Training in progress, step 7500, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4523108832
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:430cdd15f16a4454ce99c1eefc5c617f45181ccd71b78f820c1efeb3d3a232ab
|
| 3 |
size 4523108832
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2912179275
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5751a58db1d09e6480d8b29a14bd911adf8e9a0bd929a29507334812c3c0dae
|
| 3 |
size 2912179275
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a392b10968451c357d65996bfcadc4a9aa4ec0990c7a29affa02cf58b1efdc45
|
| 3 |
size 14645
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a3cfb61f5c0f850324c2cee3170c15199945aeab36c6c3dd0afd3a9ec2feea0
|
| 3 |
size 1383
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2edcc4543530e4932ef7330778fea25bdcdb4cab9e40ff4265acd64a703f2c77
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -4908,6 +4908,356 @@
|
|
| 4908 |
"learning_rate": 6.004000000000001e-05,
|
| 4909 |
"loss": 0.02357790470123291,
|
| 4910 |
"step": 7000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4911 |
}
|
| 4912 |
],
|
| 4913 |
"logging_steps": 10,
|
|
@@ -4927,7 +5277,7 @@
|
|
| 4927 |
"attributes": {}
|
| 4928 |
}
|
| 4929 |
},
|
| 4930 |
-
"total_flos":
|
| 4931 |
"train_batch_size": 1,
|
| 4932 |
"trial_name": null,
|
| 4933 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 4.0064,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 7500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 4908 |
"learning_rate": 6.004000000000001e-05,
|
| 4909 |
"loss": 0.02357790470123291,
|
| 4910 |
"step": 7000
|
| 4911 |
+
},
|
| 4912 |
+
{
|
| 4913 |
+
"epoch": 2.0198,
|
| 4914 |
+
"grad_norm": 0.17045632004737854,
|
| 4915 |
+
"learning_rate": 5.984000000000001e-05,
|
| 4916 |
+
"loss": 0.02287629395723343,
|
| 4917 |
+
"step": 7010
|
| 4918 |
+
},
|
| 4919 |
+
{
|
| 4920 |
+
"epoch": 2.0208,
|
| 4921 |
+
"grad_norm": 0.12083585560321808,
|
| 4922 |
+
"learning_rate": 5.9640000000000005e-05,
|
| 4923 |
+
"loss": 0.02358904778957367,
|
| 4924 |
+
"step": 7020
|
| 4925 |
+
},
|
| 4926 |
+
{
|
| 4927 |
+
"epoch": 2.0218,
|
| 4928 |
+
"grad_norm": 0.13300317525863647,
|
| 4929 |
+
"learning_rate": 5.944000000000001e-05,
|
| 4930 |
+
"loss": 0.023620645701885223,
|
| 4931 |
+
"step": 7030
|
| 4932 |
+
},
|
| 4933 |
+
{
|
| 4934 |
+
"epoch": 2.0228,
|
| 4935 |
+
"grad_norm": 0.15547838807106018,
|
| 4936 |
+
"learning_rate": 5.924000000000001e-05,
|
| 4937 |
+
"loss": 0.023183451592922212,
|
| 4938 |
+
"step": 7040
|
| 4939 |
+
},
|
| 4940 |
+
{
|
| 4941 |
+
"epoch": 2.0238,
|
| 4942 |
+
"grad_norm": 0.16687844693660736,
|
| 4943 |
+
"learning_rate": 5.9040000000000004e-05,
|
| 4944 |
+
"loss": 0.023912055790424346,
|
| 4945 |
+
"step": 7050
|
| 4946 |
+
},
|
| 4947 |
+
{
|
| 4948 |
+
"epoch": 2.0248,
|
| 4949 |
+
"grad_norm": 0.12975580990314484,
|
| 4950 |
+
"learning_rate": 5.8840000000000006e-05,
|
| 4951 |
+
"loss": 0.023545092344284056,
|
| 4952 |
+
"step": 7060
|
| 4953 |
+
},
|
| 4954 |
+
{
|
| 4955 |
+
"epoch": 2.0258,
|
| 4956 |
+
"grad_norm": 0.11908718943595886,
|
| 4957 |
+
"learning_rate": 5.864000000000001e-05,
|
| 4958 |
+
"loss": 0.024130003154277803,
|
| 4959 |
+
"step": 7070
|
| 4960 |
+
},
|
| 4961 |
+
{
|
| 4962 |
+
"epoch": 2.0268,
|
| 4963 |
+
"grad_norm": 0.13249467313289642,
|
| 4964 |
+
"learning_rate": 5.844e-05,
|
| 4965 |
+
"loss": 0.02512686550617218,
|
| 4966 |
+
"step": 7080
|
| 4967 |
+
},
|
| 4968 |
+
{
|
| 4969 |
+
"epoch": 2.0278,
|
| 4970 |
+
"grad_norm": 0.12178528308868408,
|
| 4971 |
+
"learning_rate": 5.8240000000000005e-05,
|
| 4972 |
+
"loss": 0.02444648593664169,
|
| 4973 |
+
"step": 7090
|
| 4974 |
+
},
|
| 4975 |
+
{
|
| 4976 |
+
"epoch": 2.0288,
|
| 4977 |
+
"grad_norm": 0.12476595491170883,
|
| 4978 |
+
"learning_rate": 5.804000000000001e-05,
|
| 4979 |
+
"loss": 0.022948546707630156,
|
| 4980 |
+
"step": 7100
|
| 4981 |
+
},
|
| 4982 |
+
{
|
| 4983 |
+
"epoch": 2.0298,
|
| 4984 |
+
"grad_norm": 0.15463438630104065,
|
| 4985 |
+
"learning_rate": 5.784000000000001e-05,
|
| 4986 |
+
"loss": 0.025882315635681153,
|
| 4987 |
+
"step": 7110
|
| 4988 |
+
},
|
| 4989 |
+
{
|
| 4990 |
+
"epoch": 2.0308,
|
| 4991 |
+
"grad_norm": 0.16765928268432617,
|
| 4992 |
+
"learning_rate": 5.7640000000000004e-05,
|
| 4993 |
+
"loss": 0.024091361463069914,
|
| 4994 |
+
"step": 7120
|
| 4995 |
+
},
|
| 4996 |
+
{
|
| 4997 |
+
"epoch": 3.0006,
|
| 4998 |
+
"grad_norm": 0.08396776765584946,
|
| 4999 |
+
"learning_rate": 5.7440000000000006e-05,
|
| 5000 |
+
"loss": 0.02553088665008545,
|
| 5001 |
+
"step": 7130
|
| 5002 |
+
},
|
| 5003 |
+
{
|
| 5004 |
+
"epoch": 3.0016,
|
| 5005 |
+
"grad_norm": 0.10819737613201141,
|
| 5006 |
+
"learning_rate": 5.724000000000001e-05,
|
| 5007 |
+
"loss": 0.019945000112056733,
|
| 5008 |
+
"step": 7140
|
| 5009 |
+
},
|
| 5010 |
+
{
|
| 5011 |
+
"epoch": 3.0026,
|
| 5012 |
+
"grad_norm": 0.1035761684179306,
|
| 5013 |
+
"learning_rate": 5.704e-05,
|
| 5014 |
+
"loss": 0.021518710255622863,
|
| 5015 |
+
"step": 7150
|
| 5016 |
+
},
|
| 5017 |
+
{
|
| 5018 |
+
"epoch": 3.0036,
|
| 5019 |
+
"grad_norm": 0.11577849835157394,
|
| 5020 |
+
"learning_rate": 5.6840000000000005e-05,
|
| 5021 |
+
"loss": 0.020588450133800507,
|
| 5022 |
+
"step": 7160
|
| 5023 |
+
},
|
| 5024 |
+
{
|
| 5025 |
+
"epoch": 3.0046,
|
| 5026 |
+
"grad_norm": 0.0943162813782692,
|
| 5027 |
+
"learning_rate": 5.6640000000000007e-05,
|
| 5028 |
+
"loss": 0.019012409448623657,
|
| 5029 |
+
"step": 7170
|
| 5030 |
+
},
|
| 5031 |
+
{
|
| 5032 |
+
"epoch": 3.0056,
|
| 5033 |
+
"grad_norm": 0.10145782679319382,
|
| 5034 |
+
"learning_rate": 5.644e-05,
|
| 5035 |
+
"loss": 0.020459675788879396,
|
| 5036 |
+
"step": 7180
|
| 5037 |
+
},
|
| 5038 |
+
{
|
| 5039 |
+
"epoch": 3.0066,
|
| 5040 |
+
"grad_norm": 0.11473723500967026,
|
| 5041 |
+
"learning_rate": 5.6240000000000004e-05,
|
| 5042 |
+
"loss": 0.02082264721393585,
|
| 5043 |
+
"step": 7190
|
| 5044 |
+
},
|
| 5045 |
+
{
|
| 5046 |
+
"epoch": 3.0076,
|
| 5047 |
+
"grad_norm": 0.14834633469581604,
|
| 5048 |
+
"learning_rate": 5.6040000000000006e-05,
|
| 5049 |
+
"loss": 0.02009241431951523,
|
| 5050 |
+
"step": 7200
|
| 5051 |
+
},
|
| 5052 |
+
{
|
| 5053 |
+
"epoch": 3.0086,
|
| 5054 |
+
"grad_norm": 0.18625617027282715,
|
| 5055 |
+
"learning_rate": 5.584e-05,
|
| 5056 |
+
"loss": 0.022254915535449983,
|
| 5057 |
+
"step": 7210
|
| 5058 |
+
},
|
| 5059 |
+
{
|
| 5060 |
+
"epoch": 3.0096,
|
| 5061 |
+
"grad_norm": 0.09577304124832153,
|
| 5062 |
+
"learning_rate": 5.564e-05,
|
| 5063 |
+
"loss": 0.02103894203901291,
|
| 5064 |
+
"step": 7220
|
| 5065 |
+
},
|
| 5066 |
+
{
|
| 5067 |
+
"epoch": 3.0106,
|
| 5068 |
+
"grad_norm": 0.11397113651037216,
|
| 5069 |
+
"learning_rate": 5.5440000000000005e-05,
|
| 5070 |
+
"loss": 0.020081984996795654,
|
| 5071 |
+
"step": 7230
|
| 5072 |
+
},
|
| 5073 |
+
{
|
| 5074 |
+
"epoch": 3.0116,
|
| 5075 |
+
"grad_norm": 0.11981774121522903,
|
| 5076 |
+
"learning_rate": 5.524e-05,
|
| 5077 |
+
"loss": 0.020449602603912355,
|
| 5078 |
+
"step": 7240
|
| 5079 |
+
},
|
| 5080 |
+
{
|
| 5081 |
+
"epoch": 3.0126,
|
| 5082 |
+
"grad_norm": 0.12391442805528641,
|
| 5083 |
+
"learning_rate": 5.504e-05,
|
| 5084 |
+
"loss": 0.021178384125232697,
|
| 5085 |
+
"step": 7250
|
| 5086 |
+
},
|
| 5087 |
+
{
|
| 5088 |
+
"epoch": 3.0136,
|
| 5089 |
+
"grad_norm": 0.10961490869522095,
|
| 5090 |
+
"learning_rate": 5.4840000000000003e-05,
|
| 5091 |
+
"loss": 0.020945684611797334,
|
| 5092 |
+
"step": 7260
|
| 5093 |
+
},
|
| 5094 |
+
{
|
| 5095 |
+
"epoch": 3.0146,
|
| 5096 |
+
"grad_norm": 0.13801056146621704,
|
| 5097 |
+
"learning_rate": 5.4640000000000005e-05,
|
| 5098 |
+
"loss": 0.01994608938694,
|
| 5099 |
+
"step": 7270
|
| 5100 |
+
},
|
| 5101 |
+
{
|
| 5102 |
+
"epoch": 3.0156,
|
| 5103 |
+
"grad_norm": 0.09954143315553665,
|
| 5104 |
+
"learning_rate": 5.444e-05,
|
| 5105 |
+
"loss": 0.0260534793138504,
|
| 5106 |
+
"step": 7280
|
| 5107 |
+
},
|
| 5108 |
+
{
|
| 5109 |
+
"epoch": 3.0166,
|
| 5110 |
+
"grad_norm": 0.09468942880630493,
|
| 5111 |
+
"learning_rate": 5.424e-05,
|
| 5112 |
+
"loss": 0.019537252187728883,
|
| 5113 |
+
"step": 7290
|
| 5114 |
+
},
|
| 5115 |
+
{
|
| 5116 |
+
"epoch": 3.0176,
|
| 5117 |
+
"grad_norm": 0.09855340421199799,
|
| 5118 |
+
"learning_rate": 5.4040000000000004e-05,
|
| 5119 |
+
"loss": 0.021300722658634186,
|
| 5120 |
+
"step": 7300
|
| 5121 |
+
},
|
| 5122 |
+
{
|
| 5123 |
+
"epoch": 3.0186,
|
| 5124 |
+
"grad_norm": 0.14724481105804443,
|
| 5125 |
+
"learning_rate": 5.384e-05,
|
| 5126 |
+
"loss": 0.021099399030208587,
|
| 5127 |
+
"step": 7310
|
| 5128 |
+
},
|
| 5129 |
+
{
|
| 5130 |
+
"epoch": 3.0196,
|
| 5131 |
+
"grad_norm": 0.1300877034664154,
|
| 5132 |
+
"learning_rate": 5.364e-05,
|
| 5133 |
+
"loss": 0.021501822769641875,
|
| 5134 |
+
"step": 7320
|
| 5135 |
+
},
|
| 5136 |
+
{
|
| 5137 |
+
"epoch": 3.0206,
|
| 5138 |
+
"grad_norm": 0.09704280644655228,
|
| 5139 |
+
"learning_rate": 5.344e-05,
|
| 5140 |
+
"loss": 0.020881634950637818,
|
| 5141 |
+
"step": 7330
|
| 5142 |
+
},
|
| 5143 |
+
{
|
| 5144 |
+
"epoch": 3.0216,
|
| 5145 |
+
"grad_norm": 0.11273547261953354,
|
| 5146 |
+
"learning_rate": 5.324e-05,
|
| 5147 |
+
"loss": 0.02265263944864273,
|
| 5148 |
+
"step": 7340
|
| 5149 |
+
},
|
| 5150 |
+
{
|
| 5151 |
+
"epoch": 3.0226,
|
| 5152 |
+
"grad_norm": 0.116827592253685,
|
| 5153 |
+
"learning_rate": 5.304e-05,
|
| 5154 |
+
"loss": 0.021155886352062225,
|
| 5155 |
+
"step": 7350
|
| 5156 |
+
},
|
| 5157 |
+
{
|
| 5158 |
+
"epoch": 3.0236,
|
| 5159 |
+
"grad_norm": 0.13560475409030914,
|
| 5160 |
+
"learning_rate": 5.284e-05,
|
| 5161 |
+
"loss": 0.02109742909669876,
|
| 5162 |
+
"step": 7360
|
| 5163 |
+
},
|
| 5164 |
+
{
|
| 5165 |
+
"epoch": 3.0246,
|
| 5166 |
+
"grad_norm": 0.11204122006893158,
|
| 5167 |
+
"learning_rate": 5.264e-05,
|
| 5168 |
+
"loss": 0.021917839348316193,
|
| 5169 |
+
"step": 7370
|
| 5170 |
+
},
|
| 5171 |
+
{
|
| 5172 |
+
"epoch": 3.0256,
|
| 5173 |
+
"grad_norm": 0.126216858625412,
|
| 5174 |
+
"learning_rate": 5.244e-05,
|
| 5175 |
+
"loss": 0.021687588095664977,
|
| 5176 |
+
"step": 7380
|
| 5177 |
+
},
|
| 5178 |
+
{
|
| 5179 |
+
"epoch": 3.0266,
|
| 5180 |
+
"grad_norm": 0.10559123009443283,
|
| 5181 |
+
"learning_rate": 5.224e-05,
|
| 5182 |
+
"loss": 0.023697364330291747,
|
| 5183 |
+
"step": 7390
|
| 5184 |
+
},
|
| 5185 |
+
{
|
| 5186 |
+
"epoch": 3.0276,
|
| 5187 |
+
"grad_norm": 0.11427237838506699,
|
| 5188 |
+
"learning_rate": 5.204e-05,
|
| 5189 |
+
"loss": 0.022155900299549103,
|
| 5190 |
+
"step": 7400
|
| 5191 |
+
},
|
| 5192 |
+
{
|
| 5193 |
+
"epoch": 3.0286,
|
| 5194 |
+
"grad_norm": 0.14890673756599426,
|
| 5195 |
+
"learning_rate": 5.184e-05,
|
| 5196 |
+
"loss": 0.021498112380504607,
|
| 5197 |
+
"step": 7410
|
| 5198 |
+
},
|
| 5199 |
+
{
|
| 5200 |
+
"epoch": 3.0296,
|
| 5201 |
+
"grad_norm": 0.09329762309789658,
|
| 5202 |
+
"learning_rate": 5.164e-05,
|
| 5203 |
+
"loss": 0.021474449336528777,
|
| 5204 |
+
"step": 7420
|
| 5205 |
+
},
|
| 5206 |
+
{
|
| 5207 |
+
"epoch": 3.0306,
|
| 5208 |
+
"grad_norm": 0.09695407003164291,
|
| 5209 |
+
"learning_rate": 5.144e-05,
|
| 5210 |
+
"loss": 0.021487128734588624,
|
| 5211 |
+
"step": 7430
|
| 5212 |
+
},
|
| 5213 |
+
{
|
| 5214 |
+
"epoch": 4.0004,
|
| 5215 |
+
"grad_norm": 0.1085561066865921,
|
| 5216 |
+
"learning_rate": 5.124e-05,
|
| 5217 |
+
"loss": 0.024187105894088744,
|
| 5218 |
+
"step": 7440
|
| 5219 |
+
},
|
| 5220 |
+
{
|
| 5221 |
+
"epoch": 4.0014,
|
| 5222 |
+
"grad_norm": 0.09849539399147034,
|
| 5223 |
+
"learning_rate": 5.104e-05,
|
| 5224 |
+
"loss": 0.01914684921503067,
|
| 5225 |
+
"step": 7450
|
| 5226 |
+
},
|
| 5227 |
+
{
|
| 5228 |
+
"epoch": 4.0024,
|
| 5229 |
+
"grad_norm": 0.12541185319423676,
|
| 5230 |
+
"learning_rate": 5.084e-05,
|
| 5231 |
+
"loss": 0.020627421140670777,
|
| 5232 |
+
"step": 7460
|
| 5233 |
+
},
|
| 5234 |
+
{
|
| 5235 |
+
"epoch": 4.0034,
|
| 5236 |
+
"grad_norm": 0.08327507227659225,
|
| 5237 |
+
"learning_rate": 5.0639999999999996e-05,
|
| 5238 |
+
"loss": 0.019001516699790954,
|
| 5239 |
+
"step": 7470
|
| 5240 |
+
},
|
| 5241 |
+
{
|
| 5242 |
+
"epoch": 4.0044,
|
| 5243 |
+
"grad_norm": 0.10637800395488739,
|
| 5244 |
+
"learning_rate": 5.044e-05,
|
| 5245 |
+
"loss": 0.017808449268341065,
|
| 5246 |
+
"step": 7480
|
| 5247 |
+
},
|
| 5248 |
+
{
|
| 5249 |
+
"epoch": 4.0054,
|
| 5250 |
+
"grad_norm": 0.09127842634916306,
|
| 5251 |
+
"learning_rate": 5.024e-05,
|
| 5252 |
+
"loss": 0.019047291576862337,
|
| 5253 |
+
"step": 7490
|
| 5254 |
+
},
|
| 5255 |
+
{
|
| 5256 |
+
"epoch": 4.0064,
|
| 5257 |
+
"grad_norm": 0.0757402703166008,
|
| 5258 |
+
"learning_rate": 5.0039999999999995e-05,
|
| 5259 |
+
"loss": 0.018700633943080903,
|
| 5260 |
+
"step": 7500
|
| 5261 |
}
|
| 5262 |
],
|
| 5263 |
"logging_steps": 10,
|
|
|
|
| 5277 |
"attributes": {}
|
| 5278 |
}
|
| 5279 |
},
|
| 5280 |
+
"total_flos": 6.3485254012126e+17,
|
| 5281 |
"train_batch_size": 1,
|
| 5282 |
"trial_name": null,
|
| 5283 |
"trial_params": null
|