sergioalves commited on
Commit
ee8bd65
·
verified ·
1 Parent(s): 71d6070

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:fdb5a9bb053473bf8e0de29d64b1e848e7c2d962c5f1869366410f7d9a8f517b
3
  size 598799664
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a13f77483aaa7eb8786805948eb206455539bf61753c3db41d54a993aed768e4
3
  size 598799664
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:21cf3e15967105794d24f1c6581ee61f20e9cb0aacd5f4e20bcae76bf847c3b0
3
  size 167923706
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21a155b61c0250b440a413982cfd5d6216b74dc17720d20601a0585bb1fe073f
3
  size 167923706
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9b0ae598748e69b3cd8972746107b8fb2745f98e4fd90cec1f97a4c6c8e80b94
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17d91bdc66f6fc21e9cb5413cf7fb9f0918a3f0ca03eaa1050cfc0b5e49411fc
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": 1.1996384859085083,
3
- "best_model_checkpoint": "miner_id_24/checkpoint-20",
4
- "epoch": 0.0006941431670281996,
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": 10.991,
90
  "eval_steps_per_second": 5.496,
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": 4064263085752320.0,
121
  "train_batch_size": 2,
122
  "trial_name": null,
123
  "trial_params": null
 
1
  {
2
+ "best_metric": 1.1903597116470337,
3
+ "best_model_checkpoint": "miner_id_24/checkpoint-30",
4
+ "epoch": 0.0010412147505422993,
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": 10.991,
90
  "eval_steps_per_second": 5.496,
91
  "step": 20
92
+ },
93
+ {
94
+ "epoch": 0.0007288503253796095,
95
+ "grad_norm": 2.0232975482940674,
96
+ "learning_rate": 4.2178276747988446e-05,
97
+ "loss": 4.4981,
98
+ "step": 21
99
+ },
100
+ {
101
+ "epoch": 0.0008329718004338395,
102
+ "grad_norm": 2.182401418685913,
103
+ "learning_rate": 2.061073738537635e-05,
104
+ "loss": 4.6256,
105
+ "step": 24
106
+ },
107
+ {
108
+ "epoch": 0.0008676789587852494,
109
+ "eval_loss": 1.1918320655822754,
110
+ "eval_runtime": 1104.7539,
111
+ "eval_samples_per_second": 10.982,
112
+ "eval_steps_per_second": 5.491,
113
+ "step": 25
114
+ },
115
+ {
116
+ "epoch": 0.0009370932754880694,
117
+ "grad_norm": 2.29049015045166,
118
+ "learning_rate": 5.449673790581611e-06,
119
+ "loss": 4.8211,
120
+ "step": 27
121
+ },
122
+ {
123
+ "epoch": 0.0010412147505422993,
124
+ "grad_norm": 4.216933727264404,
125
+ "learning_rate": 0.0,
126
+ "loss": 5.218,
127
+ "step": 30
128
+ },
129
+ {
130
+ "epoch": 0.0010412147505422993,
131
+ "eval_loss": 1.1903597116470337,
132
+ "eval_runtime": 1104.119,
133
+ "eval_samples_per_second": 10.988,
134
+ "eval_steps_per_second": 5.494,
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": 6063618313420800.0,
165
  "train_batch_size": 2,
166
  "trial_name": null,
167
  "trial_params": null