step": 2688 + }, + { + "epoch": 2.9468493150684933, + "grad_norm": 3.113130569458008, + "learning_rate": 1.27944760357433e-07, + "loss": 1.5616, + "step": 2689 + }, + { + "epoch": 2.947945205479452, + "grad_norm": 3.1505556106567383, + "learning_rate": 1.2591389114541025e-07, + "loss": 1.53, + "step": 2690 + }, + { + "epoch": 2.947945205479452, + "eval_loss": 1.789366364479065, + "eval_runtime": 37.2184, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 2690 + }, + { + "epoch": 2.949041095890411, + "grad_norm": 3.0513687133789062, + "learning_rate": 1.238830219333875e-07, + "loss": 1.4936, + "step": 2691 + }, + { + "epoch": 2.95013698630137, + "grad_norm": 2.9915530681610107, + "learning_rate": 1.2185215272136475e-07, + "loss": 1.4704, + "step": 2692 + }, + { + "epoch": 2.9512328767123286, + "grad_norm": 3.0183277130126953, + "learning_rate": 1.19821283509342e-07, + "loss": 1.4708, + "step": 2693 + }, + { + "epoch": 2.9523287671232876, + "grad_norm": 3.1279025077819824, + "learning_rate": 1.1779041429731926e-07, + "loss": 1.5471, + "step": 2694 + }, + { + "epoch": 2.9534246575342467, + "grad_norm": 3.1098203659057617, + "learning_rate": 1.1575954508529651e-07, + "loss": 1.5557, + "step": 2695 + }, + { + "epoch": 2.9545205479452052, + "grad_norm": 3.1258881092071533, + "learning_rate": 1.1372867587327377e-07, + "loss": 1.4938, + "step": 2696 + }, + { + "epoch": 2.9556164383561643, + "grad_norm": 3.079859972000122, + "learning_rate": 1.1169780666125102e-07, + "loss": 1.5093, + "step": 2697 + }, + { + "epoch": 2.9567123287671233, + "grad_norm": 3.268019437789917, + "learning_rate": 1.0966693744922827e-07, + "loss": 1.5581, + "step": 2698 + }, + { + "epoch": 2.9578082191780823, + "grad_norm": 3.02490234375, + "learning_rate": 1.0763606823720554e-07, + "loss": 1.4983, + "step": 2699 + }, + { + "epoch": 2.958904109589041, + "grad_norm": 3.195443630218506, + "learning_rate": 1.0560519902518278e-07, + "loss": 1.5028, + "step": 2700 + }, + { + "epoch": 2.958904109589041, + "eval_loss": 1.7894436120986938, + "eval_runtime": 37.2347, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 2700 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-2700/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/latest new file mode 100644 index 0000000000000000000000000000000000000000..e5bdf58d4f29d34e909da25905fad376f73e7c29 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..88dfdba1e58755a167c167632c8e889a52e5bca8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/trainer_state.json @@ -0,0 +1,3153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.4383561643835616, + "eval_steps": 10, + "global_step": 400, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-400/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fb0f9a53a507ffcdb60410deb71d60ef801bf350 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/generation_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c389b841a9a29ddd904e02d1eb0e08dcce82ad9 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/latest b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/latest new file mode 100644 index 0000000000000000000000000000000000000000..12cae1adf3af8546b4141c6f62261c8e99839a54 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/latest @@ -0,0 +1 @@ +global_step600 \ No newline at end of file diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/model.safetensors.index.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..ed64de846d720b9a7859dc20575fea8e8ca51940 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/special_tokens_map.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..6949c5975ee0e961ef61cf31010dce04df0a03f8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/tokenizer_config.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..5684b14e2756b56aa397c1324e239b442162b34a --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/tokenizer_config.json @@ -0,0 +1,2070 @@ +{ + "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 + } + }, + "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/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/trainer_state.json b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..e48b784ca854d2a073a40e4a43c7d34a7c8b5f18 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/trainer_state.json @@ -0,0 +1,4713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.6575342465753424, + "eval_steps": 10, + "global_step": 600, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001095890410958904, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9228, + "step": 1 + }, + { + "epoch": 0.002191780821917808, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.913, + "step": 2 + }, + { + "epoch": 0.0032876712328767125, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8792, + "step": 3 + }, + { + "epoch": 0.004383561643835616, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9058, + "step": 4 + }, + { + "epoch": 0.005479452054794521, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8737, + "step": 5 + }, + { + "epoch": 0.006575342465753425, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 3.0114, + "step": 6 + }, + { + "epoch": 0.007671232876712329, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9157, + "step": 7 + }, + { + "epoch": 0.008767123287671232, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9441, + "step": 8 + }, + { + "epoch": 0.009863013698630137, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9234, + "step": 9 + }, + { + "epoch": 0.010958904109589041, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9776, + "step": 10 + }, + { + "epoch": 0.010958904109589041, + "eval_loss": 2.9550302028656006, + "eval_runtime": 38.1653, + "eval_samples_per_second": 26.202, + "eval_steps_per_second": 1.1, + "step": 10 + }, + { + "epoch": 0.012054794520547946, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9516, + "step": 11 + }, + { + "epoch": 0.01315068493150685, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.8598, + "step": 12 + }, + { + "epoch": 0.014246575342465753, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9358, + "step": 13 + }, + { + "epoch": 0.015342465753424657, + "grad_norm": 0.0, + "learning_rate": 0.0, + "loss": 2.9103, + "step": 14 + }, + { + "epoch": 0.01643835616438356, + "grad_norm": 4.499549865722656, + "learning_rate": 1.8248175182481753e-08, + "loss": 2.9657, + "step": 15 + }, + { + "epoch": 0.017534246575342465, + "grad_norm": 4.579676151275635, + "learning_rate": 3.6496350364963505e-08, + "loss": 2.9001, + "step": 16 + }, + { + "epoch": 0.01863013698630137, + "grad_norm": 4.628462791442871, + "learning_rate": 5.4744525547445264e-08, + "loss": 2.9116, + "step": 17 + }, + { + "epoch": 0.019726027397260273, + "grad_norm": 4.7823486328125, + "learning_rate": 7.299270072992701e-08, + "loss": 2.9414, + "step": 18 + }, + { + "epoch": 0.020821917808219178, + "grad_norm": 4.603215217590332, + "learning_rate": 9.124087591240876e-08, + "loss": 2.8951, + "step": 19 + }, + { + "epoch": 0.021917808219178082, + "grad_norm": 4.287731647491455, + "learning_rate": 1.0948905109489053e-07, + "loss": 2.9784, + "step": 20 + }, + { + "epoch": 0.021917808219178082, + "eval_loss": 2.9541079998016357, + "eval_runtime": 37.3608, + "eval_samples_per_second": 26.766, + "eval_steps_per_second": 1.124, + "step": 20 + }, + { + "epoch": 0.023013698630136987, + "grad_norm": 4.563065052032471, + "learning_rate": 1.2773722627737227e-07, + "loss": 2.9397, + "step": 21 + }, + { + "epoch": 0.02410958904109589, + "grad_norm": 4.157818794250488, + "learning_rate": 1.4598540145985402e-07, + "loss": 2.8689, + "step": 22 + }, + { + "epoch": 0.025205479452054796, + "grad_norm": 4.302506923675537, + "learning_rate": 1.642335766423358e-07, + "loss": 2.9083, + "step": 23 + }, + { + "epoch": 0.0263013698630137, + "grad_norm": 4.426967620849609, + "learning_rate": 1.8248175182481753e-07, + "loss": 2.9395, + "step": 24 + }, + { + "epoch": 0.0273972602739726, + "grad_norm": 4.171821117401123, + "learning_rate": 2.0072992700729928e-07, + "loss": 2.9514, + "step": 25 + }, + { + "epoch": 0.028493150684931506, + "grad_norm": 4.069268703460693, + "learning_rate": 2.1897810218978106e-07, + "loss": 2.936, + "step": 26 + }, + { + "epoch": 0.02958904109589041, + "grad_norm": 3.878473997116089, + "learning_rate": 2.3722627737226278e-07, + "loss": 2.9444, + "step": 27 + }, + { + "epoch": 0.030684931506849315, + "grad_norm": 3.790377616882324, + "learning_rate": 2.5547445255474454e-07, + "loss": 2.8972, + "step": 28 + }, + { + "epoch": 0.03178082191780822, + "grad_norm": 3.6787664890289307, + "learning_rate": 2.737226277372263e-07, + "loss": 2.8909, + "step": 29 + }, + { + "epoch": 0.03287671232876712, + "grad_norm": 3.9471073150634766, + "learning_rate": 2.9197080291970804e-07, + "loss": 2.862, + "step": 30 + }, + { + "epoch": 0.03287671232876712, + "eval_loss": 2.918577194213867, + "eval_runtime": 37.1005, + "eval_samples_per_second": 26.954, + "eval_steps_per_second": 1.132, + "step": 30 + }, + { + "epoch": 0.03397260273972603, + "grad_norm": 3.1942155361175537, + "learning_rate": 3.102189781021898e-07, + "loss": 2.88, + "step": 31 + }, + { + "epoch": 0.03506849315068493, + "grad_norm": 3.4512293338775635, + "learning_rate": 3.284671532846716e-07, + "loss": 2.8781, + "step": 32 + }, + { + "epoch": 0.03616438356164384, + "grad_norm": 3.199559450149536, + "learning_rate": 3.467153284671533e-07, + "loss": 2.8464, + "step": 33 + }, + { + "epoch": 0.03726027397260274, + "grad_norm": 3.15679931640625, + "learning_rate": 3.6496350364963505e-07, + "loss": 2.8377, + "step": 34 + }, + { + "epoch": 0.038356164383561646, + "grad_norm": 3.015956163406372, + "learning_rate": 3.8321167883211686e-07, + "loss": 2.8638, + "step": 35 + }, + { + "epoch": 0.03945205479452055, + "grad_norm": 2.8539278507232666, + "learning_rate": 4.0145985401459856e-07, + "loss": 2.8819, + "step": 36 + }, + { + "epoch": 0.040547945205479455, + "grad_norm": 2.819194793701172, + "learning_rate": 4.197080291970803e-07, + "loss": 2.8232, + "step": 37 + }, + { + "epoch": 0.041643835616438356, + "grad_norm": 2.6231954097747803, + "learning_rate": 4.379562043795621e-07, + "loss": 2.783, + "step": 38 + }, + { + "epoch": 0.042739726027397264, + "grad_norm": 2.633301258087158, + "learning_rate": 4.562043795620438e-07, + "loss": 2.8905, + "step": 39 + }, + { + "epoch": 0.043835616438356165, + "grad_norm": 2.5322017669677734, + "learning_rate": 4.7445255474452557e-07, + "loss": 2.7851, + "step": 40 + }, + { + "epoch": 0.043835616438356165, + "eval_loss": 2.8536036014556885, + "eval_runtime": 37.1542, + "eval_samples_per_second": 26.915, + "eval_steps_per_second": 1.13, + "step": 40 + }, + { + "epoch": 0.044931506849315066, + "grad_norm": 2.608207941055298, + "learning_rate": 4.927007299270073e-07, + "loss": 2.82, + "step": 41 + }, + { + "epoch": 0.046027397260273974, + "grad_norm": 2.4030039310455322, + "learning_rate": 5.109489051094891e-07, + "loss": 2.8473, + "step": 42 + }, + { + "epoch": 0.047123287671232875, + "grad_norm": 2.4047887325286865, + "learning_rate": 5.291970802919708e-07, + "loss": 2.8239, + "step": 43 + }, + { + "epoch": 0.04821917808219178, + "grad_norm": 2.3790647983551025, + "learning_rate": 5.474452554744526e-07, + "loss": 2.8342, + "step": 44 + }, + { + "epoch": 0.049315068493150684, + "grad_norm": 2.2825446128845215, + "learning_rate": 5.656934306569343e-07, + "loss": 2.8208, + "step": 45 + }, + { + "epoch": 0.05041095890410959, + "grad_norm": 2.2430613040924072, + "learning_rate": 5.839416058394161e-07, + "loss": 2.7658, + "step": 46 + }, + { + "epoch": 0.05150684931506849, + "grad_norm": 2.2379651069641113, + "learning_rate": 6.021897810218979e-07, + "loss": 2.7467, + "step": 47 + }, + { + "epoch": 0.0526027397260274, + "grad_norm": 2.1349596977233887, + "learning_rate": 6.204379562043796e-07, + "loss": 2.7785, + "step": 48 + }, + { + "epoch": 0.0536986301369863, + "grad_norm": 1.9866257905960083, + "learning_rate": 6.386861313868614e-07, + "loss": 2.7127, + "step": 49 + }, + { + "epoch": 0.0547945205479452, + "grad_norm": 2.013187885284424, + "learning_rate": 6.569343065693432e-07, + "loss": 2.8342, + "step": 50 + }, + { + "epoch": 0.0547945205479452, + "eval_loss": 2.779691457748413, + "eval_runtime": 37.2292, + "eval_samples_per_second": 26.861, + "eval_steps_per_second": 1.128, + "step": 50 + }, + { + "epoch": 0.05589041095890411, + "grad_norm": 2.3823800086975098, + "learning_rate": 6.751824817518248e-07, + "loss": 2.7202, + "step": 51 + }, + { + "epoch": 0.05698630136986301, + "grad_norm": 1.9252759218215942, + "learning_rate": 6.934306569343066e-07, + "loss": 2.7472, + "step": 52 + }, + { + "epoch": 0.05808219178082192, + "grad_norm": 1.8134092092514038, + "learning_rate": 7.116788321167883e-07, + "loss": 2.6801, + "step": 53 + }, + { + "epoch": 0.05917808219178082, + "grad_norm": 1.7787110805511475, + "learning_rate": 7.299270072992701e-07, + "loss": 2.6931, + "step": 54 + }, + { + "epoch": 0.06027397260273973, + "grad_norm": 1.8264989852905273, + "learning_rate": 7.48175182481752e-07, + "loss": 2.7037, + "step": 55 + }, + { + "epoch": 0.06136986301369863, + "grad_norm": 1.779778242111206, + "learning_rate": 7.664233576642337e-07, + "loss": 2.7206, + "step": 56 + }, + { + "epoch": 0.06246575342465754, + "grad_norm": 1.9083834886550903, + "learning_rate": 7.846715328467154e-07, + "loss": 2.7215, + "step": 57 + }, + { + "epoch": 0.06356164383561644, + "grad_norm": 1.614102840423584, + "learning_rate": 8.029197080291971e-07, + "loss": 2.706, + "step": 58 + }, + { + "epoch": 0.06465753424657535, + "grad_norm": 1.6313952207565308, + "learning_rate": 8.211678832116789e-07, + "loss": 2.6451, + "step": 59 + }, + { + "epoch": 0.06575342465753424, + "grad_norm": 1.8218132257461548, + "learning_rate": 8.394160583941606e-07, + "loss": 2.6168, + "step": 60 + }, + { + "epoch": 0.06575342465753424, + "eval_loss": 2.719381332397461, + "eval_runtime": 37.2407, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 60 + }, + { + "epoch": 0.06684931506849315, + "grad_norm": 1.7891069650650024, + "learning_rate": 8.576642335766425e-07, + "loss": 2.6559, + "step": 61 + }, + { + "epoch": 0.06794520547945206, + "grad_norm": 1.6071841716766357, + "learning_rate": 8.759124087591242e-07, + "loss": 2.6428, + "step": 62 + }, + { + "epoch": 0.06904109589041096, + "grad_norm": 1.7150017023086548, + "learning_rate": 8.94160583941606e-07, + "loss": 2.6496, + "step": 63 + }, + { + "epoch": 0.07013698630136986, + "grad_norm": 1.7664421796798706, + "learning_rate": 9.124087591240876e-07, + "loss": 2.6671, + "step": 64 + }, + { + "epoch": 0.07123287671232877, + "grad_norm": 1.684327483177185, + "learning_rate": 9.306569343065694e-07, + "loss": 2.6309, + "step": 65 + }, + { + "epoch": 0.07232876712328767, + "grad_norm": 1.5630618333816528, + "learning_rate": 9.489051094890511e-07, + "loss": 2.6513, + "step": 66 + }, + { + "epoch": 0.07342465753424658, + "grad_norm": 1.622764229774475, + "learning_rate": 9.67153284671533e-07, + "loss": 2.6346, + "step": 67 + }, + { + "epoch": 0.07452054794520548, + "grad_norm": 1.7873258590698242, + "learning_rate": 9.854014598540146e-07, + "loss": 2.6784, + "step": 68 + }, + { + "epoch": 0.07561643835616438, + "grad_norm": 2.058234214782715, + "learning_rate": 1.0036496350364965e-06, + "loss": 2.6063, + "step": 69 + }, + { + "epoch": 0.07671232876712329, + "grad_norm": 1.9183952808380127, + "learning_rate": 1.0218978102189781e-06, + "loss": 2.575, + "step": 70 + }, + { + "epoch": 0.07671232876712329, + "eval_loss": 2.6536612510681152, + "eval_runtime": 37.2183, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 70 + }, + { + "epoch": 0.07780821917808219, + "grad_norm": 1.931950330734253, + "learning_rate": 1.04014598540146e-06, + "loss": 2.6429, + "step": 71 + }, + { + "epoch": 0.0789041095890411, + "grad_norm": 2.045337438583374, + "learning_rate": 1.0583941605839416e-06, + "loss": 2.5218, + "step": 72 + }, + { + "epoch": 0.08, + "grad_norm": 1.9110971689224243, + "learning_rate": 1.0766423357664233e-06, + "loss": 2.5277, + "step": 73 + }, + { + "epoch": 0.08109589041095891, + "grad_norm": 2.4468348026275635, + "learning_rate": 1.0948905109489052e-06, + "loss": 2.547, + "step": 74 + }, + { + "epoch": 0.0821917808219178, + "grad_norm": 1.887924075126648, + "learning_rate": 1.113138686131387e-06, + "loss": 2.5713, + "step": 75 + }, + { + "epoch": 0.08328767123287671, + "grad_norm": 2.891157865524292, + "learning_rate": 1.1313868613138687e-06, + "loss": 2.5812, + "step": 76 + }, + { + "epoch": 0.08438356164383562, + "grad_norm": 2.295057773590088, + "learning_rate": 1.1496350364963505e-06, + "loss": 2.4991, + "step": 77 + }, + { + "epoch": 0.08547945205479453, + "grad_norm": 2.3006889820098877, + "learning_rate": 1.1678832116788322e-06, + "loss": 2.4947, + "step": 78 + }, + { + "epoch": 0.08657534246575342, + "grad_norm": 2.439432382583618, + "learning_rate": 1.1861313868613138e-06, + "loss": 2.4912, + "step": 79 + }, + { + "epoch": 0.08767123287671233, + "grad_norm": 2.269310712814331, + "learning_rate": 1.2043795620437959e-06, + "loss": 2.4802, + "step": 80 + }, + { + "epoch": 0.08767123287671233, + "eval_loss": 2.5736074447631836, + "eval_runtime": 37.2255, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 80 + }, + { + "epoch": 0.08876712328767124, + "grad_norm": 2.1347496509552, + "learning_rate": 1.2226277372262775e-06, + "loss": 2.481, + "step": 81 + }, + { + "epoch": 0.08986301369863013, + "grad_norm": 2.369415760040283, + "learning_rate": 1.2408759124087592e-06, + "loss": 2.5204, + "step": 82 + }, + { + "epoch": 0.09095890410958904, + "grad_norm": 2.8223581314086914, + "learning_rate": 1.259124087591241e-06, + "loss": 2.4783, + "step": 83 + }, + { + "epoch": 0.09205479452054795, + "grad_norm": 2.6199405193328857, + "learning_rate": 1.2773722627737229e-06, + "loss": 2.4926, + "step": 84 + }, + { + "epoch": 0.09315068493150686, + "grad_norm": 2.776623487472534, + "learning_rate": 1.2956204379562045e-06, + "loss": 2.4781, + "step": 85 + }, + { + "epoch": 0.09424657534246575, + "grad_norm": 2.649749755859375, + "learning_rate": 1.3138686131386864e-06, + "loss": 2.4184, + "step": 86 + }, + { + "epoch": 0.09534246575342466, + "grad_norm": 2.62575101852417, + "learning_rate": 1.332116788321168e-06, + "loss": 2.4466, + "step": 87 + }, + { + "epoch": 0.09643835616438357, + "grad_norm": 3.135901689529419, + "learning_rate": 1.3503649635036497e-06, + "loss": 2.3681, + "step": 88 + }, + { + "epoch": 0.09753424657534246, + "grad_norm": 2.8516979217529297, + "learning_rate": 1.3686131386861315e-06, + "loss": 2.4107, + "step": 89 + }, + { + "epoch": 0.09863013698630137, + "grad_norm": 2.5140628814697266, + "learning_rate": 1.3868613138686132e-06, + "loss": 2.3323, + "step": 90 + }, + { + "epoch": 0.09863013698630137, + "eval_loss": 2.4967005252838135, + "eval_runtime": 37.2089, + "eval_samples_per_second": 26.875, + "eval_steps_per_second": 1.129, + "step": 90 + }, + { + "epoch": 0.09972602739726028, + "grad_norm": 3.1235268115997314, + "learning_rate": 1.405109489051095e-06, + "loss": 2.4448, + "step": 91 + }, + { + "epoch": 0.10082191780821918, + "grad_norm": 2.7377545833587646, + "learning_rate": 1.4233576642335767e-06, + "loss": 2.3931, + "step": 92 + }, + { + "epoch": 0.10191780821917808, + "grad_norm": 2.8881747722625732, + "learning_rate": 1.4416058394160583e-06, + "loss": 2.4274, + "step": 93 + }, + { + "epoch": 0.10301369863013699, + "grad_norm": 2.584059000015259, + "learning_rate": 1.4598540145985402e-06, + "loss": 2.385, + "step": 94 + }, + { + "epoch": 0.10410958904109589, + "grad_norm": 3.4064419269561768, + "learning_rate": 1.478102189781022e-06, + "loss": 2.3554, + "step": 95 + }, + { + "epoch": 0.1052054794520548, + "grad_norm": 2.2290151119232178, + "learning_rate": 1.496350364963504e-06, + "loss": 2.4137, + "step": 96 + }, + { + "epoch": 0.1063013698630137, + "grad_norm": 2.7750914096832275, + "learning_rate": 1.5145985401459856e-06, + "loss": 2.3843, + "step": 97 + }, + { + "epoch": 0.1073972602739726, + "grad_norm": 2.619952440261841, + "learning_rate": 1.5328467153284674e-06, + "loss": 2.4046, + "step": 98 + }, + { + "epoch": 0.10849315068493151, + "grad_norm": 2.54819917678833, + "learning_rate": 1.551094890510949e-06, + "loss": 2.3543, + "step": 99 + }, + { + "epoch": 0.1095890410958904, + "grad_norm": 2.44057559967041, + "learning_rate": 1.5693430656934307e-06, + "loss": 2.3001, + "step": 100 + }, + { + "epoch": 0.1095890410958904, + "eval_loss": 2.432290554046631, + "eval_runtime": 37.2517, + "eval_samples_per_second": 26.844, + "eval_steps_per_second": 1.127, + "step": 100 + }, + { + "epoch": 0.11068493150684931, + "grad_norm": 2.51607608795166, + "learning_rate": 1.5875912408759126e-06, + "loss": 2.2975, + "step": 101 + }, + { + "epoch": 0.11178082191780822, + "grad_norm": 2.2553768157958984, + "learning_rate": 1.6058394160583942e-06, + "loss": 2.3619, + "step": 102 + }, + { + "epoch": 0.11287671232876713, + "grad_norm": 2.4806911945343018, + "learning_rate": 1.624087591240876e-06, + "loss": 2.4066, + "step": 103 + }, + { + "epoch": 0.11397260273972602, + "grad_norm": 2.5546786785125732, + "learning_rate": 1.6423357664233577e-06, + "loss": 2.3172, + "step": 104 + }, + { + "epoch": 0.11506849315068493, + "grad_norm": 2.4799184799194336, + "learning_rate": 1.6605839416058394e-06, + "loss": 2.2578, + "step": 105 + }, + { + "epoch": 0.11616438356164384, + "grad_norm": 3.2751758098602295, + "learning_rate": 1.6788321167883212e-06, + "loss": 2.2898, + "step": 106 + }, + { + "epoch": 0.11726027397260275, + "grad_norm": 2.386707067489624, + "learning_rate": 1.6970802919708029e-06, + "loss": 2.3128, + "step": 107 + }, + { + "epoch": 0.11835616438356164, + "grad_norm": 2.480304002761841, + "learning_rate": 1.715328467153285e-06, + "loss": 2.2404, + "step": 108 + }, + { + "epoch": 0.11945205479452055, + "grad_norm": 2.4512832164764404, + "learning_rate": 1.7335766423357666e-06, + "loss": 2.2934, + "step": 109 + }, + { + "epoch": 0.12054794520547946, + "grad_norm": 2.758535623550415, + "learning_rate": 1.7518248175182485e-06, + "loss": 2.2941, + "step": 110 + }, + { + "epoch": 0.12054794520547946, + "eval_loss": 2.3837950229644775, + "eval_runtime": 36.9129, + "eval_samples_per_second": 27.091, + "eval_steps_per_second": 1.138, + "step": 110 + }, + { + "epoch": 0.12164383561643835, + "grad_norm": 3.0642340183258057, + "learning_rate": 1.77007299270073e-06, + "loss": 2.3243, + "step": 111 + }, + { + "epoch": 0.12273972602739726, + "grad_norm": 2.8591346740722656, + "learning_rate": 1.788321167883212e-06, + "loss": 2.3018, + "step": 112 + }, + { + "epoch": 0.12383561643835617, + "grad_norm": 3.4537875652313232, + "learning_rate": 1.8065693430656936e-06, + "loss": 2.3535, + "step": 113 + }, + { + "epoch": 0.12493150684931507, + "grad_norm": 2.427309274673462, + "learning_rate": 1.8248175182481753e-06, + "loss": 2.2885, + "step": 114 + }, + { + "epoch": 0.12602739726027398, + "grad_norm": 3.87589168548584, + "learning_rate": 1.8430656934306571e-06, + "loss": 2.2312, + "step": 115 + }, + { + "epoch": 0.12712328767123288, + "grad_norm": 2.833167314529419, + "learning_rate": 1.8613138686131388e-06, + "loss": 2.2565, + "step": 116 + }, + { + "epoch": 0.12821917808219177, + "grad_norm": 3.1881866455078125, + "learning_rate": 1.8795620437956206e-06, + "loss": 2.2555, + "step": 117 + }, + { + "epoch": 0.1293150684931507, + "grad_norm": 2.531094551086426, + "learning_rate": 1.8978102189781023e-06, + "loss": 2.25, + "step": 118 + }, + { + "epoch": 0.1304109589041096, + "grad_norm": 2.865161418914795, + "learning_rate": 1.916058394160584e-06, + "loss": 2.2533, + "step": 119 + }, + { + "epoch": 0.13150684931506848, + "grad_norm": 2.625645160675049, + "learning_rate": 1.934306569343066e-06, + "loss": 2.2679, + "step": 120 + }, + { + "epoch": 0.13150684931506848, + "eval_loss": 2.3434090614318848, + "eval_runtime": 37.0839, + "eval_samples_per_second": 26.966, + "eval_steps_per_second": 1.133, + "step": 120 + }, + { + "epoch": 0.1326027397260274, + "grad_norm": 3.7083628177642822, + "learning_rate": 1.952554744525548e-06, + "loss": 2.17, + "step": 121 + }, + { + "epoch": 0.1336986301369863, + "grad_norm": 2.7827553749084473, + "learning_rate": 1.9708029197080293e-06, + "loss": 2.2228, + "step": 122 + }, + { + "epoch": 0.13479452054794522, + "grad_norm": 3.553680419921875, + "learning_rate": 1.989051094890511e-06, + "loss": 2.2112, + "step": 123 + }, + { + "epoch": 0.1358904109589041, + "grad_norm": 2.5528950691223145, + "learning_rate": 2.007299270072993e-06, + "loss": 2.2178, + "step": 124 + }, + { + "epoch": 0.136986301369863, + "grad_norm": 3.2710938453674316, + "learning_rate": 2.025547445255475e-06, + "loss": 2.2694, + "step": 125 + }, + { + "epoch": 0.13808219178082193, + "grad_norm": 3.0480809211730957, + "learning_rate": 2.0437956204379563e-06, + "loss": 2.2046, + "step": 126 + }, + { + "epoch": 0.13917808219178082, + "grad_norm": 2.5092124938964844, + "learning_rate": 2.062043795620438e-06, + "loss": 2.2175, + "step": 127 + }, + { + "epoch": 0.14027397260273972, + "grad_norm": 2.8840532302856445, + "learning_rate": 2.08029197080292e-06, + "loss": 2.205, + "step": 128 + }, + { + "epoch": 0.14136986301369864, + "grad_norm": 2.737955093383789, + "learning_rate": 2.0985401459854014e-06, + "loss": 2.2614, + "step": 129 + }, + { + "epoch": 0.14246575342465753, + "grad_norm": 3.1381702423095703, + "learning_rate": 2.1167883211678833e-06, + "loss": 2.1636, + "step": 130 + }, + { + "epoch": 0.14246575342465753, + "eval_loss": 2.311958074569702, + "eval_runtime": 37.1344, + "eval_samples_per_second": 26.929, + "eval_steps_per_second": 1.131, + "step": 130 + }, + { + "epoch": 0.14356164383561643, + "grad_norm": 2.687152862548828, + "learning_rate": 2.135036496350365e-06, + "loss": 2.2203, + "step": 131 + }, + { + "epoch": 0.14465753424657535, + "grad_norm": 2.971111536026001, + "learning_rate": 2.1532846715328466e-06, + "loss": 2.2515, + "step": 132 + }, + { + "epoch": 0.14575342465753424, + "grad_norm": 3.1336281299591064, + "learning_rate": 2.171532846715329e-06, + "loss": 2.1551, + "step": 133 + }, + { + "epoch": 0.14684931506849316, + "grad_norm": 2.6687278747558594, + "learning_rate": 2.1897810218978103e-06, + "loss": 2.1714, + "step": 134 + }, + { + "epoch": 0.14794520547945206, + "grad_norm": 3.571277141571045, + "learning_rate": 2.208029197080292e-06, + "loss": 2.2141, + "step": 135 + }, + { + "epoch": 0.14904109589041095, + "grad_norm": 2.6429405212402344, + "learning_rate": 2.226277372262774e-06, + "loss": 2.1816, + "step": 136 + }, + { + "epoch": 0.15013698630136987, + "grad_norm": 3.824120283126831, + "learning_rate": 2.244525547445256e-06, + "loss": 2.1693, + "step": 137 + }, + { + "epoch": 0.15123287671232877, + "grad_norm": 3.0339856147766113, + "learning_rate": 2.2627737226277373e-06, + "loss": 2.2146, + "step": 138 + }, + { + "epoch": 0.15232876712328766, + "grad_norm": 3.3885395526885986, + "learning_rate": 2.281021897810219e-06, + "loss": 2.2108, + "step": 139 + }, + { + "epoch": 0.15342465753424658, + "grad_norm": 3.108035087585449, + "learning_rate": 2.299270072992701e-06, + "loss": 2.1697, + "step": 140 + }, + { + "epoch": 0.15342465753424658, + "eval_loss": 2.2911388874053955, + "eval_runtime": 37.1917, + "eval_samples_per_second": 26.888, + "eval_steps_per_second": 1.129, + "step": 140 + }, + { + "epoch": 0.15452054794520548, + "grad_norm": 3.0966873168945312, + "learning_rate": 2.3175182481751825e-06, + "loss": 2.1562, + "step": 141 + }, + { + "epoch": 0.15561643835616437, + "grad_norm": 2.5972113609313965, + "learning_rate": 2.3357664233576643e-06, + "loss": 2.166, + "step": 142 + }, + { + "epoch": 0.1567123287671233, + "grad_norm": 3.031276226043701, + "learning_rate": 2.354014598540146e-06, + "loss": 2.1402, + "step": 143 + }, + { + "epoch": 0.1578082191780822, + "grad_norm": 2.853633403778076, + "learning_rate": 2.3722627737226276e-06, + "loss": 2.1671, + "step": 144 + }, + { + "epoch": 0.1589041095890411, + "grad_norm": 3.707199811935425, + "learning_rate": 2.39051094890511e-06, + "loss": 2.1801, + "step": 145 + }, + { + "epoch": 0.16, + "grad_norm": 3.106747627258301, + "learning_rate": 2.4087591240875918e-06, + "loss": 2.1157, + "step": 146 + }, + { + "epoch": 0.1610958904109589, + "grad_norm": 2.863614320755005, + "learning_rate": 2.427007299270073e-06, + "loss": 2.1314, + "step": 147 + }, + { + "epoch": 0.16219178082191782, + "grad_norm": 3.0352959632873535, + "learning_rate": 2.445255474452555e-06, + "loss": 2.1378, + "step": 148 + }, + { + "epoch": 0.1632876712328767, + "grad_norm": 3.101146936416626, + "learning_rate": 2.463503649635037e-06, + "loss": 2.1985, + "step": 149 + }, + { + "epoch": 0.1643835616438356, + "grad_norm": 2.923638343811035, + "learning_rate": 2.4817518248175183e-06, + "loss": 2.0982, + "step": 150 + }, + { + "epoch": 0.1643835616438356, + "eval_loss": 2.265890121459961, + "eval_runtime": 37.1154, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 150 + }, + { + "epoch": 0.16547945205479453, + "grad_norm": 3.429690361022949, + "learning_rate": 2.5e-06, + "loss": 2.071, + "step": 151 + }, + { + "epoch": 0.16657534246575342, + "grad_norm": 2.803128480911255, + "learning_rate": 2.518248175182482e-06, + "loss": 2.1976, + "step": 152 + }, + { + "epoch": 0.16767123287671232, + "grad_norm": 2.969264507293701, + "learning_rate": 2.5364963503649635e-06, + "loss": 2.1014, + "step": 153 + }, + { + "epoch": 0.16876712328767124, + "grad_norm": 2.749788999557495, + "learning_rate": 2.5547445255474458e-06, + "loss": 2.0907, + "step": 154 + }, + { + "epoch": 0.16986301369863013, + "grad_norm": 2.8180391788482666, + "learning_rate": 2.5729927007299272e-06, + "loss": 2.1504, + "step": 155 + }, + { + "epoch": 0.17095890410958905, + "grad_norm": 2.956796407699585, + "learning_rate": 2.591240875912409e-06, + "loss": 2.1247, + "step": 156 + }, + { + "epoch": 0.17205479452054795, + "grad_norm": 2.518430709838867, + "learning_rate": 2.6094890510948905e-06, + "loss": 2.1235, + "step": 157 + }, + { + "epoch": 0.17315068493150684, + "grad_norm": 3.3999993801116943, + "learning_rate": 2.627737226277373e-06, + "loss": 2.1458, + "step": 158 + }, + { + "epoch": 0.17424657534246576, + "grad_norm": 3.6163454055786133, + "learning_rate": 2.6459854014598542e-06, + "loss": 2.1712, + "step": 159 + }, + { + "epoch": 0.17534246575342466, + "grad_norm": 2.844896078109741, + "learning_rate": 2.664233576642336e-06, + "loss": 2.1236, + "step": 160 + }, + { + "epoch": 0.17534246575342466, + "eval_loss": 2.247826099395752, + "eval_runtime": 37.1672, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 160 + }, + { + "epoch": 0.17643835616438355, + "grad_norm": 3.897892713546753, + "learning_rate": 2.6824817518248175e-06, + "loss": 2.0972, + "step": 161 + }, + { + "epoch": 0.17753424657534247, + "grad_norm": 3.7879202365875244, + "learning_rate": 2.7007299270072994e-06, + "loss": 2.135, + "step": 162 + }, + { + "epoch": 0.17863013698630137, + "grad_norm": 2.4167919158935547, + "learning_rate": 2.718978102189781e-06, + "loss": 2.1016, + "step": 163 + }, + { + "epoch": 0.17972602739726026, + "grad_norm": 3.054913282394409, + "learning_rate": 2.737226277372263e-06, + "loss": 2.17, + "step": 164 + }, + { + "epoch": 0.18082191780821918, + "grad_norm": 2.6863350868225098, + "learning_rate": 2.755474452554745e-06, + "loss": 2.1402, + "step": 165 + }, + { + "epoch": 0.18191780821917808, + "grad_norm": 3.4107234477996826, + "learning_rate": 2.7737226277372264e-06, + "loss": 2.0967, + "step": 166 + }, + { + "epoch": 0.18301369863013697, + "grad_norm": 2.360496997833252, + "learning_rate": 2.7919708029197082e-06, + "loss": 2.2011, + "step": 167 + }, + { + "epoch": 0.1841095890410959, + "grad_norm": 3.5195155143737793, + "learning_rate": 2.81021897810219e-06, + "loss": 2.1396, + "step": 168 + }, + { + "epoch": 0.1852054794520548, + "grad_norm": 3.1431586742401123, + "learning_rate": 2.828467153284672e-06, + "loss": 2.0842, + "step": 169 + }, + { + "epoch": 0.1863013698630137, + "grad_norm": 2.77384090423584, + "learning_rate": 2.8467153284671534e-06, + "loss": 2.1278, + "step": 170 + }, + { + "epoch": 0.1863013698630137, + "eval_loss": 2.2353579998016357, + "eval_runtime": 37.1635, + "eval_samples_per_second": 26.908, + "eval_steps_per_second": 1.13, + "step": 170 + }, + { + "epoch": 0.1873972602739726, + "grad_norm": 2.8373658657073975, + "learning_rate": 2.8649635036496353e-06, + "loss": 2.0448, + "step": 171 + }, + { + "epoch": 0.1884931506849315, + "grad_norm": 3.4246926307678223, + "learning_rate": 2.8832116788321167e-06, + "loss": 2.086, + "step": 172 + }, + { + "epoch": 0.18958904109589042, + "grad_norm": 3.76779842376709, + "learning_rate": 2.901459854014599e-06, + "loss": 2.0495, + "step": 173 + }, + { + "epoch": 0.19068493150684931, + "grad_norm": 2.7396318912506104, + "learning_rate": 2.9197080291970804e-06, + "loss": 2.1436, + "step": 174 + }, + { + "epoch": 0.1917808219178082, + "grad_norm": 2.7848520278930664, + "learning_rate": 2.9379562043795623e-06, + "loss": 2.1175, + "step": 175 + }, + { + "epoch": 0.19287671232876713, + "grad_norm": 2.672060489654541, + "learning_rate": 2.956204379562044e-06, + "loss": 2.1036, + "step": 176 + }, + { + "epoch": 0.19397260273972602, + "grad_norm": 3.0901851654052734, + "learning_rate": 2.9744525547445256e-06, + "loss": 2.1422, + "step": 177 + }, + { + "epoch": 0.19506849315068492, + "grad_norm": 2.8616273403167725, + "learning_rate": 2.992700729927008e-06, + "loss": 2.1018, + "step": 178 + }, + { + "epoch": 0.19616438356164384, + "grad_norm": 3.3321707248687744, + "learning_rate": 3.0109489051094893e-06, + "loss": 2.0624, + "step": 179 + }, + { + "epoch": 0.19726027397260273, + "grad_norm": 3.1142725944519043, + "learning_rate": 3.029197080291971e-06, + "loss": 2.0417, + "step": 180 + }, + { + "epoch": 0.19726027397260273, + "eval_loss": 2.2239184379577637, + "eval_runtime": 37.1327, + "eval_samples_per_second": 26.93, + "eval_steps_per_second": 1.131, + "step": 180 + }, + { + "epoch": 0.19835616438356166, + "grad_norm": 2.9385526180267334, + "learning_rate": 3.0474452554744526e-06, + "loss": 2.1352, + "step": 181 + }, + { + "epoch": 0.19945205479452055, + "grad_norm": 3.4400417804718018, + "learning_rate": 3.065693430656935e-06, + "loss": 2.1138, + "step": 182 + }, + { + "epoch": 0.20054794520547944, + "grad_norm": 2.871488094329834, + "learning_rate": 3.0839416058394163e-06, + "loss": 2.1048, + "step": 183 + }, + { + "epoch": 0.20164383561643837, + "grad_norm": 2.614055633544922, + "learning_rate": 3.102189781021898e-06, + "loss": 2.1185, + "step": 184 + }, + { + "epoch": 0.20273972602739726, + "grad_norm": 3.247382640838623, + "learning_rate": 3.1204379562043796e-06, + "loss": 2.1052, + "step": 185 + }, + { + "epoch": 0.20383561643835615, + "grad_norm": 2.6722798347473145, + "learning_rate": 3.1386861313868614e-06, + "loss": 2.0419, + "step": 186 + }, + { + "epoch": 0.20493150684931508, + "grad_norm": 3.482536554336548, + "learning_rate": 3.156934306569343e-06, + "loss": 2.0283, + "step": 187 + }, + { + "epoch": 0.20602739726027397, + "grad_norm": 2.8816161155700684, + "learning_rate": 3.175182481751825e-06, + "loss": 2.1435, + "step": 188 + }, + { + "epoch": 0.20712328767123286, + "grad_norm": 3.5030934810638428, + "learning_rate": 3.193430656934307e-06, + "loss": 2.0987, + "step": 189 + }, + { + "epoch": 0.20821917808219179, + "grad_norm": 2.4934983253479004, + "learning_rate": 3.2116788321167884e-06, + "loss": 2.0511, + "step": 190 + }, + { + "epoch": 0.20821917808219179, + "eval_loss": 2.2034926414489746, + "eval_runtime": 37.1051, + "eval_samples_per_second": 26.95, + "eval_steps_per_second": 1.132, + "step": 190 + }, + { + "epoch": 0.20931506849315068, + "grad_norm": 2.9186456203460693, + "learning_rate": 3.2299270072992707e-06, + "loss": 2.0801, + "step": 191 + }, + { + "epoch": 0.2104109589041096, + "grad_norm": 2.7397220134735107, + "learning_rate": 3.248175182481752e-06, + "loss": 2.0577, + "step": 192 + }, + { + "epoch": 0.2115068493150685, + "grad_norm": 2.7978568077087402, + "learning_rate": 3.266423357664234e-06, + "loss": 1.9916, + "step": 193 + }, + { + "epoch": 0.2126027397260274, + "grad_norm": 2.6294498443603516, + "learning_rate": 3.2846715328467155e-06, + "loss": 2.0868, + "step": 194 + }, + { + "epoch": 0.2136986301369863, + "grad_norm": 2.6224911212921143, + "learning_rate": 3.3029197080291973e-06, + "loss": 2.0626, + "step": 195 + }, + { + "epoch": 0.2147945205479452, + "grad_norm": 2.7921297550201416, + "learning_rate": 3.3211678832116788e-06, + "loss": 2.0906, + "step": 196 + }, + { + "epoch": 0.2158904109589041, + "grad_norm": 2.9427266120910645, + "learning_rate": 3.339416058394161e-06, + "loss": 2.0625, + "step": 197 + }, + { + "epoch": 0.21698630136986302, + "grad_norm": 2.5963096618652344, + "learning_rate": 3.3576642335766425e-06, + "loss": 2.064, + "step": 198 + }, + { + "epoch": 0.21808219178082192, + "grad_norm": 2.7042300701141357, + "learning_rate": 3.3759124087591243e-06, + "loss": 2.0615, + "step": 199 + }, + { + "epoch": 0.2191780821917808, + "grad_norm": 2.6845407485961914, + "learning_rate": 3.3941605839416058e-06, + "loss": 2.0666, + "step": 200 + }, + { + "epoch": 0.2191780821917808, + "eval_loss": 2.193758726119995, + "eval_runtime": 37.1155, + "eval_samples_per_second": 26.943, + "eval_steps_per_second": 1.132, + "step": 200 + }, + { + "epoch": 0.22027397260273973, + "grad_norm": 2.6319382190704346, + "learning_rate": 3.412408759124088e-06, + "loss": 2.0607, + "step": 201 + }, + { + "epoch": 0.22136986301369863, + "grad_norm": 2.993312120437622, + "learning_rate": 3.43065693430657e-06, + "loss": 2.0841, + "step": 202 + }, + { + "epoch": 0.22246575342465755, + "grad_norm": 3.092479944229126, + "learning_rate": 3.4489051094890513e-06, + "loss": 1.9966, + "step": 203 + }, + { + "epoch": 0.22356164383561644, + "grad_norm": 3.04301381111145, + "learning_rate": 3.467153284671533e-06, + "loss": 2.0652, + "step": 204 + }, + { + "epoch": 0.22465753424657534, + "grad_norm": 2.7950596809387207, + "learning_rate": 3.4854014598540146e-06, + "loss": 2.0167, + "step": 205 + }, + { + "epoch": 0.22575342465753426, + "grad_norm": 3.0609445571899414, + "learning_rate": 3.503649635036497e-06, + "loss": 2.0511, + "step": 206 + }, + { + "epoch": 0.22684931506849315, + "grad_norm": 3.079122543334961, + "learning_rate": 3.5218978102189783e-06, + "loss": 2.0507, + "step": 207 + }, + { + "epoch": 0.22794520547945205, + "grad_norm": 2.6499931812286377, + "learning_rate": 3.54014598540146e-06, + "loss": 2.0657, + "step": 208 + }, + { + "epoch": 0.22904109589041097, + "grad_norm": 3.3667314052581787, + "learning_rate": 3.5583941605839416e-06, + "loss": 2.0959, + "step": 209 + }, + { + "epoch": 0.23013698630136986, + "grad_norm": 2.955397129058838, + "learning_rate": 3.576642335766424e-06, + "loss": 2.128, + "step": 210 + }, + { + "epoch": 0.23013698630136986, + "eval_loss": 2.189744234085083, + "eval_runtime": 36.8022, + "eval_samples_per_second": 27.172, + "eval_steps_per_second": 1.141, + "step": 210 + }, + { + "epoch": 0.23123287671232876, + "grad_norm": 2.9283270835876465, + "learning_rate": 3.5948905109489054e-06, + "loss": 2.1244, + "step": 211 + }, + { + "epoch": 0.23232876712328768, + "grad_norm": 3.3702054023742676, + "learning_rate": 3.6131386861313872e-06, + "loss": 1.9598, + "step": 212 + }, + { + "epoch": 0.23342465753424657, + "grad_norm": 3.0767223834991455, + "learning_rate": 3.6313868613138687e-06, + "loss": 1.9655, + "step": 213 + }, + { + "epoch": 0.2345205479452055, + "grad_norm": 2.8419556617736816, + "learning_rate": 3.6496350364963505e-06, + "loss": 2.0019, + "step": 214 + }, + { + "epoch": 0.2356164383561644, + "grad_norm": 2.9517970085144043, + "learning_rate": 3.667883211678833e-06, + "loss": 2.028, + "step": 215 + }, + { + "epoch": 0.23671232876712328, + "grad_norm": 3.122650384902954, + "learning_rate": 3.6861313868613142e-06, + "loss": 1.9962, + "step": 216 + }, + { + "epoch": 0.2378082191780822, + "grad_norm": 2.7380897998809814, + "learning_rate": 3.704379562043796e-06, + "loss": 1.994, + "step": 217 + }, + { + "epoch": 0.2389041095890411, + "grad_norm": 2.9314167499542236, + "learning_rate": 3.7226277372262775e-06, + "loss": 2.0991, + "step": 218 + }, + { + "epoch": 0.24, + "grad_norm": 2.9320032596588135, + "learning_rate": 3.7408759124087594e-06, + "loss": 2.0525, + "step": 219 + }, + { + "epoch": 0.2410958904109589, + "grad_norm": 2.8559162616729736, + "learning_rate": 3.7591240875912412e-06, + "loss": 1.999, + "step": 220 + }, + { + "epoch": 0.2410958904109589, + "eval_loss": 2.168269634246826, + "eval_runtime": 37.0075, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 220 + }, + { + "epoch": 0.2421917808219178, + "grad_norm": 2.692728281021118, + "learning_rate": 3.777372262773723e-06, + "loss": 2.0338, + "step": 221 + }, + { + "epoch": 0.2432876712328767, + "grad_norm": 3.3898634910583496, + "learning_rate": 3.7956204379562045e-06, + "loss": 2.0188, + "step": 222 + }, + { + "epoch": 0.24438356164383562, + "grad_norm": 4.22472620010376, + "learning_rate": 3.8138686131386864e-06, + "loss": 2.0173, + "step": 223 + }, + { + "epoch": 0.24547945205479452, + "grad_norm": 4.254660129547119, + "learning_rate": 3.832116788321168e-06, + "loss": 2.0552, + "step": 224 + }, + { + "epoch": 0.2465753424657534, + "grad_norm": 2.9048502445220947, + "learning_rate": 3.85036496350365e-06, + "loss": 1.9933, + "step": 225 + }, + { + "epoch": 0.24767123287671233, + "grad_norm": 3.3615903854370117, + "learning_rate": 3.868613138686132e-06, + "loss": 1.9841, + "step": 226 + }, + { + "epoch": 0.24876712328767123, + "grad_norm": 3.673733711242676, + "learning_rate": 3.886861313868613e-06, + "loss": 2.008, + "step": 227 + }, + { + "epoch": 0.24986301369863015, + "grad_norm": 2.8139588832855225, + "learning_rate": 3.905109489051096e-06, + "loss": 2.0353, + "step": 228 + }, + { + "epoch": 0.25095890410958904, + "grad_norm": 3.1198532581329346, + "learning_rate": 3.923357664233577e-06, + "loss": 2.0665, + "step": 229 + }, + { + "epoch": 0.25205479452054796, + "grad_norm": 2.9415290355682373, + "learning_rate": 3.9416058394160585e-06, + "loss": 2.0086, + "step": 230 + }, + { + "epoch": 0.25205479452054796, + "eval_loss": 2.158285140991211, + "eval_runtime": 37.098, + "eval_samples_per_second": 26.956, + "eval_steps_per_second": 1.132, + "step": 230 + }, + { + "epoch": 0.25315068493150683, + "grad_norm": 2.797816514968872, + "learning_rate": 3.95985401459854e-06, + "loss": 2.0766, + "step": 231 + }, + { + "epoch": 0.25424657534246575, + "grad_norm": 2.702988386154175, + "learning_rate": 3.978102189781022e-06, + "loss": 2.0902, + "step": 232 + }, + { + "epoch": 0.2553424657534247, + "grad_norm": 2.6598644256591797, + "learning_rate": 3.996350364963504e-06, + "loss": 1.9482, + "step": 233 + }, + { + "epoch": 0.25643835616438354, + "grad_norm": 2.961045026779175, + "learning_rate": 4.014598540145986e-06, + "loss": 2.0437, + "step": 234 + }, + { + "epoch": 0.25753424657534246, + "grad_norm": 2.663412094116211, + "learning_rate": 4.032846715328467e-06, + "loss": 1.938, + "step": 235 + }, + { + "epoch": 0.2586301369863014, + "grad_norm": 2.650676965713501, + "learning_rate": 4.05109489051095e-06, + "loss": 1.9949, + "step": 236 + }, + { + "epoch": 0.25972602739726025, + "grad_norm": 3.020890951156616, + "learning_rate": 4.069343065693431e-06, + "loss": 2.0114, + "step": 237 + }, + { + "epoch": 0.2608219178082192, + "grad_norm": 2.584810972213745, + "learning_rate": 4.0875912408759126e-06, + "loss": 1.9733, + "step": 238 + }, + { + "epoch": 0.2619178082191781, + "grad_norm": 2.735201358795166, + "learning_rate": 4.105839416058395e-06, + "loss": 1.9983, + "step": 239 + }, + { + "epoch": 0.26301369863013696, + "grad_norm": 2.4640519618988037, + "learning_rate": 4.124087591240876e-06, + "loss": 2.0639, + "step": 240 + }, + { + "epoch": 0.26301369863013696, + "eval_loss": 2.1486563682556152, + "eval_runtime": 37.1245, + "eval_samples_per_second": 26.936, + "eval_steps_per_second": 1.131, + "step": 240 + }, + { + "epoch": 0.2641095890410959, + "grad_norm": 2.472024440765381, + "learning_rate": 4.1423357664233586e-06, + "loss": 1.985, + "step": 241 + }, + { + "epoch": 0.2652054794520548, + "grad_norm": 2.689582586288452, + "learning_rate": 4.16058394160584e-06, + "loss": 1.9904, + "step": 242 + }, + { + "epoch": 0.2663013698630137, + "grad_norm": 3.031712293624878, + "learning_rate": 4.1788321167883214e-06, + "loss": 2.0685, + "step": 243 + }, + { + "epoch": 0.2673972602739726, + "grad_norm": 2.758924722671509, + "learning_rate": 4.197080291970803e-06, + "loss": 1.9359, + "step": 244 + }, + { + "epoch": 0.2684931506849315, + "grad_norm": 2.8976337909698486, + "learning_rate": 4.215328467153285e-06, + "loss": 2.0103, + "step": 245 + }, + { + "epoch": 0.26958904109589044, + "grad_norm": 3.1584532260894775, + "learning_rate": 4.233576642335767e-06, + "loss": 1.9749, + "step": 246 + }, + { + "epoch": 0.2706849315068493, + "grad_norm": 2.826107978820801, + "learning_rate": 4.251824817518249e-06, + "loss": 1.9712, + "step": 247 + }, + { + "epoch": 0.2717808219178082, + "grad_norm": 2.8240978717803955, + "learning_rate": 4.27007299270073e-06, + "loss": 1.9967, + "step": 248 + }, + { + "epoch": 0.27287671232876715, + "grad_norm": 2.875216484069824, + "learning_rate": 4.288321167883212e-06, + "loss": 1.9523, + "step": 249 + }, + { + "epoch": 0.273972602739726, + "grad_norm": 2.85149884223938, + "learning_rate": 4.306569343065693e-06, + "loss": 1.967, + "step": 250 + }, + { + "epoch": 0.273972602739726, + "eval_loss": 2.1425344944000244, + "eval_runtime": 37.1683, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 250 + }, + { + "epoch": 0.27506849315068493, + "grad_norm": 2.7668282985687256, + "learning_rate": 4.3248175182481755e-06, + "loss": 1.9345, + "step": 251 + }, + { + "epoch": 0.27616438356164386, + "grad_norm": 2.5530643463134766, + "learning_rate": 4.343065693430658e-06, + "loss": 1.994, + "step": 252 + }, + { + "epoch": 0.2772602739726027, + "grad_norm": 2.8284707069396973, + "learning_rate": 4.361313868613139e-06, + "loss": 1.982, + "step": 253 + }, + { + "epoch": 0.27835616438356164, + "grad_norm": 3.2699639797210693, + "learning_rate": 4.379562043795621e-06, + "loss": 1.9636, + "step": 254 + }, + { + "epoch": 0.27945205479452057, + "grad_norm": 3.0885937213897705, + "learning_rate": 4.397810218978102e-06, + "loss": 2.0257, + "step": 255 + }, + { + "epoch": 0.28054794520547943, + "grad_norm": 2.753086805343628, + "learning_rate": 4.416058394160584e-06, + "loss": 1.9844, + "step": 256 + }, + { + "epoch": 0.28164383561643835, + "grad_norm": 2.7163126468658447, + "learning_rate": 4.434306569343066e-06, + "loss": 1.9733, + "step": 257 + }, + { + "epoch": 0.2827397260273973, + "grad_norm": 3.2078559398651123, + "learning_rate": 4.452554744525548e-06, + "loss": 1.9609, + "step": 258 + }, + { + "epoch": 0.28383561643835614, + "grad_norm": 2.6238415241241455, + "learning_rate": 4.4708029197080295e-06, + "loss": 2.0062, + "step": 259 + }, + { + "epoch": 0.28493150684931506, + "grad_norm": 2.8016085624694824, + "learning_rate": 4.489051094890512e-06, + "loss": 2.004, + "step": 260 + }, + { + "epoch": 0.28493150684931506, + "eval_loss": 2.1314096450805664, + "eval_runtime": 37.181, + "eval_samples_per_second": 26.895, + "eval_steps_per_second": 1.13, + "step": 260 + }, + { + "epoch": 0.286027397260274, + "grad_norm": 3.6612119674682617, + "learning_rate": 4.507299270072993e-06, + "loss": 2.0389, + "step": 261 + }, + { + "epoch": 0.28712328767123285, + "grad_norm": 3.0465657711029053, + "learning_rate": 4.525547445255475e-06, + "loss": 1.9978, + "step": 262 + }, + { + "epoch": 0.2882191780821918, + "grad_norm": 2.766305685043335, + "learning_rate": 4.543795620437956e-06, + "loss": 1.9734, + "step": 263 + }, + { + "epoch": 0.2893150684931507, + "grad_norm": 2.971726655960083, + "learning_rate": 4.562043795620438e-06, + "loss": 1.9313, + "step": 264 + }, + { + "epoch": 0.29041095890410956, + "grad_norm": 2.9012374877929688, + "learning_rate": 4.580291970802921e-06, + "loss": 2.031, + "step": 265 + }, + { + "epoch": 0.2915068493150685, + "grad_norm": 2.5543956756591797, + "learning_rate": 4.598540145985402e-06, + "loss": 2.0319, + "step": 266 + }, + { + "epoch": 0.2926027397260274, + "grad_norm": 3.143434762954712, + "learning_rate": 4.6167883211678835e-06, + "loss": 1.9949, + "step": 267 + }, + { + "epoch": 0.2936986301369863, + "grad_norm": 2.559908628463745, + "learning_rate": 4.635036496350365e-06, + "loss": 1.9567, + "step": 268 + }, + { + "epoch": 0.2947945205479452, + "grad_norm": 3.090074062347412, + "learning_rate": 4.653284671532847e-06, + "loss": 1.9552, + "step": 269 + }, + { + "epoch": 0.2958904109589041, + "grad_norm": 2.9278085231781006, + "learning_rate": 4.671532846715329e-06, + "loss": 2.0083, + "step": 270 + }, + { + "epoch": 0.2958904109589041, + "eval_loss": 2.120840072631836, + "eval_runtime": 37.2226, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 270 + }, + { + "epoch": 0.29698630136986304, + "grad_norm": 3.458752155303955, + "learning_rate": 4.689781021897811e-06, + "loss": 1.9958, + "step": 271 + }, + { + "epoch": 0.2980821917808219, + "grad_norm": 3.8958938121795654, + "learning_rate": 4.708029197080292e-06, + "loss": 1.969, + "step": 272 + }, + { + "epoch": 0.2991780821917808, + "grad_norm": 2.8383052349090576, + "learning_rate": 4.726277372262774e-06, + "loss": 2.0298, + "step": 273 + }, + { + "epoch": 0.30027397260273975, + "grad_norm": 4.117269039154053, + "learning_rate": 4.744525547445255e-06, + "loss": 1.9996, + "step": 274 + }, + { + "epoch": 0.3013698630136986, + "grad_norm": 3.134629249572754, + "learning_rate": 4.7627737226277375e-06, + "loss": 1.9275, + "step": 275 + }, + { + "epoch": 0.30246575342465754, + "grad_norm": 2.904829978942871, + "learning_rate": 4.78102189781022e-06, + "loss": 1.9821, + "step": 276 + }, + { + "epoch": 0.30356164383561646, + "grad_norm": 3.429172992706299, + "learning_rate": 4.799270072992701e-06, + "loss": 1.9963, + "step": 277 + }, + { + "epoch": 0.3046575342465753, + "grad_norm": 2.6045989990234375, + "learning_rate": 4.8175182481751835e-06, + "loss": 1.9388, + "step": 278 + }, + { + "epoch": 0.30575342465753425, + "grad_norm": 3.114208459854126, + "learning_rate": 4.835766423357665e-06, + "loss": 1.9813, + "step": 279 + }, + { + "epoch": 0.30684931506849317, + "grad_norm": 2.4997620582580566, + "learning_rate": 4.854014598540146e-06, + "loss": 1.9761, + "step": 280 + }, + { + "epoch": 0.30684931506849317, + "eval_loss": 2.1167938709259033, + "eval_runtime": 37.1676, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 280 + }, + { + "epoch": 0.30794520547945203, + "grad_norm": 2.8907406330108643, + "learning_rate": 4.872262773722628e-06, + "loss": 1.9097, + "step": 281 + }, + { + "epoch": 0.30904109589041096, + "grad_norm": 2.529203176498413, + "learning_rate": 4.89051094890511e-06, + "loss": 1.9296, + "step": 282 + }, + { + "epoch": 0.3101369863013699, + "grad_norm": 2.6994636058807373, + "learning_rate": 4.9087591240875915e-06, + "loss": 1.9099, + "step": 283 + }, + { + "epoch": 0.31123287671232874, + "grad_norm": 3.2694296836853027, + "learning_rate": 4.927007299270074e-06, + "loss": 1.9545, + "step": 284 + }, + { + "epoch": 0.31232876712328766, + "grad_norm": 2.5501177310943604, + "learning_rate": 4.945255474452555e-06, + "loss": 1.9679, + "step": 285 + }, + { + "epoch": 0.3134246575342466, + "grad_norm": 3.156254291534424, + "learning_rate": 4.963503649635037e-06, + "loss": 1.9237, + "step": 286 + }, + { + "epoch": 0.31452054794520545, + "grad_norm": 2.8147518634796143, + "learning_rate": 4.981751824817518e-06, + "loss": 2.0199, + "step": 287 + }, + { + "epoch": 0.3156164383561644, + "grad_norm": 2.8747735023498535, + "learning_rate": 5e-06, + "loss": 2.0092, + "step": 288 + }, + { + "epoch": 0.3167123287671233, + "grad_norm": 2.829530954360962, + "learning_rate": 4.997969130787978e-06, + "loss": 1.9589, + "step": 289 + }, + { + "epoch": 0.3178082191780822, + "grad_norm": 2.843920946121216, + "learning_rate": 4.995938261575955e-06, + "loss": 1.9746, + "step": 290 + }, + { + "epoch": 0.3178082191780822, + "eval_loss": 2.099478244781494, + "eval_runtime": 37.2052, + "eval_samples_per_second": 26.878, + "eval_steps_per_second": 1.129, + "step": 290 + }, + { + "epoch": 0.3189041095890411, + "grad_norm": 2.5574231147766113, + "learning_rate": 4.993907392363932e-06, + "loss": 1.8964, + "step": 291 + }, + { + "epoch": 0.32, + "grad_norm": 2.888542890548706, + "learning_rate": 4.9918765231519095e-06, + "loss": 1.9762, + "step": 292 + }, + { + "epoch": 0.32109589041095893, + "grad_norm": 2.7378697395324707, + "learning_rate": 4.989845653939887e-06, + "loss": 1.9051, + "step": 293 + }, + { + "epoch": 0.3221917808219178, + "grad_norm": 3.1764118671417236, + "learning_rate": 4.987814784727864e-06, + "loss": 1.9309, + "step": 294 + }, + { + "epoch": 0.3232876712328767, + "grad_norm": 3.3455569744110107, + "learning_rate": 4.985783915515841e-06, + "loss": 1.9348, + "step": 295 + }, + { + "epoch": 0.32438356164383564, + "grad_norm": 2.8268158435821533, + "learning_rate": 4.983753046303819e-06, + "loss": 1.9167, + "step": 296 + }, + { + "epoch": 0.3254794520547945, + "grad_norm": 2.625690460205078, + "learning_rate": 4.981722177091796e-06, + "loss": 1.9385, + "step": 297 + }, + { + "epoch": 0.3265753424657534, + "grad_norm": 2.5886993408203125, + "learning_rate": 4.979691307879773e-06, + "loss": 1.9129, + "step": 298 + }, + { + "epoch": 0.32767123287671235, + "grad_norm": 2.554054021835327, + "learning_rate": 4.9776604386677504e-06, + "loss": 1.9424, + "step": 299 + }, + { + "epoch": 0.3287671232876712, + "grad_norm": 2.731053113937378, + "learning_rate": 4.975629569455728e-06, + "loss": 1.9453, + "step": 300 + }, + { + "epoch": 0.3287671232876712, + "eval_loss": 2.095012664794922, + "eval_runtime": 37.2385, + "eval_samples_per_second": 26.854, + "eval_steps_per_second": 1.128, + "step": 300 + }, + { + "epoch": 0.32986301369863014, + "grad_norm": 2.7957098484039307, + "learning_rate": 4.973598700243705e-06, + "loss": 1.9849, + "step": 301 + }, + { + "epoch": 0.33095890410958906, + "grad_norm": 2.9258270263671875, + "learning_rate": 4.971567831031681e-06, + "loss": 1.8948, + "step": 302 + }, + { + "epoch": 0.3320547945205479, + "grad_norm": 3.202796459197998, + "learning_rate": 4.9695369618196595e-06, + "loss": 1.9141, + "step": 303 + }, + { + "epoch": 0.33315068493150685, + "grad_norm": 2.5909249782562256, + "learning_rate": 4.967506092607636e-06, + "loss": 1.7926, + "step": 304 + }, + { + "epoch": 0.33424657534246577, + "grad_norm": 2.7584893703460693, + "learning_rate": 4.965475223395614e-06, + "loss": 1.8774, + "step": 305 + }, + { + "epoch": 0.33534246575342463, + "grad_norm": 2.855374336242676, + "learning_rate": 4.9634443541835905e-06, + "loss": 1.9049, + "step": 306 + }, + { + "epoch": 0.33643835616438356, + "grad_norm": 2.845473289489746, + "learning_rate": 4.961413484971569e-06, + "loss": 1.8887, + "step": 307 + }, + { + "epoch": 0.3375342465753425, + "grad_norm": 2.7958481311798096, + "learning_rate": 4.959382615759545e-06, + "loss": 1.9363, + "step": 308 + }, + { + "epoch": 0.33863013698630134, + "grad_norm": 2.7589993476867676, + "learning_rate": 4.957351746547523e-06, + "loss": 1.916, + "step": 309 + }, + { + "epoch": 0.33972602739726027, + "grad_norm": 2.8347692489624023, + "learning_rate": 4.9553208773355e-06, + "loss": 1.9018, + "step": 310 + }, + { + "epoch": 0.33972602739726027, + "eval_loss": 2.0926122665405273, + "eval_runtime": 36.8827, + "eval_samples_per_second": 27.113, + "eval_steps_per_second": 1.139, + "step": 310 + }, + { + "epoch": 0.3408219178082192, + "grad_norm": 2.9393861293792725, + "learning_rate": 4.953290008123478e-06, + "loss": 2.039, + "step": 311 + }, + { + "epoch": 0.3419178082191781, + "grad_norm": 3.011475086212158, + "learning_rate": 4.951259138911454e-06, + "loss": 1.923, + "step": 312 + }, + { + "epoch": 0.343013698630137, + "grad_norm": 2.689653158187866, + "learning_rate": 4.9492282696994315e-06, + "loss": 1.8759, + "step": 313 + }, + { + "epoch": 0.3441095890410959, + "grad_norm": 2.696784257888794, + "learning_rate": 4.947197400487409e-06, + "loss": 1.9979, + "step": 314 + }, + { + "epoch": 0.3452054794520548, + "grad_norm": 3.7033591270446777, + "learning_rate": 4.945166531275386e-06, + "loss": 1.9241, + "step": 315 + }, + { + "epoch": 0.3463013698630137, + "grad_norm": 2.8621737957000732, + "learning_rate": 4.943135662063363e-06, + "loss": 1.8909, + "step": 316 + }, + { + "epoch": 0.3473972602739726, + "grad_norm": 2.570342779159546, + "learning_rate": 4.9411047928513406e-06, + "loss": 1.9197, + "step": 317 + }, + { + "epoch": 0.34849315068493153, + "grad_norm": 3.081667184829712, + "learning_rate": 4.939073923639318e-06, + "loss": 1.9666, + "step": 318 + }, + { + "epoch": 0.3495890410958904, + "grad_norm": 2.598898410797119, + "learning_rate": 4.937043054427295e-06, + "loss": 1.9616, + "step": 319 + }, + { + "epoch": 0.3506849315068493, + "grad_norm": 2.8304622173309326, + "learning_rate": 4.935012185215272e-06, + "loss": 1.8894, + "step": 320 + }, + { + "epoch": 0.3506849315068493, + "eval_loss": 2.0831871032714844, + "eval_runtime": 37.0685, + "eval_samples_per_second": 26.977, + "eval_steps_per_second": 1.133, + "step": 320 + }, + { + "epoch": 0.35178082191780824, + "grad_norm": 2.767806053161621, + "learning_rate": 4.93298131600325e-06, + "loss": 1.8689, + "step": 321 + }, + { + "epoch": 0.3528767123287671, + "grad_norm": 2.7915256023406982, + "learning_rate": 4.930950446791227e-06, + "loss": 1.9446, + "step": 322 + }, + { + "epoch": 0.353972602739726, + "grad_norm": 2.5830628871917725, + "learning_rate": 4.928919577579204e-06, + "loss": 1.9165, + "step": 323 + }, + { + "epoch": 0.35506849315068495, + "grad_norm": 2.7124366760253906, + "learning_rate": 4.9268887083671815e-06, + "loss": 1.937, + "step": 324 + }, + { + "epoch": 0.3561643835616438, + "grad_norm": 2.853339910507202, + "learning_rate": 4.924857839155159e-06, + "loss": 1.9088, + "step": 325 + }, + { + "epoch": 0.35726027397260274, + "grad_norm": 2.569329023361206, + "learning_rate": 4.922826969943136e-06, + "loss": 1.8973, + "step": 326 + }, + { + "epoch": 0.35835616438356166, + "grad_norm": 2.7952170372009277, + "learning_rate": 4.920796100731113e-06, + "loss": 1.9285, + "step": 327 + }, + { + "epoch": 0.3594520547945205, + "grad_norm": 2.518095016479492, + "learning_rate": 4.918765231519091e-06, + "loss": 1.8472, + "step": 328 + }, + { + "epoch": 0.36054794520547945, + "grad_norm": 2.5369744300842285, + "learning_rate": 4.916734362307068e-06, + "loss": 1.9084, + "step": 329 + }, + { + "epoch": 0.36164383561643837, + "grad_norm": 2.834768056869507, + "learning_rate": 4.914703493095045e-06, + "loss": 1.8692, + "step": 330 + }, + { + "epoch": 0.36164383561643837, + "eval_loss": 2.0748064517974854, + "eval_runtime": 37.1611, + "eval_samples_per_second": 26.91, + "eval_steps_per_second": 1.13, + "step": 330 + }, + { + "epoch": 0.36273972602739724, + "grad_norm": 2.496169090270996, + "learning_rate": 4.912672623883022e-06, + "loss": 1.8688, + "step": 331 + }, + { + "epoch": 0.36383561643835616, + "grad_norm": 2.615295886993408, + "learning_rate": 4.910641754671e-06, + "loss": 1.8616, + "step": 332 + }, + { + "epoch": 0.3649315068493151, + "grad_norm": 2.8769407272338867, + "learning_rate": 4.908610885458977e-06, + "loss": 1.9066, + "step": 333 + }, + { + "epoch": 0.36602739726027395, + "grad_norm": 2.5247085094451904, + "learning_rate": 4.906580016246954e-06, + "loss": 1.8801, + "step": 334 + }, + { + "epoch": 0.36712328767123287, + "grad_norm": 3.1026415824890137, + "learning_rate": 4.9045491470349315e-06, + "loss": 1.9397, + "step": 335 + }, + { + "epoch": 0.3682191780821918, + "grad_norm": 3.14855694770813, + "learning_rate": 4.902518277822909e-06, + "loss": 1.917, + "step": 336 + }, + { + "epoch": 0.3693150684931507, + "grad_norm": 2.33418345451355, + "learning_rate": 4.900487408610886e-06, + "loss": 1.9197, + "step": 337 + }, + { + "epoch": 0.3704109589041096, + "grad_norm": 3.7271599769592285, + "learning_rate": 4.898456539398863e-06, + "loss": 1.9312, + "step": 338 + }, + { + "epoch": 0.3715068493150685, + "grad_norm": 3.115157127380371, + "learning_rate": 4.89642567018684e-06, + "loss": 1.949, + "step": 339 + }, + { + "epoch": 0.3726027397260274, + "grad_norm": 2.969817638397217, + "learning_rate": 4.894394800974818e-06, + "loss": 1.9971, + "step": 340 + }, + { + "epoch": 0.3726027397260274, + "eval_loss": 2.068490743637085, + "eval_runtime": 37.2019, + "eval_samples_per_second": 26.88, + "eval_steps_per_second": 1.129, + "step": 340 + }, + { + "epoch": 0.3736986301369863, + "grad_norm": 3.2445967197418213, + "learning_rate": 4.892363931762794e-06, + "loss": 1.8629, + "step": 341 + }, + { + "epoch": 0.3747945205479452, + "grad_norm": 2.711249589920044, + "learning_rate": 4.8903330625507725e-06, + "loss": 1.9455, + "step": 342 + }, + { + "epoch": 0.37589041095890413, + "grad_norm": 2.970963716506958, + "learning_rate": 4.888302193338749e-06, + "loss": 1.8738, + "step": 343 + }, + { + "epoch": 0.376986301369863, + "grad_norm": 2.625316858291626, + "learning_rate": 4.886271324126727e-06, + "loss": 1.8396, + "step": 344 + }, + { + "epoch": 0.3780821917808219, + "grad_norm": 2.5299627780914307, + "learning_rate": 4.8842404549147034e-06, + "loss": 1.9576, + "step": 345 + }, + { + "epoch": 0.37917808219178084, + "grad_norm": 2.6957638263702393, + "learning_rate": 4.8822095857026816e-06, + "loss": 1.9354, + "step": 346 + }, + { + "epoch": 0.3802739726027397, + "grad_norm": 2.644958734512329, + "learning_rate": 4.880178716490658e-06, + "loss": 1.8891, + "step": 347 + }, + { + "epoch": 0.38136986301369863, + "grad_norm": 2.7591962814331055, + "learning_rate": 4.878147847278636e-06, + "loss": 1.8545, + "step": 348 + }, + { + "epoch": 0.38246575342465755, + "grad_norm": 2.7353060245513916, + "learning_rate": 4.8761169780666125e-06, + "loss": 1.8584, + "step": 349 + }, + { + "epoch": 0.3835616438356164, + "grad_norm": 2.9095821380615234, + "learning_rate": 4.874086108854591e-06, + "loss": 1.8294, + "step": 350 + }, + { + "epoch": 0.3835616438356164, + "eval_loss": 2.0575802326202393, + "eval_runtime": 37.2237, + "eval_samples_per_second": 26.865, + "eval_steps_per_second": 1.128, + "step": 350 + }, + { + "epoch": 0.38465753424657534, + "grad_norm": 3.0609660148620605, + "learning_rate": 4.872055239642567e-06, + "loss": 1.8584, + "step": 351 + }, + { + "epoch": 0.38575342465753426, + "grad_norm": 3.1720101833343506, + "learning_rate": 4.870024370430544e-06, + "loss": 1.8226, + "step": 352 + }, + { + "epoch": 0.3868493150684931, + "grad_norm": 2.734989881515503, + "learning_rate": 4.867993501218522e-06, + "loss": 1.8816, + "step": 353 + }, + { + "epoch": 0.38794520547945205, + "grad_norm": 2.7359373569488525, + "learning_rate": 4.865962632006499e-06, + "loss": 1.9108, + "step": 354 + }, + { + "epoch": 0.38904109589041097, + "grad_norm": 3.5148444175720215, + "learning_rate": 4.863931762794476e-06, + "loss": 1.8402, + "step": 355 + }, + { + "epoch": 0.39013698630136984, + "grad_norm": 2.5981862545013428, + "learning_rate": 4.8619008935824535e-06, + "loss": 1.8873, + "step": 356 + }, + { + "epoch": 0.39123287671232876, + "grad_norm": 3.42146372795105, + "learning_rate": 4.859870024370431e-06, + "loss": 1.9222, + "step": 357 + }, + { + "epoch": 0.3923287671232877, + "grad_norm": 2.795802354812622, + "learning_rate": 4.857839155158408e-06, + "loss": 1.9199, + "step": 358 + }, + { + "epoch": 0.3934246575342466, + "grad_norm": 3.0176706314086914, + "learning_rate": 4.855808285946385e-06, + "loss": 1.9374, + "step": 359 + }, + { + "epoch": 0.39452054794520547, + "grad_norm": 2.607785940170288, + "learning_rate": 4.8537774167343626e-06, + "loss": 1.8891, + "step": 360 + }, + { + "epoch": 0.39452054794520547, + "eval_loss": 2.0473363399505615, + "eval_runtime": 37.2406, + "eval_samples_per_second": 26.852, + "eval_steps_per_second": 1.128, + "step": 360 + }, + { + "epoch": 0.3956164383561644, + "grad_norm": 3.1632132530212402, + "learning_rate": 4.85174654752234e-06, + "loss": 1.8772, + "step": 361 + }, + { + "epoch": 0.3967123287671233, + "grad_norm": 2.6246979236602783, + "learning_rate": 4.849715678310317e-06, + "loss": 1.8952, + "step": 362 + }, + { + "epoch": 0.3978082191780822, + "grad_norm": 2.4879660606384277, + "learning_rate": 4.847684809098294e-06, + "loss": 1.9017, + "step": 363 + }, + { + "epoch": 0.3989041095890411, + "grad_norm": 3.2623815536499023, + "learning_rate": 4.845653939886272e-06, + "loss": 1.8788, + "step": 364 + }, + { + "epoch": 0.4, + "grad_norm": 2.6892051696777344, + "learning_rate": 4.843623070674249e-06, + "loss": 1.8912, + "step": 365 + }, + { + "epoch": 0.4010958904109589, + "grad_norm": 2.720777750015259, + "learning_rate": 4.841592201462226e-06, + "loss": 1.8763, + "step": 366 + }, + { + "epoch": 0.4021917808219178, + "grad_norm": 2.4111084938049316, + "learning_rate": 4.8395613322502035e-06, + "loss": 1.9202, + "step": 367 + }, + { + "epoch": 0.40328767123287673, + "grad_norm": 2.875309944152832, + "learning_rate": 4.837530463038181e-06, + "loss": 1.8335, + "step": 368 + }, + { + "epoch": 0.4043835616438356, + "grad_norm": 2.7094228267669678, + "learning_rate": 4.835499593826158e-06, + "loss": 1.8571, + "step": 369 + }, + { + "epoch": 0.4054794520547945, + "grad_norm": 2.55806565284729, + "learning_rate": 4.833468724614135e-06, + "loss": 1.9049, + "step": 370 + }, + { + "epoch": 0.4054794520547945, + "eval_loss": 2.0418052673339844, + "eval_runtime": 37.2181, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 1.128, + "step": 370 + }, + { + "epoch": 0.40657534246575344, + "grad_norm": 2.5894389152526855, + "learning_rate": 4.831437855402113e-06, + "loss": 1.8717, + "step": 371 + }, + { + "epoch": 0.4076712328767123, + "grad_norm": 2.5270323753356934, + "learning_rate": 4.82940698619009e-06, + "loss": 1.7785, + "step": 372 + }, + { + "epoch": 0.40876712328767123, + "grad_norm": 2.4306442737579346, + "learning_rate": 4.827376116978067e-06, + "loss": 1.8603, + "step": 373 + }, + { + "epoch": 0.40986301369863015, + "grad_norm": 2.7987163066864014, + "learning_rate": 4.8253452477660444e-06, + "loss": 1.8907, + "step": 374 + }, + { + "epoch": 0.410958904109589, + "grad_norm": 2.455786943435669, + "learning_rate": 4.823314378554022e-06, + "loss": 1.8965, + "step": 375 + }, + { + "epoch": 0.41205479452054794, + "grad_norm": 2.573777198791504, + "learning_rate": 4.821283509341998e-06, + "loss": 1.8986, + "step": 376 + }, + { + "epoch": 0.41315068493150686, + "grad_norm": 2.6333072185516357, + "learning_rate": 4.819252640129976e-06, + "loss": 1.8622, + "step": 377 + }, + { + "epoch": 0.41424657534246573, + "grad_norm": 2.5853562355041504, + "learning_rate": 4.817221770917953e-06, + "loss": 1.8671, + "step": 378 + }, + { + "epoch": 0.41534246575342465, + "grad_norm": 2.583556890487671, + "learning_rate": 4.815190901705931e-06, + "loss": 1.8307, + "step": 379 + }, + { + "epoch": 0.41643835616438357, + "grad_norm": 2.544290542602539, + "learning_rate": 4.813160032493907e-06, + "loss": 1.884, + "step": 380 + }, + { + "epoch": 0.41643835616438357, + "eval_loss": 2.0327165126800537, + "eval_runtime": 37.234, + "eval_samples_per_second": 26.857, + "eval_steps_per_second": 1.128, + "step": 380 + }, + { + "epoch": 0.41753424657534244, + "grad_norm": 2.63576340675354, + "learning_rate": 4.811129163281885e-06, + "loss": 1.842, + "step": 381 + }, + { + "epoch": 0.41863013698630136, + "grad_norm": 2.5061604976654053, + "learning_rate": 4.809098294069862e-06, + "loss": 1.8964, + "step": 382 + }, + { + "epoch": 0.4197260273972603, + "grad_norm": 2.737086534500122, + "learning_rate": 4.80706742485784e-06, + "loss": 1.8802, + "step": 383 + }, + { + "epoch": 0.4208219178082192, + "grad_norm": 2.686113119125366, + "learning_rate": 4.805036555645816e-06, + "loss": 1.9279, + "step": 384 + }, + { + "epoch": 0.42191780821917807, + "grad_norm": 2.530296802520752, + "learning_rate": 4.8030056864337945e-06, + "loss": 1.8442, + "step": 385 + }, + { + "epoch": 0.423013698630137, + "grad_norm": 2.854804277420044, + "learning_rate": 4.800974817221771e-06, + "loss": 1.9247, + "step": 386 + }, + { + "epoch": 0.4241095890410959, + "grad_norm": 2.6104133129119873, + "learning_rate": 4.798943948009749e-06, + "loss": 1.8614, + "step": 387 + }, + { + "epoch": 0.4252054794520548, + "grad_norm": 2.8105721473693848, + "learning_rate": 4.7969130787977254e-06, + "loss": 1.8743, + "step": 388 + }, + { + "epoch": 0.4263013698630137, + "grad_norm": 2.4551901817321777, + "learning_rate": 4.794882209585703e-06, + "loss": 1.8735, + "step": 389 + }, + { + "epoch": 0.4273972602739726, + "grad_norm": 2.4942400455474854, + "learning_rate": 4.79285134037368e-06, + "loss": 1.8989, + "step": 390 + }, + { + "epoch": 0.4273972602739726, + "eval_loss": 2.028289794921875, + "eval_runtime": 37.1679, + "eval_samples_per_second": 26.905, + "eval_steps_per_second": 1.13, + "step": 390 + }, + { + "epoch": 0.4284931506849315, + "grad_norm": 3.1998565196990967, + "learning_rate": 4.790820471161657e-06, + "loss": 1.8479, + "step": 391 + }, + { + "epoch": 0.4295890410958904, + "grad_norm": 2.4295566082000732, + "learning_rate": 4.7887896019496345e-06, + "loss": 1.8474, + "step": 392 + }, + { + "epoch": 0.43068493150684933, + "grad_norm": 2.6794519424438477, + "learning_rate": 4.786758732737612e-06, + "loss": 1.848, + "step": 393 + }, + { + "epoch": 0.4317808219178082, + "grad_norm": 2.7655680179595947, + "learning_rate": 4.784727863525589e-06, + "loss": 1.8989, + "step": 394 + }, + { + "epoch": 0.4328767123287671, + "grad_norm": 2.3869338035583496, + "learning_rate": 4.782696994313566e-06, + "loss": 1.8028, + "step": 395 + }, + { + "epoch": 0.43397260273972604, + "grad_norm": 2.776850461959839, + "learning_rate": 4.780666125101544e-06, + "loss": 1.8631, + "step": 396 + }, + { + "epoch": 0.4350684931506849, + "grad_norm": 2.678173303604126, + "learning_rate": 4.778635255889521e-06, + "loss": 1.804, + "step": 397 + }, + { + "epoch": 0.43616438356164383, + "grad_norm": 2.5821075439453125, + "learning_rate": 4.776604386677498e-06, + "loss": 1.8311, + "step": 398 + }, + { + "epoch": 0.43726027397260275, + "grad_norm": 2.8446664810180664, + "learning_rate": 4.7745735174654755e-06, + "loss": 1.8461, + "step": 399 + }, + { + "epoch": 0.4383561643835616, + "grad_norm": 2.760930061340332, + "learning_rate": 4.772542648253453e-06, + "loss": 1.9053, + "step": 400 + }, + { + "epoch": 0.4383561643835616, + "eval_loss": 2.019589900970459, + "eval_runtime": 37.1584, + "eval_samples_per_second": 26.912, + "eval_steps_per_second": 1.13, + "step": 400 + }, + { + "epoch": 0.43945205479452054, + "grad_norm": 2.6788735389709473, + "learning_rate": 4.77051177904143e-06, + "loss": 1.8893, + "step": 401 + }, + { + "epoch": 0.44054794520547946, + "grad_norm": 2.4157094955444336, + "learning_rate": 4.768480909829407e-06, + "loss": 1.8156, + "step": 402 + }, + { + "epoch": 0.44164383561643833, + "grad_norm": 2.675287961959839, + "learning_rate": 4.766450040617385e-06, + "loss": 1.8701, + "step": 403 + }, + { + "epoch": 0.44273972602739725, + "grad_norm": 2.470655918121338, + "learning_rate": 4.764419171405362e-06, + "loss": 1.8435, + "step": 404 + }, + { + "epoch": 0.4438356164383562, + "grad_norm": 2.711153984069824, + "learning_rate": 4.762388302193339e-06, + "loss": 1.9061, + "step": 405 + }, + { + "epoch": 0.4449315068493151, + "grad_norm": 3.007571220397949, + "learning_rate": 4.760357432981316e-06, + "loss": 1.867, + "step": 406 + }, + { + "epoch": 0.44602739726027396, + "grad_norm": 2.5110905170440674, + "learning_rate": 4.758326563769294e-06, + "loss": 1.8559, + "step": 407 + }, + { + "epoch": 0.4471232876712329, + "grad_norm": 2.5284948348999023, + "learning_rate": 4.756295694557271e-06, + "loss": 1.8635, + "step": 408 + }, + { + "epoch": 0.4482191780821918, + "grad_norm": 2.6240103244781494, + "learning_rate": 4.754264825345248e-06, + "loss": 1.8642, + "step": 409 + }, + { + "epoch": 0.44931506849315067, + "grad_norm": 2.4638724327087402, + "learning_rate": 4.7522339561332255e-06, + "loss": 1.8007, + "step": 410 + }, + { + "epoch": 0.44931506849315067, + "eval_loss": 2.017984628677368, + "eval_runtime": 36.7959, + "eval_samples_per_second": 27.177, + "eval_steps_per_second": 1.141, + "step": 410 + }, + { + "epoch": 0.4504109589041096, + "grad_norm": 2.5062263011932373, + "learning_rate": 4.750203086921203e-06, + "loss": 1.8261, + "step": 411 + }, + { + "epoch": 0.4515068493150685, + "grad_norm": 2.77105712890625, + "learning_rate": 4.74817221770918e-06, + "loss": 1.772, + "step": 412 + }, + { + "epoch": 0.4526027397260274, + "grad_norm": 2.376661539077759, + "learning_rate": 4.7461413484971565e-06, + "loss": 1.8772, + "step": 413 + }, + { + "epoch": 0.4536986301369863, + "grad_norm": 2.664144277572632, + "learning_rate": 4.744110479285135e-06, + "loss": 1.9047, + "step": 414 + }, + { + "epoch": 0.4547945205479452, + "grad_norm": 2.4807093143463135, + "learning_rate": 4.742079610073111e-06, + "loss": 1.7809, + "step": 415 + }, + { + "epoch": 0.4558904109589041, + "grad_norm": 2.4022881984710693, + "learning_rate": 4.740048740861089e-06, + "loss": 1.807, + "step": 416 + }, + { + "epoch": 0.456986301369863, + "grad_norm": 2.475606679916382, + "learning_rate": 4.738017871649066e-06, + "loss": 1.8418, + "step": 417 + }, + { + "epoch": 0.45808219178082193, + "grad_norm": 2.6839230060577393, + "learning_rate": 4.735987002437044e-06, + "loss": 1.8537, + "step": 418 + }, + { + "epoch": 0.4591780821917808, + "grad_norm": 3.063843011856079, + "learning_rate": 4.73395613322502e-06, + "loss": 1.8906, + "step": 419 + }, + { + "epoch": 0.4602739726027397, + "grad_norm": 2.7786505222320557, + "learning_rate": 4.731925264012998e-06, + "loss": 1.9112, + "step": 420 + }, + { + "epoch": 0.4602739726027397, + "eval_loss": 2.012242317199707, + "eval_runtime": 36.9477, + "eval_samples_per_second": 27.065, + "eval_steps_per_second": 1.137, + "step": 420 + }, + { + "epoch": 0.46136986301369864, + "grad_norm": 2.669280767440796, + "learning_rate": 4.729894394800975e-06, + "loss": 1.8901, + "step": 421 + }, + { + "epoch": 0.4624657534246575, + "grad_norm": 2.86869478225708, + "learning_rate": 4.727863525588953e-06, + "loss": 1.8101, + "step": 422 + }, + { + "epoch": 0.46356164383561643, + "grad_norm": 2.6715657711029053, + "learning_rate": 4.725832656376929e-06, + "loss": 1.7711, + "step": 423 + }, + { + "epoch": 0.46465753424657535, + "grad_norm": 2.894533395767212, + "learning_rate": 4.723801787164907e-06, + "loss": 1.815, + "step": 424 + }, + { + "epoch": 0.4657534246575342, + "grad_norm": 3.145872116088867, + "learning_rate": 4.721770917952885e-06, + "loss": 1.8788, + "step": 425 + }, + { + "epoch": 0.46684931506849314, + "grad_norm": 2.6268310546875, + "learning_rate": 4.719740048740861e-06, + "loss": 1.8018, + "step": 426 + }, + { + "epoch": 0.46794520547945206, + "grad_norm": 2.9428443908691406, + "learning_rate": 4.717709179528839e-06, + "loss": 1.8431, + "step": 427 + }, + { + "epoch": 0.469041095890411, + "grad_norm": 2.6178042888641357, + "learning_rate": 4.715678310316816e-06, + "loss": 1.8719, + "step": 428 + }, + { + "epoch": 0.47013698630136985, + "grad_norm": 2.695261001586914, + "learning_rate": 4.713647441104794e-06, + "loss": 1.8506, + "step": 429 + }, + { + "epoch": 0.4712328767123288, + "grad_norm": 2.529327630996704, + "learning_rate": 4.71161657189277e-06, + "loss": 1.8033, + "step": 430 + }, + { + "epoch": 0.4712328767123288, + "eval_loss": 2.0065560340881348, + "eval_runtime": 37.0073, + "eval_samples_per_second": 27.022, + "eval_steps_per_second": 1.135, + "step": 430 + }, + { + "epoch": 0.4723287671232877, + "grad_norm": 2.7176175117492676, + "learning_rate": 4.709585702680748e-06, + "loss": 1.801, + "step": 431 + }, + { + "epoch": 0.47342465753424656, + "grad_norm": 2.439375638961792, + "learning_rate": 4.707554833468725e-06, + "loss": 1.8774, + "step": 432 + }, + { + "epoch": 0.4745205479452055, + "grad_norm": 2.583240509033203, + "learning_rate": 4.705523964256703e-06, + "loss": 1.9315, + "step": 433 + }, + { + "epoch": 0.4756164383561644, + "grad_norm": 2.486970901489258, + "learning_rate": 4.703493095044679e-06, + "loss": 1.8089, + "step": 434 + }, + { + "epoch": 0.4767123287671233, + "grad_norm": 2.729233503341675, + "learning_rate": 4.701462225832657e-06, + "loss": 1.832, + "step": 435 + }, + { + "epoch": 0.4778082191780822, + "grad_norm": 2.707348346710205, + "learning_rate": 4.699431356620634e-06, + "loss": 1.7688, + "step": 436 + }, + { + "epoch": 0.4789041095890411, + "grad_norm": 2.61710786819458, + "learning_rate": 4.697400487408612e-06, + "loss": 1.9149, + "step": 437 + }, + { + "epoch": 0.48, + "grad_norm": 2.4853103160858154, + "learning_rate": 4.695369618196588e-06, + "loss": 1.8606, + "step": 438 + }, + { + "epoch": 0.4810958904109589, + "grad_norm": 2.7008261680603027, + "learning_rate": 4.693338748984566e-06, + "loss": 1.8466, + "step": 439 + }, + { + "epoch": 0.4821917808219178, + "grad_norm": 2.7126548290252686, + "learning_rate": 4.691307879772543e-06, + "loss": 1.8176, + "step": 440 + }, + { + "epoch": 0.4821917808219178, + "eval_loss": 1.9978634119033813, + "eval_runtime": 37.1045, + "eval_samples_per_second": 26.951, + "eval_steps_per_second": 1.132, + "step": 440 + }, + { + "epoch": 0.4832876712328767, + "grad_norm": 2.4356260299682617, + "learning_rate": 4.68927701056052e-06, + "loss": 1.7871, + "step": 441 + }, + { + "epoch": 0.4843835616438356, + "grad_norm": 2.533386468887329, + "learning_rate": 4.6872461413484975e-06, + "loss": 1.869, + "step": 442 + }, + { + "epoch": 0.48547945205479454, + "grad_norm": 2.406616687774658, + "learning_rate": 4.685215272136475e-06, + "loss": 1.8441, + "step": 443 + }, + { + "epoch": 0.4865753424657534, + "grad_norm": 2.573341131210327, + "learning_rate": 4.683184402924452e-06, + "loss": 1.8378, + "step": 444 + }, + { + "epoch": 0.4876712328767123, + "grad_norm": 2.38126540184021, + "learning_rate": 4.681153533712429e-06, + "loss": 1.8289, + "step": 445 + }, + { + "epoch": 0.48876712328767125, + "grad_norm": 2.48789381980896, + "learning_rate": 4.679122664500407e-06, + "loss": 1.9087, + "step": 446 + }, + { + "epoch": 0.4898630136986301, + "grad_norm": 2.701105833053589, + "learning_rate": 4.677091795288384e-06, + "loss": 1.8072, + "step": 447 + }, + { + "epoch": 0.49095890410958903, + "grad_norm": 2.4330027103424072, + "learning_rate": 4.675060926076361e-06, + "loss": 1.7734, + "step": 448 + }, + { + "epoch": 0.49205479452054796, + "grad_norm": 2.5766141414642334, + "learning_rate": 4.673030056864338e-06, + "loss": 1.8826, + "step": 449 + }, + { + "epoch": 0.4931506849315068, + "grad_norm": 2.516427516937256, + "learning_rate": 4.670999187652316e-06, + "loss": 1.7688, + "step": 450 + }, + { + "epoch": 0.4931506849315068, + "eval_loss": 1.9994404315948486, + "eval_runtime": 37.1559, + "eval_samples_per_second": 26.914, + "eval_steps_per_second": 1.13, + "step": 450 + }, + { + "epoch": 0.49424657534246574, + "grad_norm": 2.5246028900146484, + "learning_rate": 4.668968318440293e-06, + "loss": 1.7906, + "step": 451 + }, + { + "epoch": 0.49534246575342467, + "grad_norm": 2.5245795249938965, + "learning_rate": 4.66693744922827e-06, + "loss": 1.7793, + "step": 452 + }, + { + "epoch": 0.4964383561643836, + "grad_norm": 2.537071466445923, + "learning_rate": 4.6649065800162475e-06, + "loss": 1.8586, + "step": 453 + }, + { + "epoch": 0.49753424657534245, + "grad_norm": 2.709639072418213, + "learning_rate": 4.662875710804225e-06, + "loss": 1.7657, + "step": 454 + }, + { + "epoch": 0.4986301369863014, + "grad_norm": 2.5277726650238037, + "learning_rate": 4.660844841592202e-06, + "loss": 1.8319, + "step": 455 + }, + { + "epoch": 0.4997260273972603, + "grad_norm": 2.5228514671325684, + "learning_rate": 4.658813972380179e-06, + "loss": 1.7612, + "step": 456 + }, + { + "epoch": 0.5008219178082192, + "grad_norm": 2.784677267074585, + "learning_rate": 4.656783103168157e-06, + "loss": 1.8727, + "step": 457 + }, + { + "epoch": 0.5019178082191781, + "grad_norm": 2.570239305496216, + "learning_rate": 4.654752233956134e-06, + "loss": 1.7435, + "step": 458 + }, + { + "epoch": 0.503013698630137, + "grad_norm": 2.5987436771392822, + "learning_rate": 4.652721364744111e-06, + "loss": 1.8351, + "step": 459 + }, + { + "epoch": 0.5041095890410959, + "grad_norm": 2.610114097595215, + "learning_rate": 4.6506904955320885e-06, + "loss": 1.81, + "step": 460 + }, + { + "epoch": 0.5041095890410959, + "eval_loss": 1.9993067979812622, + "eval_runtime": 37.1803, + "eval_samples_per_second": 26.896, + "eval_steps_per_second": 1.13, + "step": 460 + }, + { + "epoch": 0.5052054794520547, + "grad_norm": 2.751809597015381, + "learning_rate": 4.648659626320066e-06, + "loss": 1.8693, + "step": 461 + }, + { + "epoch": 0.5063013698630137, + "grad_norm": 2.467066764831543, + "learning_rate": 4.646628757108043e-06, + "loss": 1.8593, + "step": 462 + }, + { + "epoch": 0.5073972602739726, + "grad_norm": 2.4966063499450684, + "learning_rate": 4.6445978878960194e-06, + "loss": 1.8136, + "step": 463 + }, + { + "epoch": 0.5084931506849315, + "grad_norm": 2.506410598754883, + "learning_rate": 4.6425670186839976e-06, + "loss": 1.7826, + "step": 464 + }, + { + "epoch": 0.5095890410958904, + "grad_norm": 2.651237964630127, + "learning_rate": 4.640536149471974e-06, + "loss": 1.8173, + "step": 465 + }, + { + "epoch": 0.5106849315068493, + "grad_norm": 2.7836928367614746, + "learning_rate": 4.638505280259952e-06, + "loss": 1.849, + "step": 466 + }, + { + "epoch": 0.5117808219178083, + "grad_norm": 2.6783299446105957, + "learning_rate": 4.6364744110479285e-06, + "loss": 1.8577, + "step": 467 + }, + { + "epoch": 0.5128767123287671, + "grad_norm": 2.4898788928985596, + "learning_rate": 4.634443541835907e-06, + "loss": 1.8237, + "step": 468 + }, + { + "epoch": 0.513972602739726, + "grad_norm": 2.74481201171875, + "learning_rate": 4.632412672623883e-06, + "loss": 1.8577, + "step": 469 + }, + { + "epoch": 0.5150684931506849, + "grad_norm": 2.56190824508667, + "learning_rate": 4.630381803411861e-06, + "loss": 1.8211, + "step": 470 + }, + { + "epoch": 0.5150684931506849, + "eval_loss": 1.9867504835128784, + "eval_runtime": 37.1359, + "eval_samples_per_second": 26.928, + "eval_steps_per_second": 1.131, + "step": 470 + }, + { + "epoch": 0.5161643835616438, + "grad_norm": 2.485008478164673, + "learning_rate": 4.628350934199838e-06, + "loss": 1.6976, + "step": 471 + }, + { + "epoch": 0.5172602739726028, + "grad_norm": 2.645906448364258, + "learning_rate": 4.626320064987816e-06, + "loss": 1.8071, + "step": 472 + }, + { + "epoch": 0.5183561643835617, + "grad_norm": 2.457845687866211, + "learning_rate": 4.624289195775792e-06, + "loss": 1.8582, + "step": 473 + }, + { + "epoch": 0.5194520547945205, + "grad_norm": 2.6392905712127686, + "learning_rate": 4.62225832656377e-06, + "loss": 1.8103, + "step": 474 + }, + { + "epoch": 0.5205479452054794, + "grad_norm": 2.8293070793151855, + "learning_rate": 4.620227457351747e-06, + "loss": 1.8021, + "step": 475 + }, + { + "epoch": 0.5216438356164383, + "grad_norm": 2.6029255390167236, + "learning_rate": 4.618196588139724e-06, + "loss": 1.8064, + "step": 476 + }, + { + "epoch": 0.5227397260273973, + "grad_norm": 2.618307590484619, + "learning_rate": 4.616165718927701e-06, + "loss": 1.8176, + "step": 477 + }, + { + "epoch": 0.5238356164383562, + "grad_norm": 2.7881240844726562, + "learning_rate": 4.6141348497156786e-06, + "loss": 1.7908, + "step": 478 + }, + { + "epoch": 0.5249315068493151, + "grad_norm": 2.4347259998321533, + "learning_rate": 4.612103980503656e-06, + "loss": 1.8788, + "step": 479 + }, + { + "epoch": 0.5260273972602739, + "grad_norm": 2.581630229949951, + "learning_rate": 4.610073111291633e-06, + "loss": 1.8469, + "step": 480 + }, + { + "epoch": 0.5260273972602739, + "eval_loss": 1.988048791885376, + "eval_runtime": 37.1521, + "eval_samples_per_second": 26.916, + "eval_steps_per_second": 1.13, + "step": 480 + }, + { + "epoch": 0.5271232876712328, + "grad_norm": 2.40920090675354, + "learning_rate": 4.60804224207961e-06, + "loss": 1.8168, + "step": 481 + }, + { + "epoch": 0.5282191780821918, + "grad_norm": 2.4435620307922363, + "learning_rate": 4.606011372867588e-06, + "loss": 1.8088, + "step": 482 + }, + { + "epoch": 0.5293150684931507, + "grad_norm": 2.460261106491089, + "learning_rate": 4.603980503655565e-06, + "loss": 1.8343, + "step": 483 + }, + { + "epoch": 0.5304109589041096, + "grad_norm": 2.472010850906372, + "learning_rate": 4.601949634443542e-06, + "loss": 1.806, + "step": 484 + }, + { + "epoch": 0.5315068493150685, + "grad_norm": 2.578524589538574, + "learning_rate": 4.5999187652315195e-06, + "loss": 1.7372, + "step": 485 + }, + { + "epoch": 0.5326027397260275, + "grad_norm": 2.654944658279419, + "learning_rate": 4.597887896019497e-06, + "loss": 1.7519, + "step": 486 + }, + { + "epoch": 0.5336986301369863, + "grad_norm": 2.6866676807403564, + "learning_rate": 4.595857026807474e-06, + "loss": 1.8048, + "step": 487 + }, + { + "epoch": 0.5347945205479452, + "grad_norm": 2.708998918533325, + "learning_rate": 4.593826157595451e-06, + "loss": 1.8096, + "step": 488 + }, + { + "epoch": 0.5358904109589041, + "grad_norm": 2.841792345046997, + "learning_rate": 4.591795288383429e-06, + "loss": 1.778, + "step": 489 + }, + { + "epoch": 0.536986301369863, + "grad_norm": 2.6368772983551025, + "learning_rate": 4.589764419171406e-06, + "loss": 1.8524, + "step": 490 + }, + { + "epoch": 0.536986301369863, + "eval_loss": 1.9821399450302124, + "eval_runtime": 37.1274, + "eval_samples_per_second": 26.934, + "eval_steps_per_second": 1.131, + "step": 490 + }, + { + "epoch": 0.538082191780822, + "grad_norm": 2.739165782928467, + "learning_rate": 4.587733549959383e-06, + "loss": 1.8284, + "step": 491 + }, + { + "epoch": 0.5391780821917809, + "grad_norm": 2.428269147872925, + "learning_rate": 4.5857026807473604e-06, + "loss": 1.7719, + "step": 492 + }, + { + "epoch": 0.5402739726027397, + "grad_norm": 2.379875659942627, + "learning_rate": 4.583671811535338e-06, + "loss": 1.7707, + "step": 493 + }, + { + "epoch": 0.5413698630136986, + "grad_norm": 2.624039888381958, + "learning_rate": 4.581640942323315e-06, + "loss": 1.7442, + "step": 494 + }, + { + "epoch": 0.5424657534246575, + "grad_norm": 2.6703460216522217, + "learning_rate": 4.579610073111292e-06, + "loss": 1.8391, + "step": 495 + }, + { + "epoch": 0.5435616438356164, + "grad_norm": 2.724379062652588, + "learning_rate": 4.5775792038992695e-06, + "loss": 1.8463, + "step": 496 + }, + { + "epoch": 0.5446575342465754, + "grad_norm": 2.375727415084839, + "learning_rate": 4.575548334687247e-06, + "loss": 1.8401, + "step": 497 + }, + { + "epoch": 0.5457534246575343, + "grad_norm": 2.542696475982666, + "learning_rate": 4.573517465475224e-06, + "loss": 1.8636, + "step": 498 + }, + { + "epoch": 0.5468493150684931, + "grad_norm": 2.5622928142547607, + "learning_rate": 4.571486596263201e-06, + "loss": 1.7639, + "step": 499 + }, + { + "epoch": 0.547945205479452, + "grad_norm": 2.5569279193878174, + "learning_rate": 4.569455727051178e-06, + "loss": 1.8198, + "step": 500 + }, + { + "epoch": 0.547945205479452, + "eval_loss": 1.9778796434402466, + "eval_runtime": 37.1657, + "eval_samples_per_second": 26.906, + "eval_steps_per_second": 1.13, + "step": 500 + }, + { + "epoch": 0.549041095890411, + "grad_norm": 2.6474053859710693, + "learning_rate": 4.567424857839156e-06, + "loss": 1.8139, + "step": 501 + }, + { + "epoch": 0.5501369863013699, + "grad_norm": 2.636495590209961, + "learning_rate": 4.565393988627132e-06, + "loss": 1.8293, + "step": 502 + }, + { + "epoch": 0.5512328767123288, + "grad_norm": 2.6525604724884033, + "learning_rate": 4.5633631194151105e-06, + "loss": 1.8475, + "step": 503 + }, + { + "epoch": 0.5523287671232877, + "grad_norm": 2.700871229171753, + "learning_rate": 4.561332250203087e-06, + "loss": 1.7908, + "step": 504 + }, + { + "epoch": 0.5534246575342465, + "grad_norm": 2.7101478576660156, + "learning_rate": 4.559301380991065e-06, + "loss": 1.8141, + "step": 505 + }, + { + "epoch": 0.5545205479452054, + "grad_norm": 2.6825270652770996, + "learning_rate": 4.5572705117790414e-06, + "loss": 1.7579, + "step": 506 + }, + { + "epoch": 0.5556164383561644, + "grad_norm": 2.5643324851989746, + "learning_rate": 4.5552396425670196e-06, + "loss": 1.7814, + "step": 507 + }, + { + "epoch": 0.5567123287671233, + "grad_norm": 2.9949214458465576, + "learning_rate": 4.553208773354996e-06, + "loss": 1.8451, + "step": 508 + }, + { + "epoch": 0.5578082191780822, + "grad_norm": 2.6214592456817627, + "learning_rate": 4.551177904142974e-06, + "loss": 1.7665, + "step": 509 + }, + { + "epoch": 0.5589041095890411, + "grad_norm": 2.4616057872772217, + "learning_rate": 4.5491470349309505e-06, + "loss": 1.8566, + "step": 510 + }, + { + "epoch": 0.5589041095890411, + "eval_loss": 1.975143551826477, + "eval_runtime": 36.8303, + "eval_samples_per_second": 27.152, + "eval_steps_per_second": 1.14, + "step": 510 + }, + { + "epoch": 0.56, + "grad_norm": 2.7507705688476562, + "learning_rate": 4.547116165718929e-06, + "loss": 1.7764, + "step": 511 + }, + { + "epoch": 0.5610958904109589, + "grad_norm": 2.7398650646209717, + "learning_rate": 4.545085296506905e-06, + "loss": 1.8143, + "step": 512 + }, + { + "epoch": 0.5621917808219178, + "grad_norm": 2.476698398590088, + "learning_rate": 4.543054427294882e-06, + "loss": 1.7684, + "step": 513 + }, + { + "epoch": 0.5632876712328767, + "grad_norm": 2.440007209777832, + "learning_rate": 4.54102355808286e-06, + "loss": 1.7734, + "step": 514 + }, + { + "epoch": 0.5643835616438356, + "grad_norm": 2.857173204421997, + "learning_rate": 4.538992688870837e-06, + "loss": 1.7694, + "step": 515 + }, + { + "epoch": 0.5654794520547946, + "grad_norm": 2.4049761295318604, + "learning_rate": 4.536961819658814e-06, + "loss": 1.7974, + "step": 516 + }, + { + "epoch": 0.5665753424657535, + "grad_norm": 2.478177309036255, + "learning_rate": 4.5349309504467915e-06, + "loss": 1.793, + "step": 517 + }, + { + "epoch": 0.5676712328767123, + "grad_norm": 2.849869728088379, + "learning_rate": 4.532900081234769e-06, + "loss": 1.8157, + "step": 518 + }, + { + "epoch": 0.5687671232876712, + "grad_norm": 2.495866537094116, + "learning_rate": 4.530869212022746e-06, + "loss": 1.7698, + "step": 519 + }, + { + "epoch": 0.5698630136986301, + "grad_norm": 2.674063205718994, + "learning_rate": 4.528838342810723e-06, + "loss": 1.8023, + "step": 520 + }, + { + "epoch": 0.5698630136986301, + "eval_loss": 1.971044659614563, + "eval_runtime": 37.0575, + "eval_samples_per_second": 26.985, + "eval_steps_per_second": 1.133, + "step": 520 + }, + { + "epoch": 0.570958904109589, + "grad_norm": 2.8688297271728516, + "learning_rate": 4.526807473598701e-06, + "loss": 1.7905, + "step": 521 + }, + { + "epoch": 0.572054794520548, + "grad_norm": 2.4092676639556885, + "learning_rate": 4.524776604386678e-06, + "loss": 1.7955, + "step": 522 + }, + { + "epoch": 0.5731506849315069, + "grad_norm": 2.4861643314361572, + "learning_rate": 4.522745735174655e-06, + "loss": 1.7912, + "step": 523 + }, + { + "epoch": 0.5742465753424657, + "grad_norm": 2.632859945297241, + "learning_rate": 4.520714865962632e-06, + "loss": 1.8499, + "step": 524 + }, + { + "epoch": 0.5753424657534246, + "grad_norm": 2.6263527870178223, + "learning_rate": 4.51868399675061e-06, + "loss": 1.7913, + "step": 525 + }, + { + "epoch": 0.5764383561643835, + "grad_norm": 2.4004664421081543, + "learning_rate": 4.516653127538587e-06, + "loss": 1.8213, + "step": 526 + }, + { + "epoch": 0.5775342465753425, + "grad_norm": 2.467395067214966, + "learning_rate": 4.514622258326564e-06, + "loss": 1.7722, + "step": 527 + }, + { + "epoch": 0.5786301369863014, + "grad_norm": 2.62382435798645, + "learning_rate": 4.5125913891145415e-06, + "loss": 1.7589, + "step": 528 + }, + { + "epoch": 0.5797260273972603, + "grad_norm": 2.5330324172973633, + "learning_rate": 4.510560519902519e-06, + "loss": 1.7847, + "step": 529 + }, + { + "epoch": 0.5808219178082191, + "grad_norm": 2.4729249477386475, + "learning_rate": 4.508529650690496e-06, + "loss": 1.8447, + "step": 530 + }, + { + "epoch": 0.5808219178082191, + "eval_loss": 1.9698253870010376, + "eval_runtime": 37.2137, + "eval_samples_per_second": 26.872, + "eval_steps_per_second": 1.129, + "step": 530 + }, + { + "epoch": 0.581917808219178, + "grad_norm": 2.7225241661071777, + "learning_rate": 4.506498781478473e-06, + "loss": 1.7383, + "step": 531 + }, + { + "epoch": 0.583013698630137, + "grad_norm": 2.677614212036133, + "learning_rate": 4.504467912266451e-06, + "loss": 1.7323, + "step": 532 + }, + { + "epoch": 0.5841095890410959, + "grad_norm": 2.3463661670684814, + "learning_rate": 4.502437043054428e-06, + "loss": 1.8017, + "step": 533 + }, + { + "epoch": 0.5852054794520548, + "grad_norm": 2.403362512588501, + "learning_rate": 4.500406173842405e-06, + "loss": 1.8743, + "step": 534 + }, + { + "epoch": 0.5863013698630137, + "grad_norm": 2.609267234802246, + "learning_rate": 4.4983753046303824e-06, + "loss": 1.802, + "step": 535 + }, + { + "epoch": 0.5873972602739727, + "grad_norm": 2.394679307937622, + "learning_rate": 4.49634443541836e-06, + "loss": 1.7896, + "step": 536 + }, + { + "epoch": 0.5884931506849315, + "grad_norm": 2.532957077026367, + "learning_rate": 4.494313566206336e-06, + "loss": 1.7716, + "step": 537 + }, + { + "epoch": 0.5895890410958904, + "grad_norm": 2.659764289855957, + "learning_rate": 4.492282696994314e-06, + "loss": 1.8528, + "step": 538 + }, + { + "epoch": 0.5906849315068493, + "grad_norm": 2.442457914352417, + "learning_rate": 4.490251827782291e-06, + "loss": 1.7947, + "step": 539 + }, + { + "epoch": 0.5917808219178082, + "grad_norm": 2.443274736404419, + "learning_rate": 4.488220958570269e-06, + "loss": 1.7351, + "step": 540 + }, + { + "epoch": 0.5917808219178082, + "eval_loss": 1.9651908874511719, + "eval_runtime": 37.1745, + "eval_samples_per_second": 26.9, + "eval_steps_per_second": 1.13, + "step": 540 + }, + { + "epoch": 0.5928767123287672, + "grad_norm": 2.402625322341919, + "learning_rate": 4.486190089358245e-06, + "loss": 1.7653, + "step": 541 + }, + { + "epoch": 0.5939726027397261, + "grad_norm": 2.4391021728515625, + "learning_rate": 4.484159220146223e-06, + "loss": 1.7177, + "step": 542 + }, + { + "epoch": 0.5950684931506849, + "grad_norm": 2.587466239929199, + "learning_rate": 4.4821283509342e-06, + "loss": 1.8174, + "step": 543 + }, + { + "epoch": 0.5961643835616438, + "grad_norm": 2.3635201454162598, + "learning_rate": 4.480097481722178e-06, + "loss": 1.7943, + "step": 544 + }, + { + "epoch": 0.5972602739726027, + "grad_norm": 2.5390195846557617, + "learning_rate": 4.478066612510154e-06, + "loss": 1.7927, + "step": 545 + }, + { + "epoch": 0.5983561643835617, + "grad_norm": 2.5881011486053467, + "learning_rate": 4.4760357432981325e-06, + "loss": 1.7499, + "step": 546 + }, + { + "epoch": 0.5994520547945206, + "grad_norm": 2.4159603118896484, + "learning_rate": 4.474004874086109e-06, + "loss": 1.7773, + "step": 547 + }, + { + "epoch": 0.6005479452054795, + "grad_norm": 2.908005475997925, + "learning_rate": 4.471974004874087e-06, + "loss": 1.9111, + "step": 548 + }, + { + "epoch": 0.6016438356164383, + "grad_norm": 2.4689557552337646, + "learning_rate": 4.4699431356620635e-06, + "loss": 1.727, + "step": 549 + }, + { + "epoch": 0.6027397260273972, + "grad_norm": 2.5731844902038574, + "learning_rate": 4.467912266450041e-06, + "loss": 1.8425, + "step": 550 + }, + { + "epoch": 0.6027397260273972, + "eval_loss": 1.9631503820419312, + "eval_runtime": 37.2189, + "eval_samples_per_second": 26.868, + "eval_steps_per_second": 1.128, + "step": 550 + }, + { + "epoch": 0.6038356164383561, + "grad_norm": 2.4711105823516846, + "learning_rate": 4.465881397238018e-06, + "loss": 1.7985, + "step": 551 + }, + { + "epoch": 0.6049315068493151, + "grad_norm": 2.5799551010131836, + "learning_rate": 4.463850528025995e-06, + "loss": 1.768, + "step": 552 + }, + { + "epoch": 0.606027397260274, + "grad_norm": 2.5204827785491943, + "learning_rate": 4.4618196588139726e-06, + "loss": 1.8141, + "step": 553 + }, + { + "epoch": 0.6071232876712329, + "grad_norm": 2.595759391784668, + "learning_rate": 4.45978878960195e-06, + "loss": 1.8273, + "step": 554 + }, + { + "epoch": 0.6082191780821918, + "grad_norm": 2.5692834854125977, + "learning_rate": 4.457757920389927e-06, + "loss": 1.8023, + "step": 555 + }, + { + "epoch": 0.6093150684931506, + "grad_norm": 2.402923345565796, + "learning_rate": 4.455727051177904e-06, + "loss": 1.7329, + "step": 556 + }, + { + "epoch": 0.6104109589041096, + "grad_norm": 2.480069160461426, + "learning_rate": 4.453696181965882e-06, + "loss": 1.7722, + "step": 557 + }, + { + "epoch": 0.6115068493150685, + "grad_norm": 2.685495138168335, + "learning_rate": 4.451665312753859e-06, + "loss": 1.8269, + "step": 558 + }, + { + "epoch": 0.6126027397260274, + "grad_norm": 2.513362407684326, + "learning_rate": 4.449634443541836e-06, + "loss": 1.7996, + "step": 559 + }, + { + "epoch": 0.6136986301369863, + "grad_norm": 2.535891532897949, + "learning_rate": 4.4476035743298135e-06, + "loss": 1.8341, + "step": 560 + }, + { + "epoch": 0.6136986301369863, + "eval_loss": 1.9630459547042847, + "eval_runtime": 37.1854, + "eval_samples_per_second": 26.892, + "eval_steps_per_second": 1.129, + "step": 560 + }, + { + "epoch": 0.6147945205479453, + "grad_norm": 2.440772771835327, + "learning_rate": 4.445572705117791e-06, + "loss": 1.7514, + "step": 561 + }, + { + "epoch": 0.6158904109589041, + "grad_norm": 2.398111343383789, + "learning_rate": 4.443541835905768e-06, + "loss": 1.7599, + "step": 562 + }, + { + "epoch": 0.616986301369863, + "grad_norm": 2.5974881649017334, + "learning_rate": 4.441510966693745e-06, + "loss": 1.8262, + "step": 563 + }, + { + "epoch": 0.6180821917808219, + "grad_norm": 2.5586795806884766, + "learning_rate": 4.439480097481723e-06, + "loss": 1.7638, + "step": 564 + }, + { + "epoch": 0.6191780821917808, + "grad_norm": 2.5561492443084717, + "learning_rate": 4.4374492282697e-06, + "loss": 1.8434, + "step": 565 + }, + { + "epoch": 0.6202739726027398, + "grad_norm": 2.7550947666168213, + "learning_rate": 4.435418359057677e-06, + "loss": 1.702, + "step": 566 + }, + { + "epoch": 0.6213698630136987, + "grad_norm": 2.6542508602142334, + "learning_rate": 4.433387489845654e-06, + "loss": 1.7973, + "step": 567 + }, + { + "epoch": 0.6224657534246575, + "grad_norm": 2.592461585998535, + "learning_rate": 4.431356620633632e-06, + "loss": 1.788, + "step": 568 + }, + { + "epoch": 0.6235616438356164, + "grad_norm": 2.7114830017089844, + "learning_rate": 4.429325751421609e-06, + "loss": 1.7536, + "step": 569 + }, + { + "epoch": 0.6246575342465753, + "grad_norm": 2.414747476577759, + "learning_rate": 4.427294882209586e-06, + "loss": 1.7875, + "step": 570 + }, + { + "epoch": 0.6246575342465753, + "eval_loss": 1.9587609767913818, + "eval_runtime": 37.2007, + "eval_samples_per_second": 26.881, + "eval_steps_per_second": 1.129, + "step": 570 + }, + { + "epoch": 0.6257534246575343, + "grad_norm": 2.369011878967285, + "learning_rate": 4.4252640129975635e-06, + "loss": 1.771, + "step": 571 + }, + { + "epoch": 0.6268493150684932, + "grad_norm": 2.6837189197540283, + "learning_rate": 4.423233143785541e-06, + "loss": 1.7064, + "step": 572 + }, + { + "epoch": 0.6279452054794521, + "grad_norm": 2.4707508087158203, + "learning_rate": 4.421202274573518e-06, + "loss": 1.7772, + "step": 573 + }, + { + "epoch": 0.6290410958904109, + "grad_norm": 2.63283109664917, + "learning_rate": 4.4191714053614945e-06, + "loss": 1.8008, + "step": 574 + }, + { + "epoch": 0.6301369863013698, + "grad_norm": 2.4014840126037598, + "learning_rate": 4.417140536149473e-06, + "loss": 1.6774, + "step": 575 + }, + { + "epoch": 0.6312328767123287, + "grad_norm": 2.4078781604766846, + "learning_rate": 4.415109666937449e-06, + "loss": 1.7338, + "step": 576 + }, + { + "epoch": 0.6323287671232877, + "grad_norm": 2.5279288291931152, + "learning_rate": 4.413078797725427e-06, + "loss": 1.8084, + "step": 577 + }, + { + "epoch": 0.6334246575342466, + "grad_norm": 2.410296678543091, + "learning_rate": 4.411047928513404e-06, + "loss": 1.7504, + "step": 578 + }, + { + "epoch": 0.6345205479452055, + "grad_norm": 2.397796630859375, + "learning_rate": 4.409017059301382e-06, + "loss": 1.7761, + "step": 579 + }, + { + "epoch": 0.6356164383561644, + "grad_norm": 2.6257503032684326, + "learning_rate": 4.406986190089358e-06, + "loss": 1.7741, + "step": 580 + }, + { + "epoch": 0.6356164383561644, + "eval_loss": 1.9512827396392822, + "eval_runtime": 37.1932, + "eval_samples_per_second": 26.887, + "eval_steps_per_second": 1.129, + "step": 580 + }, + { + "epoch": 0.6367123287671232, + "grad_norm": 2.516399383544922, + "learning_rate": 4.404955320877336e-06, + "loss": 1.8043, + "step": 581 + }, + { + "epoch": 0.6378082191780822, + "grad_norm": 2.4476537704467773, + "learning_rate": 4.402924451665313e-06, + "loss": 1.8018, + "step": 582 + }, + { + "epoch": 0.6389041095890411, + "grad_norm": 2.3977835178375244, + "learning_rate": 4.400893582453291e-06, + "loss": 1.7419, + "step": 583 + }, + { + "epoch": 0.64, + "grad_norm": 2.4341602325439453, + "learning_rate": 4.398862713241267e-06, + "loss": 1.7319, + "step": 584 + }, + { + "epoch": 0.6410958904109589, + "grad_norm": 2.611096143722534, + "learning_rate": 4.396831844029245e-06, + "loss": 1.7661, + "step": 585 + }, + { + "epoch": 0.6421917808219179, + "grad_norm": 2.5748324394226074, + "learning_rate": 4.394800974817222e-06, + "loss": 1.7834, + "step": 586 + }, + { + "epoch": 0.6432876712328767, + "grad_norm": 2.5776925086975098, + "learning_rate": 4.392770105605199e-06, + "loss": 1.7718, + "step": 587 + }, + { + "epoch": 0.6443835616438356, + "grad_norm": 2.6215028762817383, + "learning_rate": 4.390739236393176e-06, + "loss": 1.7803, + "step": 588 + }, + { + "epoch": 0.6454794520547945, + "grad_norm": 2.517838954925537, + "learning_rate": 4.388708367181154e-06, + "loss": 1.7986, + "step": 589 + }, + { + "epoch": 0.6465753424657534, + "grad_norm": 2.981196165084839, + "learning_rate": 4.386677497969131e-06, + "loss": 1.7996, + "step": 590 + }, + { + "epoch": 0.6465753424657534, + "eval_loss": 1.9492793083190918, + "eval_runtime": 37.1837, + "eval_samples_per_second": 26.893, + "eval_steps_per_second": 1.13, + "step": 590 + }, + { + "epoch": 0.6476712328767124, + "grad_norm": 2.7888052463531494, + "learning_rate": 4.384646628757108e-06, + "loss": 1.7935, + "step": 591 + }, + { + "epoch": 0.6487671232876713, + "grad_norm": 2.807997226715088, + "learning_rate": 4.3826157595450855e-06, + "loss": 1.8281, + "step": 592 + }, + { + "epoch": 0.6498630136986301, + "grad_norm": 2.5797502994537354, + "learning_rate": 4.380584890333063e-06, + "loss": 1.7655, + "step": 593 + }, + { + "epoch": 0.650958904109589, + "grad_norm": 2.68839168548584, + "learning_rate": 4.37855402112104e-06, + "loss": 1.7329, + "step": 594 + }, + { + "epoch": 0.6520547945205479, + "grad_norm": 2.6886773109436035, + "learning_rate": 4.376523151909017e-06, + "loss": 1.6756, + "step": 595 + }, + { + "epoch": 0.6531506849315069, + "grad_norm": 3.0069046020507812, + "learning_rate": 4.3744922826969946e-06, + "loss": 1.8444, + "step": 596 + }, + { + "epoch": 0.6542465753424658, + "grad_norm": 2.382343292236328, + "learning_rate": 4.372461413484972e-06, + "loss": 1.7632, + "step": 597 + }, + { + "epoch": 0.6553424657534247, + "grad_norm": 2.983272075653076, + "learning_rate": 4.370430544272949e-06, + "loss": 1.779, + "step": 598 + }, + { + "epoch": 0.6564383561643835, + "grad_norm": 2.544196844100952, + "learning_rate": 4.368399675060926e-06, + "loss": 1.7994, + "step": 599 + }, + { + "epoch": 0.6575342465753424, + "grad_norm": 2.836195468902588, + "learning_rate": 4.366368805848904e-06, + "loss": 1.8101, + "step": 600 + }, + { + "epoch": 0.6575342465753424, + "eval_loss": 1.9479612112045288, + "eval_runtime": 37.2259, + "eval_samples_per_second": 26.863, + "eval_steps_per_second": 1.128, + "step": 600 + } + ], + "logging_steps": 1, + "max_steps": 2736, + "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": 1.870385645617152e+17, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/zero_to_fp32.py b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..24cc342e78d1a006c782b3a4cd68d9ce786d8fd8 --- /dev/null +++ b/checkpoints/Llama-3.2-3B/babylm_shuffle_even_odd_10M_seed0/runs/checkpoint-600/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)