Upload 5 files
Browse files- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +81 -0
- training_args.bin +3 -0
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:35b1e2346458f1817007ae59db6a742ece72f561e7d61ff3720c0d2265faed34
|
3 |
+
size 4279546
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f99ad65665196c3aa2bce538e25605284518542acc7ac200cf264f95182c9022
|
3 |
+
size 14244
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bc30d29ba4ed588e3a7e9f1ada97ed956dd6e3c9675ccf96d7feaa7c04fa34e9
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 2.5536394119262695,
|
3 |
+
"best_model_checkpoint": "shawgpt-ft/checkpoint-13",
|
4 |
+
"epoch": 4.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 13,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.9230769230769231,
|
13 |
+
"grad_norm": 2.1487667560577393,
|
14 |
+
"learning_rate": 0.00019285714285714286,
|
15 |
+
"loss": 4.5923,
|
16 |
+
"step": 3
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.9230769230769231,
|
20 |
+
"eval_loss": 3.957515239715576,
|
21 |
+
"eval_runtime": 5.3861,
|
22 |
+
"eval_samples_per_second": 1.671,
|
23 |
+
"eval_steps_per_second": 0.557,
|
24 |
+
"step": 3
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 1.8461538461538463,
|
28 |
+
"grad_norm": 2.26593279838562,
|
29 |
+
"learning_rate": 0.00017142857142857143,
|
30 |
+
"loss": 4.0416,
|
31 |
+
"step": 6
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 1.8461538461538463,
|
35 |
+
"eval_loss": 3.429060697555542,
|
36 |
+
"eval_runtime": 5.4348,
|
37 |
+
"eval_samples_per_second": 1.656,
|
38 |
+
"eval_steps_per_second": 0.552,
|
39 |
+
"step": 6
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 2.769230769230769,
|
43 |
+
"grad_norm": 2.352947950363159,
|
44 |
+
"learning_rate": 0.00015000000000000001,
|
45 |
+
"loss": 3.4652,
|
46 |
+
"step": 9
|
47 |
+
},
|
48 |
+
{
|
49 |
+
"epoch": 2.769230769230769,
|
50 |
+
"eval_loss": 2.976196527481079,
|
51 |
+
"eval_runtime": 5.4009,
|
52 |
+
"eval_samples_per_second": 1.666,
|
53 |
+
"eval_steps_per_second": 0.555,
|
54 |
+
"step": 9
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 4.0,
|
58 |
+
"grad_norm": 2.3249692916870117,
|
59 |
+
"learning_rate": 0.00012142857142857143,
|
60 |
+
"loss": 2.2554,
|
61 |
+
"step": 13
|
62 |
+
},
|
63 |
+
{
|
64 |
+
"epoch": 4.0,
|
65 |
+
"eval_loss": 2.5536394119262695,
|
66 |
+
"eval_runtime": 5.4046,
|
67 |
+
"eval_samples_per_second": 1.665,
|
68 |
+
"eval_steps_per_second": 0.555,
|
69 |
+
"step": 13
|
70 |
+
}
|
71 |
+
],
|
72 |
+
"logging_steps": 500,
|
73 |
+
"max_steps": 30,
|
74 |
+
"num_input_tokens_seen": 0,
|
75 |
+
"num_train_epochs": 10,
|
76 |
+
"save_steps": 500,
|
77 |
+
"total_flos": 43039318818816.0,
|
78 |
+
"train_batch_size": 4,
|
79 |
+
"trial_name": null,
|
80 |
+
"trial_params": null
|
81 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:425c63576bac4c598045325cb635f0a50b610376cc1b475e29bce8e98f3bd25a
|
3 |
+
size 4984
|