fedovtt commited on
Commit
c5807bc
·
verified ·
1 Parent(s): e57c62b

Training in progress, step 30, 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:dc7bc93b51275b767c587d004cf5142405ab304e9a7c2f04ba681f82d5dcea05
3
  size 132164608
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb237924aacdef593f49e4c283eae5f1e5aaa3754397f8e1cc1ea04745129d08
3
  size 132164608
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fd528811fbf994a3267de1203cb10f3b3c9354d7425525736af58315b312a096
3
  size 264524418
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5dde6db1c4bc46737a1e5b0d82d159e55bb51a2db3d375b87d34982c29dd2770
3
  size 264524418
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:56a36f74967cf339acf114f0c1fed7fad4606d4092d5f7eedc49a82cdd32cf69
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c761132c1dbe185c924c1ad9f5b50be95bfde45448e92853de9049c8ed176a9e
3
  size 14244
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a62e6ea25099651400ff4a3142a50e40bef5b52ba883be53b2fcb9d1a5b0a98c
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2def2cd24154d8cecbaa07c36ae27e5ebb9b7273a78abfea27aa67c480e4ae2b
3
  size 1064
last-checkpoint/trainer_state.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
- "best_metric": 1.2646969556808472,
3
- "best_model_checkpoint": "miner_id_24/checkpoint-20",
4
- "epoch": 0.0003772570582437741,
5
  "eval_steps": 5,
6
- "global_step": 20,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
@@ -89,6 +89,50 @@
89
  "eval_samples_per_second": 9.454,
90
  "eval_steps_per_second": 4.727,
91
  "step": 20
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
92
  }
93
  ],
94
  "logging_steps": 3,
@@ -112,12 +156,12 @@
112
  "should_evaluate": false,
113
  "should_log": false,
114
  "should_save": true,
115
- "should_training_stop": false
116
  },
117
  "attributes": {}
118
  }
119
  },
120
- "total_flos": 6355576246272000.0,
121
  "train_batch_size": 2,
122
  "trial_name": null,
123
  "trial_params": null
 
1
  {
2
+ "best_metric": 1.1986429691314697,
3
+ "best_model_checkpoint": "miner_id_24/checkpoint-30",
4
+ "epoch": 0.0005658855873656611,
5
  "eval_steps": 5,
6
+ "global_step": 30,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
 
89
  "eval_samples_per_second": 9.454,
90
  "eval_steps_per_second": 4.727,
91
  "step": 20
92
+ },
93
+ {
94
+ "epoch": 0.0003961199111559628,
95
+ "grad_norm": 3.294726610183716,
96
+ "learning_rate": 8.435655349597689e-05,
97
+ "loss": 4.8713,
98
+ "step": 21
99
+ },
100
+ {
101
+ "epoch": 0.0004527084698925289,
102
+ "grad_norm": 3.5907199382781982,
103
+ "learning_rate": 4.12214747707527e-05,
104
+ "loss": 5.0274,
105
+ "step": 24
106
+ },
107
+ {
108
+ "epoch": 0.0004715713228047176,
109
+ "eval_loss": 1.2100105285644531,
110
+ "eval_runtime": 2361.223,
111
+ "eval_samples_per_second": 9.454,
112
+ "eval_steps_per_second": 4.727,
113
+ "step": 25
114
+ },
115
+ {
116
+ "epoch": 0.000509297028629095,
117
+ "grad_norm": 3.3346734046936035,
118
+ "learning_rate": 1.0899347581163221e-05,
119
+ "loss": 5.081,
120
+ "step": 27
121
+ },
122
+ {
123
+ "epoch": 0.0005658855873656611,
124
+ "grad_norm": 3.953141927719116,
125
+ "learning_rate": 0.0,
126
+ "loss": 5.0118,
127
+ "step": 30
128
+ },
129
+ {
130
+ "epoch": 0.0005658855873656611,
131
+ "eval_loss": 1.1986429691314697,
132
+ "eval_runtime": 2361.282,
133
+ "eval_samples_per_second": 9.453,
134
+ "eval_steps_per_second": 4.727,
135
+ "step": 30
136
  }
137
  ],
138
  "logging_steps": 3,
 
156
  "should_evaluate": false,
157
  "should_log": false,
158
  "should_save": true,
159
+ "should_training_stop": true
160
  },
161
  "attributes": {}
162
  }
163
  },
164
+ "total_flos": 9533364369408000.0,
165
  "train_batch_size": 2,
166
  "trial_name": null,
167
  "trial_params": null