duyphu commited on
Commit
f890b05
·
verified ·
1 Parent(s): a63c2f5

Training in progress, step 13, checkpoint

Browse files
last-checkpoint/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:98a95a34701c77e74bcfef6d2f10168d10eb8e4a39250ed9d49e92e61342262a
3
  size 65309632
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e30fd6ac9c7fa36ae3b36c23290bcb86600809e3cc64ff5795fb718c8933825
3
  size 65309632
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a959bf1a6d5b46894fa2370b44338d421e6be18330162c390340b1f366adda1a
3
  size 33424212
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:beeda25dbaab161d3347f8880b4307c2b71d8544b6d0bdf9f8e2dfbea62bd1c3
3
  size 33424212
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:eb1228e8df97a1d799b51d461e89faa3c8b34af3eaad77d8523d93156c1e565e
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4cc289ca89ed37bf50c2f838c7574993fbc613d6d7d723cf73cfe31d185c0abf
3
  size 14244
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ae751897b8e87ff08962a91d1d3485984775a96aa89e29a1caac3d6f449228f7
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4395bafb1cf8adc0d9dbfe550d1b98763496e37a8fadb76c1ca1168a4a757f4e
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": 0.00036088054853843375,
5
- "eval_steps": 1,
6
- "global_step": 1,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
@@ -11,17 +11,39 @@
11
  {
12
  "epoch": 0.00036088054853843375,
13
  "eval_loss": 2.960909366607666,
14
- "eval_runtime": 131.0676,
15
- "eval_samples_per_second": 8.904,
16
- "eval_steps_per_second": 4.456,
17
  "step": 1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
18
  }
19
  ],
20
  "logging_steps": 5,
21
- "max_steps": 1,
22
  "num_input_tokens_seen": 0,
23
  "num_train_epochs": 1,
24
- "save_steps": 1,
25
  "stateful_callbacks": {
26
  "TrainerControl": {
27
  "args": {
@@ -29,12 +51,12 @@
29
  "should_evaluate": false,
30
  "should_log": false,
31
  "should_save": true,
32
- "should_training_stop": true
33
  },
34
  "attributes": {}
35
  }
36
  },
37
- "total_flos": 326495606145024.0,
38
  "train_batch_size": 2,
39
  "trial_name": null,
40
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 0.004691447130999639,
5
+ "eval_steps": 10,
6
+ "global_step": 13,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
 
11
  {
12
  "epoch": 0.00036088054853843375,
13
  "eval_loss": 2.960909366607666,
14
+ "eval_runtime": 132.2763,
15
+ "eval_samples_per_second": 8.822,
16
+ "eval_steps_per_second": 4.415,
17
  "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.001804402742692169,
21
+ "grad_norm": 3.78476619720459,
22
+ "learning_rate": 5e-05,
23
+ "loss": 11.7201,
24
+ "step": 5
25
+ },
26
+ {
27
+ "epoch": 0.003608805485384338,
28
+ "grad_norm": 5.096208095550537,
29
+ "learning_rate": 0.0001,
30
+ "loss": 11.3326,
31
+ "step": 10
32
+ },
33
+ {
34
+ "epoch": 0.003608805485384338,
35
+ "eval_loss": 2.4844770431518555,
36
+ "eval_runtime": 132.2042,
37
+ "eval_samples_per_second": 8.827,
38
+ "eval_steps_per_second": 4.417,
39
+ "step": 10
40
  }
41
  ],
42
  "logging_steps": 5,
43
+ "max_steps": 50,
44
  "num_input_tokens_seen": 0,
45
  "num_train_epochs": 1,
46
+ "save_steps": 13,
47
  "stateful_callbacks": {
48
  "TrainerControl": {
49
  "args": {
 
51
  "should_evaluate": false,
52
  "should_log": false,
53
  "should_save": true,
54
+ "should_training_stop": false
55
  },
56
  "attributes": {}
57
  }
58
  },
59
+ "total_flos": 4244442879885312.0,
60
  "train_batch_size": 2,
61
  "trial_name": null,
62
  "trial_params": null
last-checkpoint/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a491a211f3df2eb16aa3f27eadc15f0e3653bb3eb5ffc59cc0e23976b514a424
3
  size 6776
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e44b392d2a58d1447c758fa231bd1ef14907d6cc6a85ebc1ea9c74bddc70c480
3
  size 6776