diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10056.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10056.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..2b6a820c2dc247535f20760f8f96e13acec30ac2 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10056.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b2b4a3ccd5b7d783a5fc00d3f98e3f715a015cff29ba7844de6c9db77337e410 +size 1222302563 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10259.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10259.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..b8ae5ffd3ea62c6bbc45648c6de99473db369f92 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10259.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78b4693b222ab14e46911127077b37087f1188b007188d56c124039577cfd11b +size 1222303073 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10472.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10472.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..0d854f83be8cc67b86258aa3c3d70dd79a3c1c54 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10472.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7ce037425a43e34a531c6122c9d9df80433907b120b397d2c4aa51a5f80fa444 +size 1222303519 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10694.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10694.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..3db2f269d3f33f659ca14239d0855872145b45d5 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10694.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f34781ad47e17e2ab4db2f29eceff6e1f0e50de1dfe264dd1a1090be96112acd +size 1222304029 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10918.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10918.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..f9549a7616573dcef1ee582ac47c61889a665671 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=10918.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6bb9978fc7d466b48c6c0a12ff3e9f391c9e200fefa33cadd773ae89d33ada36 +size 1222304475 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1108.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1108.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..941cee10a192eae8b9970aa7a8be8bfc458fcbdc --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1108.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:99f911eeb7434b96044ba0281b23ff36bb72ee990b73db90fadf158c45f88da7 +size 1222290957 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11131.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11131.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..c4ab96246184c1c24eea49774c10053ae2886f17 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11131.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3345ef67a0cb8467706d004dab635db419bb02b4a5b873406deece9b0f5d3dfe +size 1222304985 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11322.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11322.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..fec3ea2d698bf53b7add67ba4dbd0d3f0c363bd0 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11322.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03215584f3d408fea67e3bc105a8d4bdd5bbccc96bd672a25825dd953ceecfbb +size 1222305431 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11487.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11487.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..484c8ebd94a23ca505d52c9aeeffa4bd06fccf59 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11487.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:784362099449c8d19c5965405ab5c08ae0d0ab1302328f75491d9389ceb874cc +size 1222305941 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11673.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11673.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..3e811f1c404ba98aac71c30d8d461ffcd3071d02 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11673.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7ccaab083dff771ec8f79f7f8f3f4c6c73c89627816ccce3e530c2c0d3aa5cb7 +size 1222306451 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11859.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11859.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..cfa5d3337c9757c122a71cd70823f4fad3ce8f97 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=11859.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:739714416c702b153fafcf17f3a4907107579eb427fd00a223f761ce2c31e6c6 +size 1222306897 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12042.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12042.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..a25daff5718772e220c3ffa57b72730bf8a100c6 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12042.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:94d00c3a693a8e15b401ad85df74d0d8645ffa03a6ef2a54bc2c2bba762b9c34 +size 1222307407 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12217.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12217.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..69917626ac44ba6afc7845041c3ec1c966771f08 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12217.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b4efb1c00687aa12192911eb345e11f91c6ce1f5a576a37dcd311fda597ff0ba +size 1222307853 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12395.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12395.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..944b6c3dc38d4b201eb76ac3ec937badfb1ce18f --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12395.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76880badded00a6736ac309919a1d4946b2f2ea84ca6699aa7d6c5a94b58e17c +size 1222308363 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12562.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12562.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..c308c10fd1bff61fde0c8bf8eeaef96c63988a86 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12562.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:377b3216846f843fc232e2ca9452186568a92082ef33f4dad5643dfe7efa448f +size 1222308809 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12801.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12801.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..13f137a0c0eaddaf835744599becc568c6bc8238 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=12801.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:96477798251dd6056c58f717d140c949efee862707e6d28e64f825538a2faea6 +size 1222290773 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13051.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13051.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..fe590a0770aefa194d0a3fe5b1d0a7035b5c66ae --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13051.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a038dc5f10465a58e04cfb023794b91ba87efa136929164fb4ec7d14de6fccb6 +size 1222291283 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1318.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1318.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..8c1b0ce674ccf467251971c05959172d8b9542ec --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1318.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9dc784c425b2d9cf9ed89620d2b6dead5a4baaaec83fbadda40bacd48fa3e8dc +size 1222291467 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13335.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13335.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..761116e62b369724e96ca63a8581506a4d8291ee --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13335.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0dce7f6a2d0733ce113ca6640c431b3bc9aad10be3819ed2f7532e8724b9afb7 +size 1222291793 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13558.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13558.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..10cf0b6d53f0f2c23a3e6f52dca8dd9dfb956b0a --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13558.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:093b2b81653d0a6ad0bab9088025143503654579fe4e9130c66db2a4e83b5de3 +size 1222292239 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13761.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13761.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..ef2cb3373cf61234051f41c44f24f8121d8b55b3 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13761.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fcf0edcab1b106c66a62783654150d48db3ea3255e41fe98745623d9ed508a35 +size 1222292749 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13953.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13953.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..1c67103768b7136f06ea2f951e4230ba79f50dd9 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=13953.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c6cad125dab82bc6ed85c9f7ce61be1749590f27d491f2ebdba5670e187dfd0 +size 1222293195 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14151.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14151.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..7139d5d06b85100ce82f1fcea44158a9f9ffc61b --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14151.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:40e4595ee2dbf05b07692898499ce6e9fc48f09e7bd43b29002612badcdb0922 +size 1222293705 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14361.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14361.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..3dac444bb3fa5fa90f4b4c62aa376ee2f0b74a14 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14361.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d3cc10fc0c87d4482ea472159e58be6b05297fee58c5be4c3f38817915a5377 +size 1222294151 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14554.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14554.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..d9b701ef5019b785e67dbc4732a3cecb314d34c2 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14554.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b3863035eeb2d08c9097b5ab1c1ede963bdd84c23f8c038782f093a0be6c697f +size 1222294597 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14754.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14754.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..6826e0635f6e69f7828708f72ce5bc987a569279 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14754.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b639729a2eaedd0034c508f800f3d55515d52e361b01c36e7bb9b368da2532bd +size 1222295043 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14935.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14935.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..abd5ab9659b308a90b85d6762a7357cf0fd2e773 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=14935.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7a85aea9df9a6b5fff5d5fd1cb3f01f5b7996a50dbe88f58e04a4b1bc8bac3a +size 1222295489 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15110.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15110.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..e33ae4513270a61187a159edebb13b3c7d202689 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15110.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:58bc1789e974f9c0f6b4f97baffde8fdea5c2ee6aac2a68bd93ea7da07a6e0a0 +size 1222296063 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15290.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15290.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..832344540b31107bd39de62645a0b6f5d83e7ad0 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15290.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:534c73de7f18992f5a9e02f6681650cf3ed7956d957a8dff2c5fd1b51cccb67e +size 1222296445 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1535.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1535.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..cc242890b17963c8dc88f92a879013a148dc6fe5 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1535.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b4452e5aa0fb324e941744a85d712e40de087b487d04aab4f59b086bf8f67a8 +size 1222291913 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15460.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15460.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..bc6853a3ef19a235e7b1454889fd8926d65524fa --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15460.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:723e9991b0c993870d9bbe2a4bc1aaffdefe294db9a62a762dbf7a9717688e45 +size 1222296891 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15635.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15635.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..7692633759cded1d2fadd57ed49b7f39cddd1bf4 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15635.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52263a426236a4f9f2cc0673b1ab466129d1af7f8b1f97b5ea2abc73247a3fe0 +size 1222297401 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15816.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15816.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..be1fb93a4463407e5a7c40c90380d1d8fba756d5 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=15816.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea07bf632e7f7a2b726409ff74d4274cab20cb08a7aa7495cbd6aaa8f89e12d4 +size 1222297847 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=16001.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=16001.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..79260adfc0cfb54307567796c7c2e87f69bb2088 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=16001.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:58d2089cca0fb3b827610a5f51d7bab6aaa424a9183944ac5f214608debb9720 +size 1222298293 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=16189.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=16189.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..a5ed4b829b88586e639fbb3aba92507a117cfe6e --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=16189.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cec79f13b10768490136a28ac6ec4bbe3df95852bfe914b0bac43a56b52a7b7f +size 1222298739 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1759.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1759.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..10f754e389999fb8a5754d5cebe3e8ba13661d38 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1759.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a2109cb2f81ebec5621b5fd6981b59af32d2cccfab613114ffce370571737720 +size 1222292359 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1977.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1977.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..77f3bcd01e27453e6817a21425f0df53fa714524 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=1977.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a67627d45cca3a7c0a62953b54e072c28ccf4502110a2dc48568123c42df829 +size 1222292869 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2211.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2211.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..de5ae03d5ee083092b604eb3a53505351db20f59 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2211.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1ced0f60b5328425095ee5e73df73b68ffe13037ca4ababfe0afddc1ba8b5c9f +size 1222293251 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=227.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=227.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..0d5384dbfefacb32edec4f8fd5dc49f63e2f0f26 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=227.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a52b279f4f9baafade1c2e3c55726f699fa58a239657c2280ee527288da130b1 +size 1222288981 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2436.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2436.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..19379de09419a9a5c6bc10fb459e64c53e4bca32 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2436.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b48b50cc6a19309ed5441bb91fd8d5cb2d8037ef86d357bd21a1277a2b7d01f +size 1222293761 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2649.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2649.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..843d21fed7548c23c8063d30d19575cbcf4c780c --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2649.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:54be6d5c8120801c972666bd1e5ec41ac485022cbb5a209c36e7202fd58f5716 +size 1222294271 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2884.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2884.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..4c7071a4570a2c469fb295f9ae936f5106b8d279 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=2884.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf6ebce57dceb1a9141e564e4d6c2006846ec846b1580a049ca1833b63c94fcb +size 1222294717 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3106.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3106.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..348685520d0625299fdc6de2a0cc606870da8228 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3106.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b49a879b1c96accce5f0f4c54d00644f81dd6213c1f14aa5942d36e6dbcafb2a +size 1222295099 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3340.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3340.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..6d0a0032ce6642a64fb723b892c153c8614109e0 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3340.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:53ae8a394de4bed4c0d9a70d593563fa725f1c10bf85d2efb192116757cd38f7 +size 1222295673 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3577.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3577.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..be0ab2d7741740d68c106d8ec0acf5ecfbcfe5e7 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3577.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6972c41f5788de0699d0f4c50c7f021e8708473ce00178668991f3ab1166323d +size 1222296119 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3809.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3809.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..36f5b2eea469680439c53d2f77e796d469989765 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3809.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e4a3af39ade98e82ad557475b683b9ae3570f75f3274ccc9daab426b78db891f +size 1222296501 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3990.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3990.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..114ee9a41a4fc86185f7eb32136367e579bdf3ba --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3990.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f21ee06ca5c1ba591955b33769fe5921938730a4014e035a4d700a7a6a6c2a35 +size 1222297011 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4241.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4241.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..ff7150d650d231740931c5432f748a5c6b198730 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4241.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3379f2cd5f575127ecc11d6897badf45e595b4458ad5409630453f22747ebec8 +size 1222290773 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4479.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4479.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..429bf8844f98d8707f78b6f23fddb63fdba1d3ab --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4479.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e19093295c5bfcf40daf48211edcfb311fde634525842b8f452c65cd3499e580 +size 1222291283 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=465.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=465.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..4069792769de56e8b0d5e51ff0bd01847931569c --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=465.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d931120b8c619e248529e0ab8e97b24fc3b5876d1156093b30f0b6adc38b3d46 +size 1222289555 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4688.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4688.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..6ace1ba812e3e3c0ab95c3e910d38e964d7e9b48 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4688.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1b20f209d898ab5837543fd95333d52f6d2fe4c0ef7d0a90df1163dd8e2aa655 +size 1222291793 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4895.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4895.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..62faa46953ad8bf9ba32d264e6426e61160f95a7 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=4895.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:db3af2eedd657bf29ea625da42e19516ec82641888c4f523972319639497486d +size 1222292239 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5097.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5097.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..2f483853a8339cf3b4b72a293295d1d9f3683499 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5097.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ab0dc6ba19b88011101070dbc6de6a809ca020aa7b597064b5dd437e0133dbc +size 1222292749 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5298.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5298.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..930027fade01ed5ae96840a37894c73d9c48d99c --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5298.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e34b3903663a0fc6ff05a89fc2a46168d20a2d0e7cef5189de58540bff3d90fa +size 1222293195 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5524.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5524.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..a66b598ea0ccb1bb3a7f0047c1a4de19cbb8c79d --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5524.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f645e45a988a344fe30120e18804049c797e0bc251281a9f11ab67369b4906dc +size 1222293641 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5742.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5742.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..b3f0e016e563f7bd0b975775a6ac9ec147931309 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5742.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c79a7a684623a488cc38133d81e788ee14f297f1b25503f5fad764cc9635958 +size 1222294151 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5973.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5973.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..b320e87eba259807cdaf33889895aa9cae953227 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=5973.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0cbdc41526bcf9ddddaeb0b4abc4f267b1297f5343cae1c9214a98feb739a366 +size 1222294597 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6218.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6218.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..8734076cdc3bb4fc2f6c191b00669f3a6196cb9b --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6218.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca250a4a0383887842b0430d26a5d685b2038a16d4a1a72837251e301f187cfd +size 1222295043 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6449.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6449.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..beb8cd056b92792372b5584ad8d2528fd4830155 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6449.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b1276a72a2fed1b0944cb24d64a25b04a9046a42e14bc3a99d984f6b95d36f7 +size 1222295489 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6683.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6683.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..ee73f26752a0b5a7c07f83f249b5434c5c615630 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6683.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7f8dcbe8e6f6d32b5d85e642cd1d2391e77c70a46e6338c537c290b25f6f6c79 +size 1222296063 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=670.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=670.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..d279890c966f804338a4c6b8ec1af6b49a455530 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=670.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a4b47dc6b3cfe9d6f467a3d0aa9152c305074af9332183cd86a039f04ed0db53 +size 1222290001 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6921.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6921.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..ea424fa0c5b2ac1c312c9cd84a6c55ceb4d20592 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=6921.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb151306d4e08b27327d6f850016a73f7eb61825410f959a7a3f864091afa4ed +size 1222296445 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7153.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7153.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..9ea6e806748711a82c817a32ed64d0d7c8291d90 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7153.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9f334bd7a3efe507ef288a2d8a73bfdbc5b7d12a51a96c9aa176f6bfa152efbe +size 1222296827 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7408.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7408.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..07edd659737b2b03481362575977fba7075d3ace --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7408.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:feef699d5ca61a0a9bdc806913cb0c260072e13000266fd3728e76ee57540dee +size 1222297401 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7643.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7643.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..e2dcd80c960e3030feff5fe85a804f0db95853c1 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7643.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:123c656be398a963d4c9f6eac92cf9b32c07c6736b095befa9a695cbab4de91b +size 1222297847 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7906.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7906.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..b5fdbd89fcf84e449a7eed803854893ac9316eca --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=7906.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ef8a63bb3df84c4d340b002a1b16a38fd13090cdb21f29c77f467f793cb95536 +size 1222298293 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8160.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8160.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..d6963bf7f34183aeb667882bcbdad8be729fd998 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8160.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7c927c390cf89c9358f0416f2946676ab8e0bb75767b36b8be18cee6b4660eed +size 1222298739 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8380.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8380.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..b993e5531707b43972b3f1cabdd634112d860d7f --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8380.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47229388c3c8875dd13e872ed18eba13d3bb04bcc9cf951be622978428baafa0 +size 1222299249 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8626.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8626.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..5d897efc0a24abc30f011e08fc5499657af599df --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8626.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0060e7149075acb1a700e70f1da3df9bda9da5e3ef7071a2bf2620e7b71dea15 +size 1222299759 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8850.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8850.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..982b7c522ae533a7d193cf062fbc99e1d8d2ecdf --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=8850.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c8b8cdadbafa852d739d1d7ede530296dc8d6910117e434afeb3beba2ecb7f62 +size 1222300141 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=894.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=894.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..1286c91c837122a0acbc63e6acd23d2f3fe33543 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=894.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8cd9c8793f7e765e870541d0b19ce37cc65f383e82185060f0ccfc3e7cc9440 +size 1222290511 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9114.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9114.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..d47cc9c0ed813e681cef63520f003f704506a93e --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9114.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a8b9373c41f2868ebf0d4c97063bcbcbbea6ec2fe0b350a999b7ddab3a826d65 +size 1222300651 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9340.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9340.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..4cb79767f71263016cf40aaaa33c0abc7d2d621f --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9340.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52115706e380e97421e7de606f2b559935d94d5ce5fd7d070221bd4e0dc9f61c +size 1222301161 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9583.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9583.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..6a3a8425a275932cc98be9fe97110a03cb9eefa8 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9583.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5110a37c6c88a6d290a2f9a9089f5cc2cb1e1dc2cbb365907eb2764a743a6dbe +size 1222301671 diff --git a/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9807.ckpt b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9807.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..d936675b012c9713f4c0513c5a7904d837710a06 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=9807.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:50a4f545791b39a48d7f15ce55585642a5c5983d6c2467d778c24bf205630835 +size 1222302117 diff --git a/half_bs4080_seed5501650_bf16-mixed/last.ckpt b/half_bs4080_seed5501650_bf16-mixed/last.ckpt new file mode 100644 index 0000000000000000000000000000000000000000..278976014cbaae7ef0533a8d809f204292c0f7e6 --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/last.ckpt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7f2470ebdb0f097d01d12f59a141fe9f208d92d40689e18838a5f84bc9a952b8 +size 1222298739 diff --git a/half_bs4080_seed5501650_bf16-mixed/resume_singlenode_2023-08-18-11:12:45.log b/half_bs4080_seed5501650_bf16-mixed/resume_singlenode_2023-08-18-11:12:45.log new file mode 100644 index 0000000000000000000000000000000000000000..f1cfac30feb861d96109d3679f6be2ac799651de --- /dev/null +++ b/half_bs4080_seed5501650_bf16-mixed/resume_singlenode_2023-08-18-11:12:45.log @@ -0,0 +1,6370 @@ +wandb: Currently logged in as: tamariucai (rycolab). Use `wandb login --relogin` to force relogin +wandb: wandb version 0.15.8 is available! To upgrade, please run: +wandb: $ pip install wandb --upgrade +wandb: Tracking run with wandb version 0.15.4 +wandb: Run data is saved locally in ./wandb/run-20230818_111634-cbsw5iqa +wandb: Run `wandb offline` to turn off syncing. +wandb: Resuming run swift-sweep-4 +wandb: โญ๏ธ View project at https://wandb.ai/rycolab/alkmi-wit +wandb: ๐Ÿงน View sweep at https://wandb.ai/rycolab/alkmi-wit/sweeps/p9z42he9 +wandb: ๐Ÿš€ View run at https://wandb.ai/rycolab/alkmi-wit/runs/cbsw5iqa +[rank: 0] Global seed set to 5501650 +Found cached dataset parquet (/cluster/scratch/tamariucai/HuggingfaceDatasets/theodor1289___parquet/theodor1289--wit-bf532b06ea2d7bd2/0.0.0/14a00e99c0d15a23649d0db8944380ac81082d4b021f398733dd84f3a6c569a7) +Loading cached processed dataset at theodor1289wit-test +The current host is eu-a65-04 +The current cuDNN version is 8801 +Parameter 'text_perc' already had the value 100. +Parameter 'vision_perc' already had the value 0. +Parameter 'seed' already had the value 5501650. +Total GPU memory: 42.296 GB +Detected half-sized run, (almost) doubling batch size to 48. +Detected half-sized run, (almost) halving gradient accumulation to 85 +[update_ckeckpoint_dir] Detected hyperparameter run! +[update_ckeckpoint_dir] Detected half-size run! +[update_ckeckpoint_dir] Found and updated model name flava in the checkpoint dirpath. +[update_ckeckpoint_dir] Setting checkpoint dirpath to /cluster/work/cotterell/tamariucai/HuggingfaceCheckpoints/flava_half-wit/date(2023-08-18)_time(11:16:41)/seed(5501650)-magic({'enable': True})-text_perc(100)-vision_perc(0/ +Text is the predominant modality (100 v.s. 0 vision), will sample proportionally for optimal BLiMP performance. +Registering basic callbacks +[PPL Evaluation] Loading dataset 'theodor1289/wit' with split 'test' +[PPL Evaluation] Length of the dataset is 1295591 +Building model 'flava' +Enabling TensorFloat32 tensor cores for float32 matrix multiplication +Compiling FLAVA's text model... +Found cached dataset parquet (/cluster/scratch/tamariucai/HuggingfaceDatasets/theodor1289___parquet/theodor1289--wit-bf532b06ea2d7bd2/0.0.0/14a00e99c0d15a23649d0db8944380ac81082d4b021f398733dd84f3a6c569a7) +Found cached dataset parquet (/cluster/scratch/tamariucai/HuggingfaceDatasets/theodor1289___parquet/theodor1289--wit-bf532b06ea2d7bd2/0.0.0/14a00e99c0d15a23649d0db8944380ac81082d4b021f398733dd84f3a6c569a7) +Loading cached processed dataset at theodor1289wit-train[100%] +Loading cached processed dataset at theodor1289wit-test +/cluster/home/tamariucai/.cache/pypoetry/virtualenvs/alkmi-CJbY6op3-py3.10/lib64/python3.10/site-packages/lightning_fabric/plugins/environments/slurm.py:165: PossibleUserWarning: The `srun` command is available on your system but is not used. HINT: If your intention is to run Lightning on SLURM, prepend your python command with `srun` like so: srun python /cluster/work/cotterell/tamariucai/acquiring-linguis ... + rank_zero_warn( +Using bfloat16 Automatic Mixed Precision (AMP) +GPU available: True (cuda), used: True +TPU available: False, using: 0 TPU cores +IPU available: False, using: 0 IPUs +HPU available: False, using: 0 HPUs +Found cached dataset parquet (/cluster/scratch/tamariucai/HuggingfaceDatasets/theodor1289___parquet/theodor1289--wit-bf532b06ea2d7bd2/0.0.0/14a00e99c0d15a23649d0db8944380ac81082d4b021f398733dd84f3a6c569a7) +Found cached dataset parquet (/cluster/scratch/tamariucai/HuggingfaceDatasets/theodor1289___parquet/theodor1289--wit-bf532b06ea2d7bd2/0.0.0/14a00e99c0d15a23649d0db8944380ac81082d4b021f398733dd84f3a6c569a7) +Loading cached processed dataset at theodor1289wit-train[100%] +Loading cached processed dataset at theodor1289wit-test +Restoring states from the checkpoint path at /cluster/work/cotterell/tamariucai/HuggingfaceCheckpoints/flava-wit/text100-vision0/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3990.ckpt +/cluster/home/tamariucai/.cache/pypoetry/virtualenvs/alkmi-CJbY6op3-py3.10/lib64/python3.10/site-packages/pytorch_lightning/callbacks/model_checkpoint.py:337: UserWarning: The dirpath has changed from "/cluster/work/cotterell/tamariucai/HuggingfaceCheckpoints/flava_half-wit/date(2023-08-15)_time(10:33:06)/seed(5501650)-text_perc(100)-vision_perc(0)-magic({'enable': True}" to "/cluster/work/cotterell/tamariucai/HuggingfaceCheckpoints/flava_half-wit/date(2023-08-18)_time(11:16:41)/seed(5501650)-magic({'enable': True})-text_perc(100)-vision_perc(0", therefore `best_model_score`, `kth_best_model_path`, `kth_value`, `last_model_path` and `best_k_models` won't be reloaded. Only `best_model_path` will be reloaded. + warnings.warn( +LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0] + + | Name | Type | Params +---------------------------------------------- +0 | model | FlavaForPreTraining | 189 M +---------------------------------------------- +136 M Trainable params +53.8 M Non-trainable params +189 M Total params +759.666 Total estimated model params size (MB) +Restored all states from the checkpoint at /cluster/work/cotterell/tamariucai/HuggingfaceCheckpoints/flava-wit/text100-vision0/half_bs4080_seed5501650_bf16-mixed/flava-epoch=00-step=3990.ckpt +Compiling FLAVA's image model... +Compiling FLAVA's multimodal model... +Compiling FLAVA's image codebook... +FLAVA will use a different learning rate for its text submodel (0.00075) compared to its other submodels (0.001) +Initializing datamodule +MLMDataModule: batch_size is 48, num_workers is 4 +Only one modality, sampling weight will be [1.0] +2023-08-18 11:18:28.212545 Setting up datamodule: MLMDataModule +Registering multimodal overfitting monitors +Callbacks registered: ['LearningRateMonitor', 'PseudoPerplexityCallback', 'LMEvalHarnessCallback', 'ModelCheckpoint', 'MultimodalOverfittingMonitor'] +Trainer successfully initialized (strategy=SingleDeviceStrategy) +Starting training +2023-08-18 11:37:46.211747 Preparing data for datamodule: MLMDataModule +2023-08-18 11:37:46.217506 Setting up datamodule: MLMDataModule +Resetting wait count for `mlm` since training is resuming and they are special tasks. +2023-08-18 11:56:48.363135 Building train dataloader for datamodule: MLMDataModule +2023-08-18 11:56:50.369680 Building val dataloader for datamodule: MLMDataModule +Found cached dataset json (/cluster/scratch/tamariucai/HuggingfaceDatasets/json/default-7795097c7b181601/0.0.0/8bb11242116d547c741b2e8a1f18598ffdd40a1d4f2a2872c7a28b697434bc96) +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 12:30:31.117975] Computing e^(tensor([5854.2983], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 12:30:31.182789] Ending with PPL=18.674316681422145 (duration: 0:04:11.773315) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.448 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 13:33:13.705835] Computing e^(tensor([5797.6025], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 13:33:13.711104] Ending with PPL=18.152372336883413 (duration: 0:04:06.625458) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.419 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 14:40:23.834766] Computing e^(tensor([5748.7880], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 14:40:23.840083] Ending with PPL=17.714685935014472 (duration: 0:04:07.686097) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.393 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 15:50:00.725552] Computing e^(tensor([5708.4172], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 15:50:00.730856] Ending with PPL=17.360692279081878 (duration: 0:04:11.000440) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.376 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 17:06:01.577975] Computing e^(tensor([5671.3757], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 17:06:01.583349] Ending with PPL=17.04211891949142 (duration: 0:04:06.996822) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.351 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 18:24:19.144117] Computing e^(tensor([5622.1666], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 18:24:19.149432] Ending with PPL=16.62792145416098 (duration: 0:04:10.250881) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.337 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 19:47:12.311438] Computing e^(tensor([5586.3956], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 19:47:12.316683] Ending with PPL=16.33316618694188 (duration: 0:04:06.529068) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.317 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 21:10:58.025457] Computing e^(tensor([5512.2741], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 21:10:58.030875] Ending with PPL=15.738926485914865 (duration: 0:04:05.244375) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.299 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-18 22:32:23.775011] Computing e^(tensor([5480.0492], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-18 22:32:23.780279] Ending with PPL=15.487366238584793 (duration: 0:04:04.702080) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.279 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 00:00:19.336860] Computing e^(tensor([5460.4286], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 00:00:19.342238] Ending with PPL=15.336172955392422 (duration: 0:04:05.950853) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.264 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 01:22:14.909715] Computing e^(tensor([5424.6010], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 01:22:14.915004] Ending with PPL=15.063890371147805 (duration: 0:04:04.558032) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.248 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 02:48:19.012416] Computing e^(tensor([5373.3947], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 02:48:19.020031] Ending with PPL=14.683102694863766 (duration: 0:04:11.139082) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.232 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 04:11:27.735574] Computing e^(tensor([5347.0620], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 04:11:27.740821] Ending with PPL=14.491046686704946 (duration: 0:04:06.528941) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.219 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 05:41:12.705006] Computing e^(tensor([5312.9766], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 05:41:12.710192] Ending with PPL=14.246172678332353 (duration: 0:04:08.745970) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.203 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 07:01:39.639585] Computing e^(tensor([5289.0615], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 07:01:39.645001] Ending with PPL=14.076838258265582 (duration: 0:04:04.903897) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.180 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 08:20:11.424634] Computing e^(tensor([5256.1836], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 08:20:11.429842] Ending with PPL=13.84732142111972 (duration: 0:04:08.907348) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.174 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 09:35:34.982295] Computing e^(tensor([5217.1983], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 09:35:34.987644] Ending with PPL=13.580014028820898 (duration: 0:04:07.126162) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.155 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 10:50:50.206175] Computing e^(tensor([5204.1837], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 10:50:50.211469] Ending with PPL=13.491931685934745 (duration: 0:04:03.309598) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.152 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 12:12:58.707116] Computing e^(tensor([5186.8826], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 12:12:58.712506] Ending with PPL=13.375722621504986 (duration: 0:04:08.043749) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.136 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 13:34:56.547710] Computing e^(tensor([5163.0370], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 13:34:56.553181] Ending with PPL=13.217193373211165 (duration: 0:04:06.506829) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.123 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 14:50:58.651919] Computing e^(tensor([5142.3186], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 14:50:58.657452] Ending with PPL=13.080980704764924 (duration: 0:04:12.355856) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.119 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 15:53:58.350763] Computing e^(tensor([5110.0834], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 15:53:58.356329] Ending with PPL=12.871836164453207 (duration: 0:04:06.401250) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.111 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 16:56:56.054557] Computing e^(tensor([5097.5775], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 16:56:56.060169] Ending with PPL=12.791600742074502 (duration: 0:04:08.100344) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.098 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 18:08:24.364438] Computing e^(tensor([5078.5512], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 18:08:24.369695] Ending with PPL=12.670488977911505 (duration: 0:04:10.742021) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.087 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 19:23:03.243850] Computing e^(tensor([5057.3129], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 19:23:03.249091] Ending with PPL=12.536650948872797 (duration: 0:04:14.751701) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.082 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 20:37:54.294073] Computing e^(tensor([5041.1628], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 20:37:54.299353] Ending with PPL=12.435824802374572 (duration: 0:04:06.727012) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.074 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 21:55:30.644461] Computing e^(tensor([5025.7633], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 21:55:30.649661] Ending with PPL=12.340439492237719 (duration: 0:04:07.563573) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.060 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-19 23:06:47.273268] Computing e^(tensor([5002.3437], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-19 23:06:47.278646] Ending with PPL=12.19677837671533 (duration: 0:04:06.342370) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.055 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 00:15:51.125971] Computing e^(tensor([4992.7054], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 00:15:51.131185] Ending with PPL=12.138141607919353 (duration: 0:04:11.049304) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.043 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 01:34:08.844610] Computing e^(tensor([4974.9525], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 01:34:08.849927] Ending with PPL=12.030874818077981 (duration: 0:04:06.050276) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.035 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 02:45:19.491691] Computing e^(tensor([4960.1071], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 02:45:19.496933] Ending with PPL=11.941903627599615 (duration: 0:04:04.453264) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.033 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 03:57:01.738518] Computing e^(tensor([4944.9001], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 03:57:01.743869] Ending with PPL=11.851448126676102 (duration: 0:04:08.807739) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.019 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 05:16:52.630941] Computing e^(tensor([4930.8073], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 05:16:52.636143] Ending with PPL=11.768231385438824 (duration: 0:04:04.784745) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.018 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 06:28:27.656713] Computing e^(tensor([4914.3553], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 06:28:27.661985] Ending with PPL=11.67182326125226 (duration: 0:04:12.646386) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 3.010 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 07:39:01.427485] Computing e^(tensor([4885.4845], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 07:39:01.432809] Ending with PPL=11.504545629500143 (duration: 0:06:01.725028) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.997 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 08:48:22.195112] Computing e^(tensor([4888.3414], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 08:48:22.200560] Ending with PPL=11.520991394215201 (duration: 0:04:11.536302) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.987 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 11:04:06.710782] Computing e^(tensor([4864.2305], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 11:04:06.716043] Ending with PPL=11.382934502703131 (duration: 0:04:06.550594) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.984 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 13:24:25.027175] Computing e^(tensor([4832.4234], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 13:24:25.032575] Ending with PPL=11.20333733326268 (duration: 0:04:13.794903) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.976 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 14:28:29.424776] Computing e^(tensor([4820.2656], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 14:28:29.430329] Ending with PPL=11.135439599945297 (duration: 0:04:14.160233) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.970 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 15:35:23.684402] Computing e^(tensor([4818.4150], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 15:35:23.689681] Ending with PPL=11.125140740177013 (duration: 0:04:41.968019) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.963 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 16:42:30.939002] Computing e^(tensor([4797.1808], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 16:42:30.944305] Ending with PPL=11.007649230493227 (duration: 0:04:08.643158) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.954 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 17:55:09.660446] Computing e^(tensor([4786.3808], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 17:55:09.665740] Ending with PPL=10.948368142816838 (duration: 0:04:06.614415) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.945 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 20:15:08.123801] Computing e^(tensor([4775.0292], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 20:15:08.129006] Ending with PPL=10.886403411120844 (duration: 0:04:08.468338) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.938 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 21:31:55.055490] Computing e^(tensor([4766.9564], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 21:31:55.060825] Ending with PPL=10.842550071278994 (duration: 0:04:11.926519) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.937 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 22:49:03.014312] Computing e^(tensor([4752.3625], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 22:49:03.019517] Ending with PPL=10.76372055420455 (duration: 0:04:09.794846) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.936 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-20 23:56:46.369425] Computing e^(tensor([4750.1119], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-20 23:56:46.374798] Ending with PPL=10.751614641610134 (duration: 0:04:32.147155) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.925 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 01:11:27.366170] Computing e^(tensor([4731.4294], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 01:11:27.371391] Ending with PPL=10.651648659543163 (duration: 0:04:05.715039) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.921 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 02:21:05.265962] Computing e^(tensor([4743.1062], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 02:21:05.271251] Ending with PPL=10.714019460153027 (duration: 0:04:05.622948) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.920 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 03:32:57.568633] Computing e^(tensor([4716.2548], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 03:32:57.573935] Ending with PPL=10.571137212189424 (duration: 0:04:11.295268) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.917 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 04:39:41.900103] Computing e^(tensor([4705.4605], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 04:39:41.905448] Ending with PPL=10.514236864280958 (duration: 0:04:04.606835) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.907 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 05:48:37.315836] Computing e^(tensor([4705.8025], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 05:48:37.321194] Ending with PPL=10.516035176027536 (duration: 0:04:08.896878) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.904 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 06:58:21.653654] Computing e^(tensor([4710.9679], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 06:58:21.659071] Ending with PPL=10.54323001880966 (duration: 0:04:05.823238) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.902 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 08:08:18.344001] Computing e^(tensor([4684.3312], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 08:08:18.349488] Ending with PPL=10.403742281571613 (duration: 0:04:31.605361) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.900 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 09:19:51.954576] Computing e^(tensor([4672.3352], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 09:19:51.960002] Ending with PPL=10.341527471432507 (duration: 0:04:12.107120) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.893 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 11:48:13.080653] Computing e^(tensor([4660.0735], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 11:48:13.086040] Ending with PPL=10.278319034532913 (duration: 0:04:08.246148) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.885 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 13:00:25.285459] Computing e^(tensor([4656.7201], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 13:00:25.290933] Ending with PPL=10.261099953637759 (duration: 0:04:09.198197) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.882 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 14:09:16.679404] Computing e^(tensor([4656.4192], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 14:09:16.684718] Ending with PPL=10.25955661991089 (duration: 0:04:11.082532) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.876 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 16:30:01.795293] Computing e^(tensor([4643.3288], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 16:30:01.800568] Ending with PPL=10.192624682703888 (duration: 0:04:08.363543) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.870 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 18:59:54.824218] Computing e^(tensor([4617.1822], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 18:59:54.829558] Ending with PPL=10.060240969209659 (duration: 0:04:09.962868) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.868 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 20:18:51.106037] Computing e^(tensor([4611.1823], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 20:18:51.111211] Ending with PPL=10.030105580834839 (duration: 0:04:13.794927) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.861 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 21:34:30.339751] Computing e^(tensor([4604.9462], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 21:34:30.345001] Ending with PPL=9.998880114262002 (duration: 0:04:06.505589) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.857 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-21 22:56:39.652254] Computing e^(tensor([4591.3892], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-21 22:56:39.657521] Ending with PPL=9.931331872230903 (duration: 0:04:12.747888) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.856 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 00:14:44.141074] Computing e^(tensor([4589.2305], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 00:14:44.146278] Ending with PPL=9.920618320419045 (duration: 0:04:12.053340) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.846 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 01:33:28.193503] Computing e^(tensor([4573.6026], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 01:33:28.198808] Ending with PPL=9.843401286615727 (duration: 0:04:13.369971) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.845 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 02:54:14.554362] Computing e^(tensor([4576.4268], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 02:54:14.559604] Ending with PPL=9.857310974731506 (duration: 0:04:11.316089) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.839 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 04:12:25.243481] Computing e^(tensor([4571.0249], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 04:12:25.248709] Ending with PPL=9.830722500788653 (duration: 0:04:12.353502) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.837 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 05:28:03.331057] Computing e^(tensor([4561.8707], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 05:28:03.336266] Ending with PPL=9.785829157329488 (duration: 0:04:11.410911) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.833 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 06:43:51.831242] Computing e^(tensor([4567.3987], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 06:43:51.836562] Ending with PPL=9.812914813826726 (duration: 0:04:10.072543) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.830 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 08:04:36.157418] Computing e^(tensor([4539.6419], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 08:04:36.162904] Ending with PPL=9.67766771575397 (duration: 0:04:08.214351) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.829 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 09:23:10.290078] Computing e^(tensor([4530.7885], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 09:23:10.293903] Ending with PPL=9.634922642540829 (duration: 0:04:07.883473) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.820 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 10:44:04.664827] Computing e^(tensor([4511.3006], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 10:44:04.670138] Ending with PPL=9.541496260692693 (duration: 0:04:11.605005) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.817 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 12:12:13.993848] Computing e^(tensor([4511.9349], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 12:12:13.999077] Ending with PPL=9.544522706595622 (duration: 0:04:09.858757) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.812 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 14:01:36.359471] Computing e^(tensor([4502.8925], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 14:01:36.365235] Ending with PPL=9.501467483312117 (duration: 0:04:11.727102) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.810 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 15:38:42.915752] Computing e^(tensor([4496.3426], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 15:38:42.921300] Ending with PPL=9.470401469945948 (duration: 0:04:17.245874) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.805 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 17:15:29.572676] Computing e^(tensor([4501.0965], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 17:15:29.576775] Ending with PPL=9.49293911158625 (duration: 0:05:19.592848) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.804 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 18:40:40.777777] Computing e^(tensor([4495.4599], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 18:40:40.783293] Ending with PPL=9.466222557696021 (duration: 0:04:15.319902) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.802 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 20:13:07.431546] Computing e^(tensor([4491.4211], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 20:13:07.437191] Ending with PPL=9.447125791761222 (duration: 0:04:11.657171) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.796 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 21:48:32.701316] Computing e^(tensor([4469.2196], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 21:48:32.706953] Ending with PPL=9.342835754923493 (duration: 0:04:55.971797) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.791 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-22 23:25:59.510689] Computing e^(tensor([4478.0459], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-22 23:25:59.516123] Ending with PPL=9.384158145590694 (duration: 0:04:09.522414) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.791 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-23 02:38:34.354474] Computing e^(tensor([4455.7269], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-23 02:38:34.359741] Ending with PPL=9.280017616339524 (duration: 0:04:13.334302) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.777 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-23 04:20:14.197094] Computing e^(tensor([4459.1773], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-23 04:20:14.202428] Ending with PPL=9.29604137134959 (duration: 0:04:07.422752) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.776 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-23 05:49:35.237109] Computing e^(tensor([4457.5102], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-23 05:49:35.242515] Ending with PPL=9.288295800581354 (duration: 0:04:11.142845) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.773 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-23 10:50:17.925921] Computing e^(tensor([4443.7614], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-23 10:50:17.931201] Ending with PPL=9.224663297913382 (duration: 0:04:08.446181) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.769 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 09:04:01.151777] Computing e^(tensor([4435.6536], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 09:04:01.157087] Ending with PPL=9.187343147441634 (duration: 0:04:06.072894) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.768 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 10:17:33.831776] Computing e^(tensor([4411.8024], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 10:17:33.837032] Ending with PPL=9.078429531008014 (duration: 0:04:09.685901) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.766 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 11:27:20.845183] Computing e^(tensor([4413.9332], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 11:27:20.850530] Ending with PPL=9.088106614563467 (duration: 0:04:09.169997) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.756 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 12:35:16.012693] Computing e^(tensor([4407.2476], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 12:35:16.018103] Ending with PPL=9.057777501494582 (duration: 0:04:08.762557) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.753 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 14:29:48.023318] Computing e^(tensor([4384.5908], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 14:29:48.028629] Ending with PPL=8.955746497936126 (duration: 0:04:07.205418) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.750 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 17:59:59.303787] Computing e^(tensor([4377.9996], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 17:59:59.309172] Ending with PPL=8.9262804114866 (duration: 0:04:12.652538) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.748 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 19:29:44.601646] Computing e^(tensor([4365.5540], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 19:29:44.607023] Ending with PPL=8.870906681837365 (duration: 0:04:11.592946) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.744 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 20:48:37.423923] Computing e^(tensor([4374.5143], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 20:48:37.429281] Ending with PPL=8.910738690201045 (duration: 0:04:10.036115) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.742 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-24 22:13:36.034476] Computing e^(tensor([4361.7591], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-24 22:13:36.039827] Ending with PPL=8.854090425096933 (duration: 0:04:13.189269) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.732 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 01:11:57.930029] Computing e^(tensor([4356.4631], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 01:11:57.935569] Ending with PPL=8.830675685047595 (duration: 0:04:16.228542) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.730 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 04:01:25.861141] Computing e^(tensor([4358.7021], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 04:01:25.866504] Ending with PPL=8.840567335374116 (duration: 0:04:08.818851) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.727 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 06:48:36.150331] Computing e^(tensor([4343.8268], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 06:48:36.155783] Ending with PPL=8.775058369240591 (duration: 0:04:13.829226) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.724 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 08:17:12.673572] Computing e^(tensor([4324.9447], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 08:17:12.678946] Ending with PPL=8.692602161033825 (duration: 0:04:16.771953) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.721 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 12:41:24.321911] Computing e^(tensor([4330.0550], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 12:41:24.327356] Ending with PPL=8.714841436398697 (duration: 0:04:15.603147) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.720 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 14:13:01.734613] Computing e^(tensor([4327.3482], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 14:13:01.739942] Ending with PPL=8.703054828409858 (duration: 0:04:13.338853) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.717 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 15:46:58.768597] Computing e^(tensor([4316.7414], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 15:46:58.774050] Ending with PPL=8.657021400393226 (duration: 0:04:16.603764) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.715 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 18:53:54.053859] Computing e^(tensor([4312.5284], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 18:53:54.059271] Ending with PPL=8.638804636015111 (duration: 0:04:08.311607) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.712 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 20:29:26.242217] Computing e^(tensor([4303.1128], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 20:29:26.247622] Ending with PPL=8.598230260084058 (duration: 0:04:13.428419) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.707 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-25 22:07:19.674858] Computing e^(tensor([4307.5211], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-25 22:07:19.680275] Ending with PPL=8.617202754779155 (duration: 0:04:13.454478) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.700 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-26 02:51:58.418197] Computing e^(tensor([4293.1118], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-26 02:51:58.423730] Ending with PPL=8.555342245197496 (duration: 0:04:13.371352) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.698 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-26 07:33:26.764842] Computing e^(tensor([4288.7298], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-26 07:33:26.770135] Ending with PPL=8.536617882180927 (duration: 0:04:15.071664) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00= min_delta = 0.0. New best score: 2.691 +[PPL Evaluation] Starting from index 0 to index 2000. +[PPL Evaluation 2023-08-26 10:44:41.790045] Computing e^(tensor([4288.6244], dtype=torch.float64) / 2000) +[PPL Evaluation 2023-08-26 10:44:41.794006] Ending with PPL=8.536168053752988 (duration: 0:04:15.254232) +Starting LM Evaluation Harness +Running on blimp... + 0%| | 0/1 [00:00