cvoffer commited on
Commit
d9d619b
·
verified ·
1 Parent(s): ad33536

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 sha256:8e0fcf4bbf1f9f1970a46c89e2a233572a7ba3e29e3a7a3b624008c93730d734
3
  size 101752088
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:283ccdb1e3fd7ff85ae843d34de2dbb727ecf3cfa3b9fc786c66703e8162ca42
3
  size 101752088
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a43367a4318965846711a981dde352d79975ad377cf0920ddb76f167a4a286f1
3
  size 203713238
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65d351a31fc65b75bda7a5d542179857c1074f01bdecc8e89b1e9cbc6cada9d3
3
  size 203713238
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c2cf22bc849b00644528515dfcd0ca607f9eb83dcba27f583e44fe76c71bd4dd
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b51a95a270690d66680cc127d27ddddb92ee7e685d5c451d8b23bcb09c97488
3
  size 14244
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:46d68217ddf3e899bf1495ee33b9ec14e2c91ed912ffcb460a7bb01192b04b7d
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0e9cc72c20ddd925ef39b6005e82a4d8730b1dde32cfcd070d74c83a8a3564a
3
  size 1064
last-checkpoint/trainer_state.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
- "best_metric": 1.9001489877700806,
3
- "best_model_checkpoint": "miner_id_24/checkpoint-20",
4
- "epoch": 0.001828278903946797,
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": 3.147,
90
  "eval_steps_per_second": 1.574,
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": 6132638865162240.0,
121
  "train_batch_size": 2,
122
  "trial_name": null,
123
  "trial_params": null
 
1
  {
2
+ "best_metric": 1.86117422580719,
3
+ "best_model_checkpoint": "miner_id_24/checkpoint-30",
4
+ "epoch": 0.0027424183559201956,
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": 3.147,
90
  "eval_steps_per_second": 1.574,
91
  "step": 20
92
+ },
93
+ {
94
+ "epoch": 0.001919692849144137,
95
+ "grad_norm": 0.5045371651649475,
96
+ "learning_rate": 4.2178276747988446e-05,
97
+ "loss": 1.9161,
98
+ "step": 21
99
+ },
100
+ {
101
+ "epoch": 0.0021939346847361565,
102
+ "grad_norm": 0.44254088401794434,
103
+ "learning_rate": 2.061073738537635e-05,
104
+ "loss": 1.9082,
105
+ "step": 24
106
+ },
107
+ {
108
+ "epoch": 0.0022853486299334964,
109
+ "eval_loss": 1.8675042390823364,
110
+ "eval_runtime": 1464.9858,
111
+ "eval_samples_per_second": 3.145,
112
+ "eval_steps_per_second": 1.573,
113
+ "step": 25
114
+ },
115
+ {
116
+ "epoch": 0.0024681765203281763,
117
+ "grad_norm": 0.4314767122268677,
118
+ "learning_rate": 5.449673790581611e-06,
119
+ "loss": 1.8267,
120
+ "step": 27
121
+ },
122
+ {
123
+ "epoch": 0.0027424183559201956,
124
+ "grad_norm": 0.4271681606769562,
125
+ "learning_rate": 0.0,
126
+ "loss": 1.8588,
127
+ "step": 30
128
+ },
129
+ {
130
+ "epoch": 0.0027424183559201956,
131
+ "eval_loss": 1.86117422580719,
132
+ "eval_runtime": 1465.0617,
133
+ "eval_samples_per_second": 3.145,
134
+ "eval_steps_per_second": 1.573,
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": 9178378972692480.0,
165
  "train_batch_size": 2,
166
  "trial_name": null,
167
  "trial_params": null