Training in progress, step 156, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:510e88c394d7b67aa131e16eb2d4057894848ee91683b790cfe04a0ffd4371fe
+oid sha256:8546044019154ce3968e6a6ba6a86f7aa3175824cef87135b2149628694d455c
 size 550593856

last-checkpoint/global_step156/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fb5be0a5f26cb07c799f46bb1862098d91fa16518bfc8875ac9fcc5e7740561
+size 243590592

last-checkpoint/global_step156/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa645ee95c39b5b497ff5f86cfd818320fe51bb72fd47f138e22d48a26dab541
+size 243590592

last-checkpoint/global_step156/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:210fc2037e25678118657e27309536a0fa50b5ba86c636be21bdb2209f2dec97
+size 243590592

last-checkpoint/global_step156/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec460de6bd172eae661c430cf806a16a0c1607eb52adf5bc5c7e171411e85ab6
+size 243590592

last-checkpoint/global_step156/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95dd69a185a59513ffa3ff81a5b3f6388c84710694de7b43dfb078f8c70f1e85
+size 243590592

last-checkpoint/global_step156/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23918528a67b93abd5356257022ed6da16f23ebcee411b5992ff681cea30903d
+size 243590592

last-checkpoint/global_step156/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82a6353c0c6295705f9ded5e47fb1513485a975279c507405959431a9d1cc963
+size 243590592

last-checkpoint/global_step156/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:164b62b107af40c179352364ffb2009aec6bfe4fd593b2869afded7221558fe1
+size 243590592

last-checkpoint/global_step156/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:debf1185e8b75b8e355f0cee329a6fcb38feca0910d724b7eb7dc2ec8a38efd6
+size 211435686

last-checkpoint/global_step156/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da1a2705bb1b4552cdea7af94d241262202b6b8a67c862f451ac3d7b98c34e8b
+size 211435686

last-checkpoint/global_step156/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8f491814d9c49f2165e4d8c5b31531f16593aaf1661d2506585d3e6b993eeb9
+size 211435686

last-checkpoint/global_step156/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48be9ef8d59d015becf0ed68c0f3bd3cc5cf2a42ccf7623e178b595c196e6ec9
+size 211435686

last-checkpoint/global_step156/zero_pp_rank_4_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a82ac3de9087af93f59e8236dacc634c86e0e30a109e3ef84dba0aab387c66a
+size 211435686

last-checkpoint/global_step156/zero_pp_rank_5_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7e8cc083f831ef3443f1f1702cde406b2390f4d408a93dca6eb700156c23082
+size 211435686

last-checkpoint/global_step156/zero_pp_rank_6_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c36dd49ee3aae7b73530658214d7872f7c323aa63363c6bb64e5cf2444719300
+size 211435686

last-checkpoint/global_step156/zero_pp_rank_7_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:766f01511652728b366ce75467fcb3e68f9361004823a91082d03ec486dd3f23
+size 211435686

last-checkpoint/latest

@@ -1 +1 @@
-global_step117
+global_step156

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e1c39d9c1614e5a5db49c2adcaea58c50f91741ffdc85357e20965a732ab918
+oid sha256:a4c561e839b2d36fc5a0e0f7b6936e77fe5a5c8df4881367ae16771a48d03931
 size 15984

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:302ede7ac90ff693eab149e42ad40ee9fc950c1ef8919c67ed57abb0686bbbe5
+oid sha256:bb40f6e44cd62eb5b3d663b4cd4a22621261ff4270ac7334a82a6c5654582d4f
 size 15984

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cd760eaa5a69675cd82d86a72d467afbc30b5630d5157f32efb4ea7ae17fc2f2
+oid sha256:809d109a794ef0cf1a0ad4ebb3119098360aab2adb6f08bddff73f40f315a7e5
 size 15984

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ba53ce4c9ec1239ebb539d4b9d5b4d60afc5de8260b83004180fbef5fa1ad12
+oid sha256:b412bb58b079b8554c3ec8e287b647f0d5eb9af588672a244913fe25977d512c
 size 15984

last-checkpoint/rng_state_4.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:37f68cc323d6a100d3e96e363031f94ca17676b3152edfe25789da3092f4e267
+oid sha256:500bd67adda8cf35848a173b29f37f06b58546a6456d0acafb829bcd30658146
 size 15984

last-checkpoint/rng_state_5.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f9ed76e42e0966ecee5df95b284ed0f8c2c9eef1e477abfa1252598e3566c6a
+oid sha256:a2ef4aff8cc70904fad22aaf8933fc74acea3612ad213c1998d6fbe418d76763
 size 15984

last-checkpoint/rng_state_6.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:81c7b8456e2d33481d89db677988b4ad26b87e262b4951b9a31bbe8c34add21b
+oid sha256:f5aa0392cb422a7717e1c27ead3b9ec1f2b7fa04f0e7c4a432e11c7d9fd4d50c
 size 15984

last-checkpoint/rng_state_7.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6154b2373b0985a816ea00d8d6b2c75a97f53aa8e7a463dabbbf8cd42f0a62c
+oid sha256:0d820cd9be3a77ff0f6d736eb3ab23e99a61accdfc90219866962f22853a7422
 size 15984

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:317bfb15ccbbc1daa774ce78869237c5810b7c8eb3f18edff11bd8175bd36feb
+oid sha256:22af69e6b655a005d672b3a204eb6484f3ef29e640fcbdffa3f8d613a5d04cf0
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.3038961038961039,
+  "epoch": 0.4051948051948052,
   "eval_steps": 39,
-  "global_step": 117,
+  "global_step": 156,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -858,6 +858,287 @@
       "eval_samples_per_second": 1.233,
       "eval_steps_per_second": 0.154,
       "step": 117
+    },
+    {
+      "epoch": 0.3064935064935065,
+      "grad_norm": 1.11071544369134,
+      "learning_rate": 8.49184929083929e-05,
+      "loss": 2.6445,
+      "step": 118
+    },
+    {
+      "epoch": 0.3090909090909091,
+      "grad_norm": 1.1571880416140317,
+      "learning_rate": 8.462807131141221e-05,
+      "loss": 2.5551,
+      "step": 119
+    },
+    {
+      "epoch": 0.3116883116883117,
+      "grad_norm": 0.9303010187060028,
+      "learning_rate": 8.433545481454206e-05,
+      "loss": 2.3873,
+      "step": 120
+    },
+    {
+      "epoch": 0.3142857142857143,
+      "grad_norm": 0.9575555447726172,
+      "learning_rate": 8.404066509532978e-05,
+      "loss": 2.5088,
+      "step": 121
+    },
+    {
+      "epoch": 0.3168831168831169,
+      "grad_norm": 0.9710292387067033,
+      "learning_rate": 8.374372399231873e-05,
+      "loss": 2.4983,
+      "step": 122
+    },
+    {
+      "epoch": 0.3194805194805195,
+      "grad_norm": 1.048408170787146,
+      "learning_rate": 8.344465350343069e-05,
+      "loss": 2.3834,
+      "step": 123
+    },
+    {
+      "epoch": 0.3220779220779221,
+      "grad_norm": 0.9705702680470921,
+      "learning_rate": 8.314347578433608e-05,
+      "loss": 2.4382,
+      "step": 124
+    },
+    {
+      "epoch": 0.3246753246753247,
+      "grad_norm": 0.9614277096166124,
+      "learning_rate": 8.284021314681265e-05,
+      "loss": 2.3393,
+      "step": 125
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 0.9281949363762984,
+      "learning_rate": 8.253488805709259e-05,
+      "loss": 2.6182,
+      "step": 126
+    },
+    {
+      "epoch": 0.32987012987012987,
+      "grad_norm": 0.9623514960091324,
+      "learning_rate": 8.222752313419822e-05,
+      "loss": 2.4305,
+      "step": 127
+    },
+    {
+      "epoch": 0.33246753246753247,
+      "grad_norm": 0.9531142482825348,
+      "learning_rate": 8.191814114826625e-05,
+      "loss": 2.5015,
+      "step": 128
+    },
+    {
+      "epoch": 0.33506493506493507,
+      "grad_norm": 1.031934013410237,
+      "learning_rate": 8.160676501886102e-05,
+      "loss": 2.4596,
+      "step": 129
+    },
+    {
+      "epoch": 0.33766233766233766,
+      "grad_norm": 1.055900967871222,
+      "learning_rate": 8.129341781327658e-05,
+      "loss": 2.7157,
+      "step": 130
+    },
+    {
+      "epoch": 0.34025974025974026,
+      "grad_norm": 0.9833403621365481,
+      "learning_rate": 8.097812274482774e-05,
+      "loss": 2.3566,
+      "step": 131
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 1.0871003085779585,
+      "learning_rate": 8.066090317113043e-05,
+      "loss": 2.4638,
+      "step": 132
+    },
+    {
+      "epoch": 0.34545454545454546,
+      "grad_norm": 0.9406528674917229,
+      "learning_rate": 8.034178259237134e-05,
+      "loss": 2.2718,
+      "step": 133
+    },
+    {
+      "epoch": 0.34805194805194806,
+      "grad_norm": 0.9582596034300516,
+      "learning_rate": 8.002078464956701e-05,
+      "loss": 2.4254,
+      "step": 134
+    },
+    {
+      "epoch": 0.35064935064935066,
+      "grad_norm": 1.0907653310302854,
+      "learning_rate": 7.969793312281237e-05,
+      "loss": 2.4145,
+      "step": 135
+    },
+    {
+      "epoch": 0.35324675324675325,
+      "grad_norm": 1.0102280592637967,
+      "learning_rate": 7.937325192951917e-05,
+      "loss": 2.4132,
+      "step": 136
+    },
+    {
+      "epoch": 0.35584415584415585,
+      "grad_norm": 1.06644159749959,
+      "learning_rate": 7.904676512264406e-05,
+      "loss": 2.4369,
+      "step": 137
+    },
+    {
+      "epoch": 0.35844155844155845,
+      "grad_norm": 0.9673017199924268,
+      "learning_rate": 7.871849688890674e-05,
+      "loss": 2.4244,
+      "step": 138
+    },
+    {
+      "epoch": 0.36103896103896105,
+      "grad_norm": 1.2268344248013403,
+      "learning_rate": 7.838847154699821e-05,
+      "loss": 2.2285,
+      "step": 139
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 1.073462737631204,
+      "learning_rate": 7.805671354577908e-05,
+      "loss": 2.3111,
+      "step": 140
+    },
+    {
+      "epoch": 0.36623376623376624,
+      "grad_norm": 1.574004120111183,
+      "learning_rate": 7.772324746246842e-05,
+      "loss": 2.4494,
+      "step": 141
+    },
+    {
+      "epoch": 0.36883116883116884,
+      "grad_norm": 0.9241427879924983,
+      "learning_rate": 7.738809800082314e-05,
+      "loss": 2.3341,
+      "step": 142
+    },
+    {
+      "epoch": 0.37142857142857144,
+      "grad_norm": 0.8766148723800662,
+      "learning_rate": 7.705128998930766e-05,
+      "loss": 2.2851,
+      "step": 143
+    },
+    {
+      "epoch": 0.37402597402597404,
+      "grad_norm": 1.0218314429150428,
+      "learning_rate": 7.671284837925483e-05,
+      "loss": 2.3307,
+      "step": 144
+    },
+    {
+      "epoch": 0.37662337662337664,
+      "grad_norm": 1.0085214768978241,
+      "learning_rate": 7.637279824301728e-05,
+      "loss": 2.407,
+      "step": 145
+    },
+    {
+      "epoch": 0.37922077922077924,
+      "grad_norm": 1.1498069068985362,
+      "learning_rate": 7.60311647721101e-05,
+      "loss": 2.3903,
+      "step": 146
+    },
+    {
+      "epoch": 0.38181818181818183,
+      "grad_norm": 0.9898512411601635,
+      "learning_rate": 7.56879732753447e-05,
+      "loss": 2.4284,
+      "step": 147
+    },
+    {
+      "epoch": 0.38441558441558443,
+      "grad_norm": 1.0768194487299194,
+      "learning_rate": 7.53432491769537e-05,
+      "loss": 2.4555,
+      "step": 148
+    },
+    {
+      "epoch": 0.38701298701298703,
+      "grad_norm": 1.2604006103307979,
+      "learning_rate": 7.49970180147076e-05,
+      "loss": 2.5891,
+      "step": 149
+    },
+    {
+      "epoch": 0.38961038961038963,
+      "grad_norm": 1.0237041933548676,
+      "learning_rate": 7.464930543802289e-05,
+      "loss": 2.4212,
+      "step": 150
+    },
+    {
+      "epoch": 0.3922077922077922,
+      "grad_norm": 1.0931014839822084,
+      "learning_rate": 7.430013720606176e-05,
+      "loss": 2.4502,
+      "step": 151
+    },
+    {
+      "epoch": 0.3948051948051948,
+      "grad_norm": 1.2220277182692068,
+      "learning_rate": 7.394953918582403e-05,
+      "loss": 2.4569,
+      "step": 152
+    },
+    {
+      "epoch": 0.3974025974025974,
+      "grad_norm": 1.1663502019072647,
+      "learning_rate": 7.35975373502307e-05,
+      "loss": 2.3729,
+      "step": 153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0686488749921856,
+      "learning_rate": 7.324415777619988e-05,
+      "loss": 2.3302,
+      "step": 154
+    },
+    {
+      "epoch": 0.4025974025974026,
+      "grad_norm": 0.929261927962379,
+      "learning_rate": 7.288942664271503e-05,
+      "loss": 2.2966,
+      "step": 155
+    },
+    {
+      "epoch": 0.4051948051948052,
+      "grad_norm": 0.957275997972424,
+      "learning_rate": 7.253337022888546e-05,
+      "loss": 2.2596,
+      "step": 156
+    },
+    {
+      "epoch": 0.4051948051948052,
+      "eval_loss": 2.42930269241333,
+      "eval_runtime": 64.9704,
+      "eval_samples_per_second": 1.231,
+      "eval_steps_per_second": 0.154,
+      "step": 156
     }
   ],
   "logging_steps": 1,
@@ -877,7 +1158,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 38677656895488.0,
+  "total_flos": 51570209193984.0,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null