flytech commited on
Commit
b198169
·
1 Parent(s): 10d7f9d

Training in progress, step 600, checkpoint

Browse files
last-checkpoint/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:143eb999e252a44bfaec3f6dc65ccaadd527995ccf2e5917e89c6ae977c6b0e8
3
  size 250422888
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c65bcca65d19aa2ce91cf78e873e97774a0f133efbbd00e80df3d7e91e648ec
3
  size 250422888
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fc2af1b572ea70609763b0c346a5e5569c0bd0ad35d5d5abe812e24aabdff71a
3
  size 126034975
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73a61a4a50ca575121fe9a2305e202119e4c4af2e3e95758ba3fd117cc32e2dc
3
  size 126034975
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8c51ea51ba43ad9523a9def313e3c5a6e90b133318988b18624df83ba715cc8b
3
  size 14575
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac85f7c570c87d0ccd6b1529e54544af0f78bb6036760cfd999b7ec2a67fb410
3
  size 14575
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:13276f15dd2b6acc19b970176aa2db4ac9b58241843e72c89b50e3094e903b19
3
  size 627
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c1a2ac1c11599601eeac95feb1dbfd49ec5c625e61dcce18b3f094491f9cf2d
3
  size 627
last-checkpoint/trainer_state.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 1.3029315960912053,
5
  "eval_steps": 200,
6
- "global_step": 400,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
@@ -33,13 +33,26 @@
33
  "eval_samples_per_second": 4.006,
34
  "eval_steps_per_second": 0.132,
35
  "step": 400
 
 
 
 
 
 
 
 
 
 
 
 
 
36
  }
37
  ],
38
  "logging_steps": 200,
39
  "max_steps": 1535,
40
  "num_train_epochs": 5,
41
  "save_steps": 200,
42
- "total_flos": 5.0683133362176e+17,
43
  "trial_name": null,
44
  "trial_params": null
45
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 1.9543973941368078,
5
  "eval_steps": 200,
6
+ "global_step": 600,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
 
33
  "eval_samples_per_second": 4.006,
34
  "eval_steps_per_second": 0.132,
35
  "step": 400
36
+ },
37
+ {
38
+ "epoch": 1.95,
39
+ "learning_rate": 0.0002,
40
+ "loss": 0.132,
41
+ "step": 600
42
+ },
43
+ {
44
+ "epoch": 1.95,
45
+ "eval_runtime": 128.8118,
46
+ "eval_samples_per_second": 4.006,
47
+ "eval_steps_per_second": 0.132,
48
+ "step": 600
49
  }
50
  ],
51
  "logging_steps": 200,
52
  "max_steps": 1535,
53
  "num_train_epochs": 5,
54
  "save_steps": 200,
55
+ "total_flos": 7.6074292150272e+17,
56
  "trial_name": null,
57
  "trial_params": null
58
  }