tarabukinivan commited on
Commit
0331169
·
verified ·
1 Parent(s): c30fde4

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:a2aed92cf51be26a47d2f05f92678d2accc93d0eed0bf127575c4dd1336bd054
3
  size 27024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63bf8bbd797eb426a4ef80ffeb9cb6eae50bcec93717147fd461f296b754eb86
3
  size 27024
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d286af48bb1cf23b8b4d660a4f5cddd0ba4dc82b179158bcfaf98836ff7bb77d
3
  size 70922
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c3ece9b9dd2dddafd3838c5bae7107b7274e060244fe9b89e6b4936cf71f3e5
3
  size 70922
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:22cdbc90fe5ab190ddc8c8d596470e7106be9ee2374df85af28061b6a32f91c4
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4542ee77b862586921f712605c9da4a157f7c6fb74687d0b6cef8d69e7a87363
3
  size 14244
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ef527702b4e2147a500c0b29beb26751d245ebb60338b3910b449b6549ac4d59
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": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 0.007052186177715092,
5
  "eval_steps": 5,
6
- "global_step": 15,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
@@ -74,6 +74,65 @@
74
  "eval_samples_per_second": 255.67,
75
  "eval_steps_per_second": 127.835,
76
  "step": 15
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
77
  }
78
  ],
79
  "logging_steps": 3,
@@ -88,12 +147,12 @@
88
  "should_evaluate": false,
89
  "should_log": false,
90
  "should_save": true,
91
- "should_training_stop": false
92
  },
93
  "attributes": {}
94
  }
95
  },
96
- "total_flos": 400858152960.0,
97
  "train_batch_size": 2,
98
  "trial_name": null,
99
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 0.014104372355430184,
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,
 
74
  "eval_samples_per_second": 255.67,
75
  "eval_steps_per_second": 127.835,
76
  "step": 15
77
+ },
78
+ {
79
+ "epoch": 0.00846262341325811,
80
+ "grad_norm": 0.03283863142132759,
81
+ "learning_rate": 0.00018090169943749476,
82
+ "loss": 10.3714,
83
+ "step": 18
84
+ },
85
+ {
86
+ "epoch": 0.009402914903620122,
87
+ "eval_loss": 10.370256423950195,
88
+ "eval_runtime": 3.566,
89
+ "eval_samples_per_second": 251.261,
90
+ "eval_steps_per_second": 125.631,
91
+ "step": 20
92
+ },
93
+ {
94
+ "epoch": 0.009873060648801129,
95
+ "grad_norm": 0.028590697795152664,
96
+ "learning_rate": 0.00013090169943749476,
97
+ "loss": 10.3723,
98
+ "step": 21
99
+ },
100
+ {
101
+ "epoch": 0.011283497884344146,
102
+ "grad_norm": 0.04093828424811363,
103
+ "learning_rate": 6.909830056250527e-05,
104
+ "loss": 10.3711,
105
+ "step": 24
106
+ },
107
+ {
108
+ "epoch": 0.011753643629525154,
109
+ "eval_loss": 10.369890213012695,
110
+ "eval_runtime": 3.5026,
111
+ "eval_samples_per_second": 255.812,
112
+ "eval_steps_per_second": 127.906,
113
+ "step": 25
114
+ },
115
+ {
116
+ "epoch": 0.012693935119887164,
117
+ "grad_norm": 0.043786004185676575,
118
+ "learning_rate": 1.9098300562505266e-05,
119
+ "loss": 10.3714,
120
+ "step": 27
121
+ },
122
+ {
123
+ "epoch": 0.014104372355430184,
124
+ "grad_norm": 0.04143758863210678,
125
+ "learning_rate": 0.0,
126
+ "loss": 10.3741,
127
+ "step": 30
128
+ },
129
+ {
130
+ "epoch": 0.014104372355430184,
131
+ "eval_loss": 10.369821548461914,
132
+ "eval_runtime": 3.4959,
133
+ "eval_samples_per_second": 256.298,
134
+ "eval_steps_per_second": 128.149,
135
+ "step": 30
136
  }
137
  ],
138
  "logging_steps": 3,
 
147
  "should_evaluate": false,
148
  "should_log": false,
149
  "should_save": true,
150
+ "should_training_stop": true
151
  },
152
  "attributes": {}
153
  }
154
  },
155
+ "total_flos": 795250851840.0,
156
  "train_batch_size": 2,
157
  "trial_name": null,
158
  "trial_params": null