kostiantynk commited on
Commit
358acbc
·
verified ·
1 Parent(s): 2a7648b

Training in progress, step 39, checkpoint

Browse files
last-checkpoint/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5694e85c4d1f12abb9b3136db72ce1d1ecd2a6be2d591f344d622e7acfa58a07
3
  size 111454040
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e5aa389ff7e85cf3d2b25e4e927d57cc31a2fb2129999c1e9aaa294d8445039
3
  size 111454040
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:718c07af6906c0dbc0e71a86a19b027167b1cb78234cba4d970fe26d3f53b096
3
  size 56950676
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a50eadb5444935deb31d49845d3de25892f84c85fb588ada80c84ba1a9d91b13
3
  size 56950676
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:70bfa699cedc5042dd9d0795988fedefecb9f8eb551ad3936666e9dde92cde92
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25ab8bb2fadb5ea724f330fd1bd2f36ff03259f68279e6767ad6577c056e063f
3
  size 14244
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:59af5f5d66ecb18e9abb84f5d5074a625502d490e59b29de1e9974286a967cbf
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:216e1a9a1c53e7de93f56319b4855910155cb98dd7ebb5fa08b3717fd303059f
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": 0.0014689680499449136,
5
  "eval_steps": 13,
6
- "global_step": 26,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
@@ -45,6 +45,21 @@
45
  "eval_samples_per_second": 9.591,
46
  "eval_steps_per_second": 4.796,
47
  "step": 26
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
48
  }
49
  ],
50
  "logging_steps": 10,
@@ -64,7 +79,7 @@
64
  "attributes": {}
65
  }
66
  },
67
- "total_flos": 1.57977582501888e+16,
68
  "train_batch_size": 2,
69
  "trial_name": null,
70
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 0.0022034520749173708,
5
  "eval_steps": 13,
6
+ "global_step": 39,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
 
45
  "eval_samples_per_second": 9.591,
46
  "eval_steps_per_second": 4.796,
47
  "step": 26
48
+ },
49
+ {
50
+ "epoch": 0.0016949631345518235,
51
+ "grad_norm": 2.3616435527801514,
52
+ "learning_rate": 8.263518223330697e-05,
53
+ "loss": 9.2361,
54
+ "step": 30
55
+ },
56
+ {
57
+ "epoch": 0.0022034520749173708,
58
+ "eval_loss": 2.33744215965271,
59
+ "eval_runtime": 777.3036,
60
+ "eval_samples_per_second": 9.588,
61
+ "eval_steps_per_second": 4.795,
62
+ "step": 39
63
  }
64
  ],
65
  "logging_steps": 10,
 
79
  "attributes": {}
80
  }
81
  },
82
+ "total_flos": 2.36966373752832e+16,
83
  "train_batch_size": 2,
84
  "trial_name": null,
85
  "trial_params": null