sergioalves commited on
Commit
c546d86
·
verified ·
1 Parent(s): 09e8a13

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 sha256:8c87f29e60c5e09cd5df4311dee83f99b94bea618b9355301adc10a4eb39dcc9
3
  size 250422888
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a06289978f7a553fa0245f2343142e44d309554970b1f607423c994d5a724f85
3
  size 250422888
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:852988a45cbd28b2a5c0415ccd0a86dcde134d50d9c802edaae5072023bfad73
3
  size 501028338
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab9e9314379ca7b1e135a7028220f8c619fbb376d1998618e8fe54a3ea2d9f92
3
  size 501028338
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c7332cedcc7b3b36b7974e070bd9fa7171e6a96e191919b27b6014d872048e1e
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83fdd6955092c161405108c24aff329154c04f5feb7c98b00794794c0262bcda
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": 2.66746187210083,
3
- "best_model_checkpoint": "miner_id_24/checkpoint-20",
4
- "epoch": 0.003270913402567667,
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": 5.041,
90
  "eval_steps_per_second": 2.522,
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": 1.2695579394048e+16,
121
  "train_batch_size": 2,
122
  "trial_name": null,
123
  "trial_params": null
 
1
  {
2
+ "best_metric": 2.6111176013946533,
3
+ "best_model_checkpoint": "miner_id_24/checkpoint-30",
4
+ "epoch": 0.0049063701038515005,
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": 5.041,
90
  "eval_steps_per_second": 2.522,
91
  "step": 20
92
+ },
93
+ {
94
+ "epoch": 0.0034344590726960505,
95
+ "grad_norm": 1.4085278511047363,
96
+ "learning_rate": 4.2178276747988446e-05,
97
+ "loss": 2.2877,
98
+ "step": 21
99
+ },
100
+ {
101
+ "epoch": 0.0039250960830812,
102
+ "grad_norm": 1.4254778623580933,
103
+ "learning_rate": 2.061073738537635e-05,
104
+ "loss": 2.4096,
105
+ "step": 24
106
+ },
107
+ {
108
+ "epoch": 0.004088641753209584,
109
+ "eval_loss": 2.6312296390533447,
110
+ "eval_runtime": 510.5738,
111
+ "eval_samples_per_second": 5.043,
112
+ "eval_steps_per_second": 2.523,
113
+ "step": 25
114
+ },
115
+ {
116
+ "epoch": 0.004415733093466351,
117
+ "grad_norm": 1.1322522163391113,
118
+ "learning_rate": 5.449673790581611e-06,
119
+ "loss": 2.4184,
120
+ "step": 27
121
+ },
122
+ {
123
+ "epoch": 0.0049063701038515005,
124
+ "grad_norm": 2.1908793449401855,
125
+ "learning_rate": 0.0,
126
+ "loss": 2.9997,
127
+ "step": 30
128
+ },
129
+ {
130
+ "epoch": 0.0049063701038515005,
131
+ "eval_loss": 2.6111176013946533,
132
+ "eval_runtime": 510.3088,
133
+ "eval_samples_per_second": 5.046,
134
+ "eval_steps_per_second": 2.524,
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": 1.9043369091072e+16,
165
  "train_batch_size": 2,
166
  "trial_name": null,
167
  "trial_params": null