Training in progress, step 372, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 80013120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:48a6f0994f83dc96cc6f751735e6fb97678ad4bc393d97a9dc093fd599ec6bc4
|
| 3 |
size 80013120
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 41120084
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b9a822bab0e997a745b26087d054e11d6c01f3452497bdb451ee15d4a3a6c0d7
|
| 3 |
size 41120084
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f79e2f3a046f199a98a51df125904aa982942d6fc81655bc2badc3d61093187d
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07e62fa0f3a9a79aa024e3df97b6370a7c2c27576fa400f2d11682e04348f4b3
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 93,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1984,6 +1984,665 @@
|
|
| 1984 |
"eval_samples_per_second": 14.63,
|
| 1985 |
"eval_steps_per_second": 7.362,
|
| 1986 |
"step": 279
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1987 |
}
|
| 1988 |
],
|
| 1989 |
"logging_steps": 1,
|
|
@@ -1998,12 +2657,12 @@
|
|
| 1998 |
"should_evaluate": false,
|
| 1999 |
"should_log": false,
|
| 2000 |
"should_save": true,
|
| 2001 |
-
"should_training_stop":
|
| 2002 |
},
|
| 2003 |
"attributes": {}
|
| 2004 |
}
|
| 2005 |
},
|
| 2006 |
-
"total_flos":
|
| 2007 |
"train_batch_size": 2,
|
| 2008 |
"trial_name": null,
|
| 2009 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0013449899125757,
|
| 5 |
"eval_steps": 93,
|
| 6 |
+
"global_step": 372,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1984 |
"eval_samples_per_second": 14.63,
|
| 1985 |
"eval_steps_per_second": 7.362,
|
| 1986 |
"step": 279
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 0.7531943510423672,
|
| 1990 |
+
"grad_norm": 3.4270105361938477,
|
| 1991 |
+
"learning_rate": 3.0215773883115706e-05,
|
| 1992 |
+
"loss": 0.5658,
|
| 1993 |
+
"step": 280
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.7558843308675185,
|
| 1997 |
+
"grad_norm": 3.4167211055755615,
|
| 1998 |
+
"learning_rate": 2.9596815548187908e-05,
|
| 1999 |
+
"loss": 0.1781,
|
| 2000 |
+
"step": 281
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 0.7585743106926698,
|
| 2004 |
+
"grad_norm": 3.9443657398223877,
|
| 2005 |
+
"learning_rate": 2.8983159609539635e-05,
|
| 2006 |
+
"loss": 0.5545,
|
| 2007 |
+
"step": 282
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.7612642905178211,
|
| 2011 |
+
"grad_norm": 3.164463758468628,
|
| 2012 |
+
"learning_rate": 2.8374852284497446e-05,
|
| 2013 |
+
"loss": 0.334,
|
| 2014 |
+
"step": 283
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 0.7639542703429725,
|
| 2018 |
+
"grad_norm": 3.6277055740356445,
|
| 2019 |
+
"learning_rate": 2.7771939387558554e-05,
|
| 2020 |
+
"loss": 0.411,
|
| 2021 |
+
"step": 284
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.7666442501681238,
|
| 2025 |
+
"grad_norm": 4.296345233917236,
|
| 2026 |
+
"learning_rate": 2.717446632694025e-05,
|
| 2027 |
+
"loss": 0.3111,
|
| 2028 |
+
"step": 285
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 0.769334229993275,
|
| 2032 |
+
"grad_norm": 4.06040096282959,
|
| 2033 |
+
"learning_rate": 2.6582478101160167e-05,
|
| 2034 |
+
"loss": 0.4634,
|
| 2035 |
+
"step": 286
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.7720242098184263,
|
| 2039 |
+
"grad_norm": 4.600436687469482,
|
| 2040 |
+
"learning_rate": 2.599601929564709e-05,
|
| 2041 |
+
"loss": 0.6998,
|
| 2042 |
+
"step": 287
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 0.7747141896435776,
|
| 2046 |
+
"grad_norm": 3.8486735820770264,
|
| 2047 |
+
"learning_rate": 2.5415134079383006e-05,
|
| 2048 |
+
"loss": 0.3987,
|
| 2049 |
+
"step": 288
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.777404169468729,
|
| 2053 |
+
"grad_norm": 5.362851142883301,
|
| 2054 |
+
"learning_rate": 2.4839866201576646e-05,
|
| 2055 |
+
"loss": 0.3466,
|
| 2056 |
+
"step": 289
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 0.7800941492938803,
|
| 2060 |
+
"grad_norm": 3.8688018321990967,
|
| 2061 |
+
"learning_rate": 2.4270258988368376e-05,
|
| 2062 |
+
"loss": 0.2902,
|
| 2063 |
+
"step": 290
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 0.7827841291190316,
|
| 2067 |
+
"grad_norm": 4.354773044586182,
|
| 2068 |
+
"learning_rate": 2.3706355339567286e-05,
|
| 2069 |
+
"loss": 0.4149,
|
| 2070 |
+
"step": 291
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 0.7854741089441829,
|
| 2074 |
+
"grad_norm": 7.11607027053833,
|
| 2075 |
+
"learning_rate": 2.3148197725419983e-05,
|
| 2076 |
+
"loss": 0.7291,
|
| 2077 |
+
"step": 292
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 0.7881640887693342,
|
| 2081 |
+
"grad_norm": 5.43526029586792,
|
| 2082 |
+
"learning_rate": 2.2595828183412172e-05,
|
| 2083 |
+
"loss": 0.2716,
|
| 2084 |
+
"step": 293
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.7908540685944856,
|
| 2088 |
+
"grad_norm": 3.004659414291382,
|
| 2089 |
+
"learning_rate": 2.2049288315102412e-05,
|
| 2090 |
+
"loss": 0.3067,
|
| 2091 |
+
"step": 294
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.7935440484196369,
|
| 2095 |
+
"grad_norm": 4.5855560302734375,
|
| 2096 |
+
"learning_rate": 2.1508619282989084e-05,
|
| 2097 |
+
"loss": 0.2618,
|
| 2098 |
+
"step": 295
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 0.7962340282447882,
|
| 2102 |
+
"grad_norm": 4.773977756500244,
|
| 2103 |
+
"learning_rate": 2.097386180741019e-05,
|
| 2104 |
+
"loss": 0.5023,
|
| 2105 |
+
"step": 296
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 0.7989240080699395,
|
| 2109 |
+
"grad_norm": 9.166229248046875,
|
| 2110 |
+
"learning_rate": 2.0445056163476374e-05,
|
| 2111 |
+
"loss": 0.4224,
|
| 2112 |
+
"step": 297
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 0.8016139878950908,
|
| 2116 |
+
"grad_norm": 6.276297092437744,
|
| 2117 |
+
"learning_rate": 1.9922242178037864e-05,
|
| 2118 |
+
"loss": 0.8068,
|
| 2119 |
+
"step": 298
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.8043039677202422,
|
| 2123 |
+
"grad_norm": 5.523612976074219,
|
| 2124 |
+
"learning_rate": 1.940545922668472e-05,
|
| 2125 |
+
"loss": 0.4406,
|
| 2126 |
+
"step": 299
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.8069939475453934,
|
| 2130 |
+
"grad_norm": 1.5128313302993774,
|
| 2131 |
+
"learning_rate": 1.88947462307814e-05,
|
| 2132 |
+
"loss": 0.0216,
|
| 2133 |
+
"step": 300
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.8096839273705447,
|
| 2137 |
+
"grad_norm": 2.8309073448181152,
|
| 2138 |
+
"learning_rate": 1.8390141654535265e-05,
|
| 2139 |
+
"loss": 1.299,
|
| 2140 |
+
"step": 301
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 0.812373907195696,
|
| 2144 |
+
"grad_norm": 3.6739649772644043,
|
| 2145 |
+
"learning_rate": 1.789168350209983e-05,
|
| 2146 |
+
"loss": 1.5798,
|
| 2147 |
+
"step": 302
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.8150638870208473,
|
| 2151 |
+
"grad_norm": 3.935307741165161,
|
| 2152 |
+
"learning_rate": 1.739940931471239e-05,
|
| 2153 |
+
"loss": 1.295,
|
| 2154 |
+
"step": 303
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.8177538668459986,
|
| 2158 |
+
"grad_norm": 4.4844865798950195,
|
| 2159 |
+
"learning_rate": 1.6913356167866578e-05,
|
| 2160 |
+
"loss": 1.225,
|
| 2161 |
+
"step": 304
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.82044384667115,
|
| 2165 |
+
"grad_norm": 4.518765449523926,
|
| 2166 |
+
"learning_rate": 1.6433560668520176e-05,
|
| 2167 |
+
"loss": 1.4111,
|
| 2168 |
+
"step": 305
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.8231338264963013,
|
| 2172 |
+
"grad_norm": 4.362013339996338,
|
| 2173 |
+
"learning_rate": 1.5960058952337887e-05,
|
| 2174 |
+
"loss": 1.1839,
|
| 2175 |
+
"step": 306
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.8258238063214526,
|
| 2179 |
+
"grad_norm": 4.76102352142334,
|
| 2180 |
+
"learning_rate": 1.5492886680969963e-05,
|
| 2181 |
+
"loss": 1.2118,
|
| 2182 |
+
"step": 307
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 0.8285137861466039,
|
| 2186 |
+
"grad_norm": 5.4755539894104,
|
| 2187 |
+
"learning_rate": 1.5032079039366209e-05,
|
| 2188 |
+
"loss": 1.4798,
|
| 2189 |
+
"step": 308
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 0.8312037659717552,
|
| 2193 |
+
"grad_norm": 3.792975902557373,
|
| 2194 |
+
"learning_rate": 1.4577670733126203e-05,
|
| 2195 |
+
"loss": 0.7013,
|
| 2196 |
+
"step": 309
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.8338937457969066,
|
| 2200 |
+
"grad_norm": 5.135954856872559,
|
| 2201 |
+
"learning_rate": 1.4129695985885228e-05,
|
| 2202 |
+
"loss": 1.5141,
|
| 2203 |
+
"step": 310
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.8365837256220578,
|
| 2207 |
+
"grad_norm": 3.417525291442871,
|
| 2208 |
+
"learning_rate": 1.3688188536736968e-05,
|
| 2209 |
+
"loss": 0.8687,
|
| 2210 |
+
"step": 311
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.8392737054472091,
|
| 2214 |
+
"grad_norm": 4.7601728439331055,
|
| 2215 |
+
"learning_rate": 1.3253181637692324e-05,
|
| 2216 |
+
"loss": 0.9127,
|
| 2217 |
+
"step": 312
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.8419636852723604,
|
| 2221 |
+
"grad_norm": 4.601919174194336,
|
| 2222 |
+
"learning_rate": 1.2824708051175016e-05,
|
| 2223 |
+
"loss": 1.0878,
|
| 2224 |
+
"step": 313
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 0.8446536650975117,
|
| 2228 |
+
"grad_norm": 3.320221185684204,
|
| 2229 |
+
"learning_rate": 1.2402800047554208e-05,
|
| 2230 |
+
"loss": 0.6061,
|
| 2231 |
+
"step": 314
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.8473436449226631,
|
| 2235 |
+
"grad_norm": 4.236156463623047,
|
| 2236 |
+
"learning_rate": 1.1987489402713981e-05,
|
| 2237 |
+
"loss": 0.7456,
|
| 2238 |
+
"step": 315
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 0.8500336247478144,
|
| 2242 |
+
"grad_norm": 6.007240295410156,
|
| 2243 |
+
"learning_rate": 1.1578807395660207e-05,
|
| 2244 |
+
"loss": 1.5298,
|
| 2245 |
+
"step": 316
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.8527236045729657,
|
| 2249 |
+
"grad_norm": 5.775532245635986,
|
| 2250 |
+
"learning_rate": 1.1176784806164676e-05,
|
| 2251 |
+
"loss": 0.7343,
|
| 2252 |
+
"step": 317
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 0.855413584398117,
|
| 2256 |
+
"grad_norm": 5.709627628326416,
|
| 2257 |
+
"learning_rate": 1.078145191244706e-05,
|
| 2258 |
+
"loss": 1.2876,
|
| 2259 |
+
"step": 318
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.8581035642232683,
|
| 2263 |
+
"grad_norm": 5.935501575469971,
|
| 2264 |
+
"learning_rate": 1.0392838488894463e-05,
|
| 2265 |
+
"loss": 0.9374,
|
| 2266 |
+
"step": 319
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 0.8607935440484197,
|
| 2270 |
+
"grad_norm": 4.249516010284424,
|
| 2271 |
+
"learning_rate": 1.0010973803818857e-05,
|
| 2272 |
+
"loss": 0.5061,
|
| 2273 |
+
"step": 320
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.863483523873571,
|
| 2277 |
+
"grad_norm": 4.154758453369141,
|
| 2278 |
+
"learning_rate": 9.635886617252975e-06,
|
| 2279 |
+
"loss": 0.1188,
|
| 2280 |
+
"step": 321
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 0.8661735036987223,
|
| 2284 |
+
"grad_norm": 3.874020576477051,
|
| 2285 |
+
"learning_rate": 9.267605178784033e-06,
|
| 2286 |
+
"loss": 0.4923,
|
| 2287 |
+
"step": 322
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.8688634835238735,
|
| 2291 |
+
"grad_norm": 3.575878143310547,
|
| 2292 |
+
"learning_rate": 8.906157225426315e-06,
|
| 2293 |
+
"loss": 0.3217,
|
| 2294 |
+
"step": 323
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 0.8715534633490248,
|
| 2298 |
+
"grad_norm": 4.050719261169434,
|
| 2299 |
+
"learning_rate": 8.55156997953197e-06,
|
| 2300 |
+
"loss": 0.4612,
|
| 2301 |
+
"step": 324
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 0.8742434431741762,
|
| 2305 |
+
"grad_norm": 3.588498830795288,
|
| 2306 |
+
"learning_rate": 8.203870146740932e-06,
|
| 2307 |
+
"loss": 0.2259,
|
| 2308 |
+
"step": 325
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 0.8769334229993275,
|
| 2312 |
+
"grad_norm": 5.262954235076904,
|
| 2313 |
+
"learning_rate": 7.86308391396956e-06,
|
| 2314 |
+
"loss": 0.7654,
|
| 2315 |
+
"step": 326
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.8796234028244788,
|
| 2319 |
+
"grad_norm": 5.5735087394714355,
|
| 2320 |
+
"learning_rate": 7.529236947438256e-06,
|
| 2321 |
+
"loss": 0.5849,
|
| 2322 |
+
"step": 327
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 0.8823133826496301,
|
| 2326 |
+
"grad_norm": 4.838580131530762,
|
| 2327 |
+
"learning_rate": 7.202354390738608e-06,
|
| 2328 |
+
"loss": 0.3913,
|
| 2329 |
+
"step": 328
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.8850033624747814,
|
| 2333 |
+
"grad_norm": 5.6935038566589355,
|
| 2334 |
+
"learning_rate": 6.882460862939522e-06,
|
| 2335 |
+
"loss": 0.7206,
|
| 2336 |
+
"step": 329
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.8876933422999328,
|
| 2340 |
+
"grad_norm": 2.3508174419403076,
|
| 2341 |
+
"learning_rate": 6.5695804567332044e-06,
|
| 2342 |
+
"loss": 0.1703,
|
| 2343 |
+
"step": 330
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 0.8903833221250841,
|
| 2347 |
+
"grad_norm": 5.699828624725342,
|
| 2348 |
+
"learning_rate": 6.263736736620551e-06,
|
| 2349 |
+
"loss": 0.4676,
|
| 2350 |
+
"step": 331
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.8930733019502354,
|
| 2354 |
+
"grad_norm": 4.048695087432861,
|
| 2355 |
+
"learning_rate": 5.964952737136353e-06,
|
| 2356 |
+
"loss": 0.5628,
|
| 2357 |
+
"step": 332
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.8957632817753867,
|
| 2361 |
+
"grad_norm": 4.811221599578857,
|
| 2362 |
+
"learning_rate": 5.673250961114529e-06,
|
| 2363 |
+
"loss": 0.7418,
|
| 2364 |
+
"step": 333
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.898453261600538,
|
| 2368 |
+
"grad_norm": 3.3414437770843506,
|
| 2369 |
+
"learning_rate": 5.388653377993324e-06,
|
| 2370 |
+
"loss": 0.3143,
|
| 2371 |
+
"step": 334
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.9011432414256894,
|
| 2375 |
+
"grad_norm": 5.924250602722168,
|
| 2376 |
+
"learning_rate": 5.111181422160671e-06,
|
| 2377 |
+
"loss": 0.5284,
|
| 2378 |
+
"step": 335
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.9038332212508406,
|
| 2382 |
+
"grad_norm": 6.767046928405762,
|
| 2383 |
+
"learning_rate": 4.840855991339799e-06,
|
| 2384 |
+
"loss": 0.6351,
|
| 2385 |
+
"step": 336
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.9065232010759919,
|
| 2389 |
+
"grad_norm": 4.555798053741455,
|
| 2390 |
+
"learning_rate": 4.577697445015472e-06,
|
| 2391 |
+
"loss": 0.5253,
|
| 2392 |
+
"step": 337
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.9092131809011432,
|
| 2396 |
+
"grad_norm": 5.7803730964660645,
|
| 2397 |
+
"learning_rate": 4.321725602900473e-06,
|
| 2398 |
+
"loss": 0.7582,
|
| 2399 |
+
"step": 338
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.9119031607262945,
|
| 2403 |
+
"grad_norm": 4.016640663146973,
|
| 2404 |
+
"learning_rate": 4.072959743443017e-06,
|
| 2405 |
+
"loss": 0.2845,
|
| 2406 |
+
"step": 339
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.9145931405514459,
|
| 2410 |
+
"grad_norm": 5.46890926361084,
|
| 2411 |
+
"learning_rate": 3.83141860237467e-06,
|
| 2412 |
+
"loss": 0.6128,
|
| 2413 |
+
"step": 340
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.9172831203765972,
|
| 2417 |
+
"grad_norm": 4.543710708618164,
|
| 2418 |
+
"learning_rate": 3.5971203712993894e-06,
|
| 2419 |
+
"loss": 0.5227,
|
| 2420 |
+
"step": 341
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.9199731002017485,
|
| 2424 |
+
"grad_norm": 4.0189008712768555,
|
| 2425 |
+
"learning_rate": 3.3700826963233735e-06,
|
| 2426 |
+
"loss": 0.4072,
|
| 2427 |
+
"step": 342
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.9226630800268998,
|
| 2431 |
+
"grad_norm": 5.0270490646362305,
|
| 2432 |
+
"learning_rate": 3.1503226767260252e-06,
|
| 2433 |
+
"loss": 0.5361,
|
| 2434 |
+
"step": 343
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.9253530598520511,
|
| 2438 |
+
"grad_norm": 7.237580299377441,
|
| 2439 |
+
"learning_rate": 2.9378568636721835e-06,
|
| 2440 |
+
"loss": 0.9466,
|
| 2441 |
+
"step": 344
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.9280430396772025,
|
| 2445 |
+
"grad_norm": 8.795455932617188,
|
| 2446 |
+
"learning_rate": 2.732701258965531e-06,
|
| 2447 |
+
"loss": 0.6604,
|
| 2448 |
+
"step": 345
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.9307330195023538,
|
| 2452 |
+
"grad_norm": 11.6528959274292,
|
| 2453 |
+
"learning_rate": 2.5348713138434564e-06,
|
| 2454 |
+
"loss": 0.5807,
|
| 2455 |
+
"step": 346
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.933422999327505,
|
| 2459 |
+
"grad_norm": 8.07696533203125,
|
| 2460 |
+
"learning_rate": 2.3443819278132996e-06,
|
| 2461 |
+
"loss": 0.7975,
|
| 2462 |
+
"step": 347
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.9361129791526563,
|
| 2466 |
+
"grad_norm": 4.788589954376221,
|
| 2467 |
+
"learning_rate": 2.161247447530268e-06,
|
| 2468 |
+
"loss": 0.6227,
|
| 2469 |
+
"step": 348
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 0.9388029589778076,
|
| 2473 |
+
"grad_norm": 7.453376293182373,
|
| 2474 |
+
"learning_rate": 1.985481665716882e-06,
|
| 2475 |
+
"loss": 0.4651,
|
| 2476 |
+
"step": 349
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.9414929388029589,
|
| 2480 |
+
"grad_norm": 4.3519392013549805,
|
| 2481 |
+
"learning_rate": 1.8170978201241474e-06,
|
| 2482 |
+
"loss": 0.1668,
|
| 2483 |
+
"step": 350
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.9441829186281103,
|
| 2487 |
+
"grad_norm": 3.087855577468872,
|
| 2488 |
+
"learning_rate": 1.6561085925346332e-06,
|
| 2489 |
+
"loss": 1.2559,
|
| 2490 |
+
"step": 351
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 0.9468728984532616,
|
| 2494 |
+
"grad_norm": 3.9484481811523438,
|
| 2495 |
+
"learning_rate": 1.5025261078073005e-06,
|
| 2496 |
+
"loss": 1.0505,
|
| 2497 |
+
"step": 352
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.9495628782784129,
|
| 2501 |
+
"grad_norm": 4.509681701660156,
|
| 2502 |
+
"learning_rate": 1.3563619329643119e-06,
|
| 2503 |
+
"loss": 1.316,
|
| 2504 |
+
"step": 353
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 0.9522528581035642,
|
| 2508 |
+
"grad_norm": 4.409306049346924,
|
| 2509 |
+
"learning_rate": 1.2176270763198828e-06,
|
| 2510 |
+
"loss": 0.9114,
|
| 2511 |
+
"step": 354
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 0.9549428379287155,
|
| 2515 |
+
"grad_norm": 5.652538299560547,
|
| 2516 |
+
"learning_rate": 1.0863319866512346e-06,
|
| 2517 |
+
"loss": 1.1458,
|
| 2518 |
+
"step": 355
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 0.9576328177538669,
|
| 2522 |
+
"grad_norm": 6.170865535736084,
|
| 2523 |
+
"learning_rate": 9.624865524115346e-07,
|
| 2524 |
+
"loss": 1.1232,
|
| 2525 |
+
"step": 356
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.9603227975790182,
|
| 2529 |
+
"grad_norm": 5.357152938842773,
|
| 2530 |
+
"learning_rate": 8.461001009852809e-07,
|
| 2531 |
+
"loss": 0.9592,
|
| 2532 |
+
"step": 357
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.9630127774041695,
|
| 2536 |
+
"grad_norm": 4.322149753570557,
|
| 2537 |
+
"learning_rate": 7.371813979857312e-07,
|
| 2538 |
+
"loss": 0.7773,
|
| 2539 |
+
"step": 358
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.9657027572293208,
|
| 2543 |
+
"grad_norm": 3.6123275756835938,
|
| 2544 |
+
"learning_rate": 6.357386465947301e-07,
|
| 2545 |
+
"loss": 0.5652,
|
| 2546 |
+
"step": 359
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 0.968392737054472,
|
| 2550 |
+
"grad_norm": 3.7311031818389893,
|
| 2551 |
+
"learning_rate": 5.417794869449377e-07,
|
| 2552 |
+
"loss": 0.6096,
|
| 2553 |
+
"step": 360
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.9710827168796234,
|
| 2557 |
+
"grad_norm": 5.762843608856201,
|
| 2558 |
+
"learning_rate": 4.5531099554435576e-07,
|
| 2559 |
+
"loss": 0.9279,
|
| 2560 |
+
"step": 361
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 0.9737726967047747,
|
| 2564 |
+
"grad_norm": 4.97388219833374,
|
| 2565 |
+
"learning_rate": 3.763396847433875e-07,
|
| 2566 |
+
"loss": 0.5789,
|
| 2567 |
+
"step": 362
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 0.976462676529926,
|
| 2571 |
+
"grad_norm": 4.815624713897705,
|
| 2572 |
+
"learning_rate": 3.048715022443749e-07,
|
| 2573 |
+
"loss": 0.5138,
|
| 2574 |
+
"step": 363
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.9791526563550773,
|
| 2578 |
+
"grad_norm": 3.541781425476074,
|
| 2579 |
+
"learning_rate": 2.409118306536229e-07,
|
| 2580 |
+
"loss": 0.259,
|
| 2581 |
+
"step": 364
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.9818426361802286,
|
| 2585 |
+
"grad_norm": 2.7444493770599365,
|
| 2586 |
+
"learning_rate": 1.8446548707604648e-07,
|
| 2587 |
+
"loss": 0.2707,
|
| 2588 |
+
"step": 365
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.98453261600538,
|
| 2592 |
+
"grad_norm": 5.796267986297607,
|
| 2593 |
+
"learning_rate": 1.3553672275230523e-07,
|
| 2594 |
+
"loss": 0.5347,
|
| 2595 |
+
"step": 366
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 0.9872225958305313,
|
| 2599 |
+
"grad_norm": 5.090404987335205,
|
| 2600 |
+
"learning_rate": 9.412922273871471e-08,
|
| 2601 |
+
"loss": 0.3201,
|
| 2602 |
+
"step": 367
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 0.9899125756556826,
|
| 2606 |
+
"grad_norm": 4.630456924438477,
|
| 2607 |
+
"learning_rate": 6.024610562962441e-08,
|
| 2608 |
+
"loss": 0.4391,
|
| 2609 |
+
"step": 368
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.9926025554808339,
|
| 2613 |
+
"grad_norm": 4.325840473175049,
|
| 2614 |
+
"learning_rate": 3.388992332259422e-08,
|
| 2615 |
+
"loss": 0.3675,
|
| 2616 |
+
"step": 369
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 0.9952925353059852,
|
| 2620 |
+
"grad_norm": 9.686969757080078,
|
| 2621 |
+
"learning_rate": 1.506266082615948e-08,
|
| 2622 |
+
"loss": 0.6909,
|
| 2623 |
+
"step": 370
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.9979825151311366,
|
| 2627 |
+
"grad_norm": 4.668429851531982,
|
| 2628 |
+
"learning_rate": 3.7657361103837776e-09,
|
| 2629 |
+
"loss": 0.285,
|
| 2630 |
+
"step": 371
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 1.0013449899125757,
|
| 2634 |
+
"grad_norm": 4.755204200744629,
|
| 2635 |
+
"learning_rate": 0.0,
|
| 2636 |
+
"loss": 0.9972,
|
| 2637 |
+
"step": 372
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 1.0013449899125757,
|
| 2641 |
+
"eval_loss": 0.7320420145988464,
|
| 2642 |
+
"eval_runtime": 10.7106,
|
| 2643 |
+
"eval_samples_per_second": 14.658,
|
| 2644 |
+
"eval_steps_per_second": 7.376,
|
| 2645 |
+
"step": 372
|
| 2646 |
}
|
| 2647 |
],
|
| 2648 |
"logging_steps": 1,
|
|
|
|
| 2657 |
"should_evaluate": false,
|
| 2658 |
"should_log": false,
|
| 2659 |
"should_save": true,
|
| 2660 |
+
"should_training_stop": true
|
| 2661 |
},
|
| 2662 |
"attributes": {}
|
| 2663 |
}
|
| 2664 |
},
|
| 2665 |
+
"total_flos": 1.211789436077998e+17,
|
| 2666 |
"train_batch_size": 2,
|
| 2667 |
"trial_name": null,
|
| 2668 |
"trial_params": null
|