Training in progress, step 18000, checkpoint
Browse files
checkpoint-18000/config.json
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"annotation_transformer": false,
|
3 |
+
"architectures": [
|
4 |
+
"CAMP"
|
5 |
+
],
|
6 |
+
"diff": false,
|
7 |
+
"dropout": 0.05,
|
8 |
+
"hidden_dim": 640,
|
9 |
+
"input_dim": 384,
|
10 |
+
"intermediate_dim": 2560,
|
11 |
+
"kernel_size": 11,
|
12 |
+
"latent": false,
|
13 |
+
"mlm": false,
|
14 |
+
"mnr": false,
|
15 |
+
"model_type": "CAMP",
|
16 |
+
"nhead": 8,
|
17 |
+
"nlp_path": "allenai/scibert_scivocab_uncased",
|
18 |
+
"num_hidden_layers": 1,
|
19 |
+
"out_dim": 512,
|
20 |
+
"plm_path": "facebook/esm2_t33_650m_UR50D",
|
21 |
+
"pooling": "avg",
|
22 |
+
"token": null,
|
23 |
+
"torch_dtype": "bfloat16",
|
24 |
+
"transformers_version": "4.41.2"
|
25 |
+
}
|
checkpoint-18000/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6096d3a9659c0e65ddfdadbc8ac004fb76d30286752a2abba5fc44e2158596a9
|
3 |
+
size 1570427506
|
checkpoint-18000/trainer_state.json
ADDED
@@ -0,0 +1,1491 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.403171616718183,
|
5 |
+
"eval_steps": 1000,
|
6 |
+
"global_step": 18000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.002239842315101017,
|
13 |
+
"grad_norm": 71.10852813720703,
|
14 |
+
"learning_rate": 0.0001,
|
15 |
+
"loss": 3.5057,
|
16 |
+
"step": 100
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.004479684630202034,
|
20 |
+
"grad_norm": 57.756534576416016,
|
21 |
+
"learning_rate": 9.9998756572327e-05,
|
22 |
+
"loss": 0.8254,
|
23 |
+
"step": 200
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.006719526945303051,
|
27 |
+
"grad_norm": 44.61080551147461,
|
28 |
+
"learning_rate": 9.999502635115246e-05,
|
29 |
+
"loss": 0.6935,
|
30 |
+
"step": 300
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.008959369260404068,
|
34 |
+
"grad_norm": 50.82619094848633,
|
35 |
+
"learning_rate": 9.998880952200681e-05,
|
36 |
+
"loss": 0.5972,
|
37 |
+
"step": 400
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.011199211575505085,
|
41 |
+
"grad_norm": 44.26677703857422,
|
42 |
+
"learning_rate": 9.998010639409713e-05,
|
43 |
+
"loss": 0.5408,
|
44 |
+
"step": 500
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.013439053890606102,
|
48 |
+
"grad_norm": 42.8045654296875,
|
49 |
+
"learning_rate": 9.996891740029186e-05,
|
50 |
+
"loss": 0.5509,
|
51 |
+
"step": 600
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.01567889620570712,
|
55 |
+
"grad_norm": 30.989139556884766,
|
56 |
+
"learning_rate": 9.995524309709913e-05,
|
57 |
+
"loss": 0.4823,
|
58 |
+
"step": 700
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.017918738520808136,
|
62 |
+
"grad_norm": 34.01952362060547,
|
63 |
+
"learning_rate": 9.993908416463927e-05,
|
64 |
+
"loss": 0.5111,
|
65 |
+
"step": 800
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.020158580835909153,
|
69 |
+
"grad_norm": 34.079307556152344,
|
70 |
+
"learning_rate": 9.992044140661079e-05,
|
71 |
+
"loss": 0.4635,
|
72 |
+
"step": 900
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.02239842315101017,
|
76 |
+
"grad_norm": 26.16071128845215,
|
77 |
+
"learning_rate": 9.989931575025056e-05,
|
78 |
+
"loss": 0.4883,
|
79 |
+
"step": 1000
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.02239842315101017,
|
83 |
+
"eval_avg_non_pair_similarity": 0.0020343252948339737,
|
84 |
+
"eval_avg_pair_similarity": 0.008852629057131708,
|
85 |
+
"eval_loss": 0.5431402921676636,
|
86 |
+
"eval_runtime": 19.6178,
|
87 |
+
"eval_samples_per_second": 25.487,
|
88 |
+
"eval_similarity_ratio": 4.35162905343228,
|
89 |
+
"eval_steps_per_second": 0.816,
|
90 |
+
"step": 1000
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"epoch": 0.024638265466111187,
|
94 |
+
"grad_norm": 35.38695526123047,
|
95 |
+
"learning_rate": 9.987570824628759e-05,
|
96 |
+
"loss": 0.4655,
|
97 |
+
"step": 1100
|
98 |
+
},
|
99 |
+
{
|
100 |
+
"epoch": 0.026878107781212204,
|
101 |
+
"grad_norm": 32.071346282958984,
|
102 |
+
"learning_rate": 9.984962006889084e-05,
|
103 |
+
"loss": 0.4342,
|
104 |
+
"step": 1200
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 0.029117950096313218,
|
108 |
+
"grad_norm": 39.6610221862793,
|
109 |
+
"learning_rate": 9.982105251561082e-05,
|
110 |
+
"loss": 0.458,
|
111 |
+
"step": 1300
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 0.03135779241141424,
|
115 |
+
"grad_norm": 31.493322372436523,
|
116 |
+
"learning_rate": 9.979000700731491e-05,
|
117 |
+
"loss": 0.4525,
|
118 |
+
"step": 1400
|
119 |
+
},
|
120 |
+
{
|
121 |
+
"epoch": 0.03359763472651525,
|
122 |
+
"grad_norm": 34.453399658203125,
|
123 |
+
"learning_rate": 9.975648508811693e-05,
|
124 |
+
"loss": 0.41,
|
125 |
+
"step": 1500
|
126 |
+
},
|
127 |
+
{
|
128 |
+
"epoch": 0.03583747704161627,
|
129 |
+
"grad_norm": 33.990074157714844,
|
130 |
+
"learning_rate": 9.972048842530012e-05,
|
131 |
+
"loss": 0.4097,
|
132 |
+
"step": 1600
|
133 |
+
},
|
134 |
+
{
|
135 |
+
"epoch": 0.038077319356717286,
|
136 |
+
"grad_norm": 30.44228172302246,
|
137 |
+
"learning_rate": 9.968201880923439e-05,
|
138 |
+
"loss": 0.4257,
|
139 |
+
"step": 1700
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"epoch": 0.040317161671818307,
|
143 |
+
"grad_norm": 31.427162170410156,
|
144 |
+
"learning_rate": 9.964107815328711e-05,
|
145 |
+
"loss": 0.3821,
|
146 |
+
"step": 1800
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 0.04255700398691932,
|
150 |
+
"grad_norm": 25.877887725830078,
|
151 |
+
"learning_rate": 9.959766849372808e-05,
|
152 |
+
"loss": 0.3788,
|
153 |
+
"step": 1900
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 0.04479684630202034,
|
157 |
+
"grad_norm": 25.36798095703125,
|
158 |
+
"learning_rate": 9.955179198962817e-05,
|
159 |
+
"loss": 0.3854,
|
160 |
+
"step": 2000
|
161 |
+
},
|
162 |
+
{
|
163 |
+
"epoch": 0.04479684630202034,
|
164 |
+
"eval_avg_non_pair_similarity": 0.0016289287904792565,
|
165 |
+
"eval_avg_pair_similarity": 0.0032495629731565715,
|
166 |
+
"eval_loss": 0.4394480586051941,
|
167 |
+
"eval_runtime": 19.5064,
|
168 |
+
"eval_samples_per_second": 25.633,
|
169 |
+
"eval_similarity_ratio": 1.9949079371360972,
|
170 |
+
"eval_steps_per_second": 0.82,
|
171 |
+
"step": 2000
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.047036688617121354,
|
175 |
+
"grad_norm": 24.537384033203125,
|
176 |
+
"learning_rate": 9.950345092275198e-05,
|
177 |
+
"loss": 0.3886,
|
178 |
+
"step": 2100
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.049276530932222375,
|
182 |
+
"grad_norm": 26.4981632232666,
|
183 |
+
"learning_rate": 9.945264769744431e-05,
|
184 |
+
"loss": 0.3926,
|
185 |
+
"step": 2200
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.05151637324732339,
|
189 |
+
"grad_norm": 40.585941314697266,
|
190 |
+
"learning_rate": 9.939938484051063e-05,
|
191 |
+
"loss": 0.3796,
|
192 |
+
"step": 2300
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.05375621556242441,
|
196 |
+
"grad_norm": 26.374149322509766,
|
197 |
+
"learning_rate": 9.934366500109132e-05,
|
198 |
+
"loss": 0.3723,
|
199 |
+
"step": 2400
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.05599605787752542,
|
203 |
+
"grad_norm": 31.54728126525879,
|
204 |
+
"learning_rate": 9.928549095053001e-05,
|
205 |
+
"loss": 0.3639,
|
206 |
+
"step": 2500
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.058235900192626436,
|
210 |
+
"grad_norm": 26.00455093383789,
|
211 |
+
"learning_rate": 9.922486558223567e-05,
|
212 |
+
"loss": 0.3387,
|
213 |
+
"step": 2600
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.060475742507727456,
|
217 |
+
"grad_norm": 23.684057235717773,
|
218 |
+
"learning_rate": 9.916179191153873e-05,
|
219 |
+
"loss": 0.3494,
|
220 |
+
"step": 2700
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.06271558482282848,
|
224 |
+
"grad_norm": 24.034656524658203,
|
225 |
+
"learning_rate": 9.909627307554108e-05,
|
226 |
+
"loss": 0.362,
|
227 |
+
"step": 2800
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.06495542713792948,
|
231 |
+
"grad_norm": 31.07891082763672,
|
232 |
+
"learning_rate": 9.902831233296009e-05,
|
233 |
+
"loss": 0.3481,
|
234 |
+
"step": 2900
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.0671952694530305,
|
238 |
+
"grad_norm": 24.183626174926758,
|
239 |
+
"learning_rate": 9.895791306396644e-05,
|
240 |
+
"loss": 0.3535,
|
241 |
+
"step": 3000
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.0671952694530305,
|
245 |
+
"eval_avg_non_pair_similarity": -0.0007878901720614375,
|
246 |
+
"eval_avg_pair_similarity": -0.008789425778668375,
|
247 |
+
"eval_loss": 0.4187028110027313,
|
248 |
+
"eval_runtime": 19.4839,
|
249 |
+
"eval_samples_per_second": 25.662,
|
250 |
+
"eval_similarity_ratio": 11.155648452463497,
|
251 |
+
"eval_steps_per_second": 0.821,
|
252 |
+
"step": 3000
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 0.06943511176813152,
|
256 |
+
"grad_norm": 20.178213119506836,
|
257 |
+
"learning_rate": 9.888507877001616e-05,
|
258 |
+
"loss": 0.3375,
|
259 |
+
"step": 3100
|
260 |
+
},
|
261 |
+
{
|
262 |
+
"epoch": 0.07167495408323254,
|
263 |
+
"grad_norm": 31.757244110107422,
|
264 |
+
"learning_rate": 9.880981307367627e-05,
|
265 |
+
"loss": 0.3489,
|
266 |
+
"step": 3200
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 0.07391479639833355,
|
270 |
+
"grad_norm": 27.889982223510742,
|
271 |
+
"learning_rate": 9.873211971844477e-05,
|
272 |
+
"loss": 0.3463,
|
273 |
+
"step": 3300
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"epoch": 0.07615463871343457,
|
277 |
+
"grad_norm": 23.282093048095703,
|
278 |
+
"learning_rate": 9.865200256856437e-05,
|
279 |
+
"loss": 0.3235,
|
280 |
+
"step": 3400
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"epoch": 0.07839448102853559,
|
284 |
+
"grad_norm": 24.573863983154297,
|
285 |
+
"learning_rate": 9.856946560883034e-05,
|
286 |
+
"loss": 0.3474,
|
287 |
+
"step": 3500
|
288 |
+
},
|
289 |
+
{
|
290 |
+
"epoch": 0.08063432334363661,
|
291 |
+
"grad_norm": 20.7421932220459,
|
292 |
+
"learning_rate": 9.848451294439224e-05,
|
293 |
+
"loss": 0.3533,
|
294 |
+
"step": 3600
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"epoch": 0.08287416565873762,
|
298 |
+
"grad_norm": 22.051513671875,
|
299 |
+
"learning_rate": 9.839714880054987e-05,
|
300 |
+
"loss": 0.3277,
|
301 |
+
"step": 3700
|
302 |
+
},
|
303 |
+
{
|
304 |
+
"epoch": 0.08511400797383864,
|
305 |
+
"grad_norm": 22.470027923583984,
|
306 |
+
"learning_rate": 9.830737752254294e-05,
|
307 |
+
"loss": 0.3261,
|
308 |
+
"step": 3800
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 0.08735385028893966,
|
312 |
+
"grad_norm": 20.80890464782715,
|
313 |
+
"learning_rate": 9.821520357533513e-05,
|
314 |
+
"loss": 0.3474,
|
315 |
+
"step": 3900
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 0.08959369260404068,
|
319 |
+
"grad_norm": 20.187280654907227,
|
320 |
+
"learning_rate": 9.812063154339191e-05,
|
321 |
+
"loss": 0.2984,
|
322 |
+
"step": 4000
|
323 |
+
},
|
324 |
+
{
|
325 |
+
"epoch": 0.08959369260404068,
|
326 |
+
"eval_avg_non_pair_similarity": -0.0008396646829899625,
|
327 |
+
"eval_avg_pair_similarity": -0.011315496074035763,
|
328 |
+
"eval_loss": 0.32538020610809326,
|
329 |
+
"eval_runtime": 19.4916,
|
330 |
+
"eval_samples_per_second": 25.652,
|
331 |
+
"eval_similarity_ratio": 13.476208185560939,
|
332 |
+
"eval_steps_per_second": 0.821,
|
333 |
+
"step": 4000
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 0.09183353491914169,
|
337 |
+
"grad_norm": 26.496036529541016,
|
338 |
+
"learning_rate": 9.802366613045254e-05,
|
339 |
+
"loss": 0.3326,
|
340 |
+
"step": 4100
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 0.09407337723424271,
|
344 |
+
"grad_norm": 23.102359771728516,
|
345 |
+
"learning_rate": 9.792431215929613e-05,
|
346 |
+
"loss": 0.3341,
|
347 |
+
"step": 4200
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 0.09631321954934373,
|
351 |
+
"grad_norm": 21.27369499206543,
|
352 |
+
"learning_rate": 9.782257457150177e-05,
|
353 |
+
"loss": 0.3254,
|
354 |
+
"step": 4300
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 0.09855306186444475,
|
358 |
+
"grad_norm": 19.25406265258789,
|
359 |
+
"learning_rate": 9.771845842720274e-05,
|
360 |
+
"loss": 0.3334,
|
361 |
+
"step": 4400
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 0.10079290417954576,
|
365 |
+
"grad_norm": 15.666335105895996,
|
366 |
+
"learning_rate": 9.761196890483482e-05,
|
367 |
+
"loss": 0.3064,
|
368 |
+
"step": 4500
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 0.10303274649464678,
|
372 |
+
"grad_norm": 24.592592239379883,
|
373 |
+
"learning_rate": 9.75031113008788e-05,
|
374 |
+
"loss": 0.2902,
|
375 |
+
"step": 4600
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 0.1052725888097478,
|
379 |
+
"grad_norm": 20.0572566986084,
|
380 |
+
"learning_rate": 9.739189102959696e-05,
|
381 |
+
"loss": 0.3121,
|
382 |
+
"step": 4700
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 0.10751243112484882,
|
386 |
+
"grad_norm": 19.887725830078125,
|
387 |
+
"learning_rate": 9.727831362276381e-05,
|
388 |
+
"loss": 0.3014,
|
389 |
+
"step": 4800
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 0.10975227343994982,
|
393 |
+
"grad_norm": 15.561097145080566,
|
394 |
+
"learning_rate": 9.716238472939101e-05,
|
395 |
+
"loss": 0.316,
|
396 |
+
"step": 4900
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 0.11199211575505084,
|
400 |
+
"grad_norm": 17.96786880493164,
|
401 |
+
"learning_rate": 9.704411011544629e-05,
|
402 |
+
"loss": 0.3215,
|
403 |
+
"step": 5000
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 0.11199211575505084,
|
407 |
+
"eval_avg_non_pair_similarity": 0.00024146916974524413,
|
408 |
+
"eval_avg_pair_similarity": -0.009633154251612723,
|
409 |
+
"eval_loss": 0.2951599061489105,
|
410 |
+
"eval_runtime": 19.5025,
|
411 |
+
"eval_samples_per_second": 25.638,
|
412 |
+
"eval_similarity_ratio": -39.893930400207765,
|
413 |
+
"eval_steps_per_second": 0.82,
|
414 |
+
"step": 5000
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"epoch": 0.11423195807015186,
|
418 |
+
"grad_norm": 14.672723770141602,
|
419 |
+
"learning_rate": 9.692349566356677e-05,
|
420 |
+
"loss": 0.2954,
|
421 |
+
"step": 5100
|
422 |
+
},
|
423 |
+
{
|
424 |
+
"epoch": 0.11647180038525287,
|
425 |
+
"grad_norm": 17.881025314331055,
|
426 |
+
"learning_rate": 9.680054737276638e-05,
|
427 |
+
"loss": 0.2968,
|
428 |
+
"step": 5200
|
429 |
+
},
|
430 |
+
{
|
431 |
+
"epoch": 0.11871164270035389,
|
432 |
+
"grad_norm": 16.55130958557129,
|
433 |
+
"learning_rate": 9.667527135813737e-05,
|
434 |
+
"loss": 0.2707,
|
435 |
+
"step": 5300
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 0.12095148501545491,
|
439 |
+
"grad_norm": 21.650575637817383,
|
440 |
+
"learning_rate": 9.654767385054627e-05,
|
441 |
+
"loss": 0.3068,
|
442 |
+
"step": 5400
|
443 |
+
},
|
444 |
+
{
|
445 |
+
"epoch": 0.12319132733055593,
|
446 |
+
"grad_norm": 20.990921020507812,
|
447 |
+
"learning_rate": 9.641776119632397e-05,
|
448 |
+
"loss": 0.3037,
|
449 |
+
"step": 5500
|
450 |
+
},
|
451 |
+
{
|
452 |
+
"epoch": 0.12543116964565695,
|
453 |
+
"grad_norm": 21.97833824157715,
|
454 |
+
"learning_rate": 9.628553985695005e-05,
|
455 |
+
"loss": 0.3307,
|
456 |
+
"step": 5600
|
457 |
+
},
|
458 |
+
{
|
459 |
+
"epoch": 0.12767101196075797,
|
460 |
+
"grad_norm": 20.353246688842773,
|
461 |
+
"learning_rate": 9.61510164087314e-05,
|
462 |
+
"loss": 0.2907,
|
463 |
+
"step": 5700
|
464 |
+
},
|
465 |
+
{
|
466 |
+
"epoch": 0.12991085427585897,
|
467 |
+
"grad_norm": 16.254976272583008,
|
468 |
+
"learning_rate": 9.601419754247514e-05,
|
469 |
+
"loss": 0.3025,
|
470 |
+
"step": 5800
|
471 |
+
},
|
472 |
+
{
|
473 |
+
"epoch": 0.13215069659096,
|
474 |
+
"grad_norm": 24.14662742614746,
|
475 |
+
"learning_rate": 9.587509006315585e-05,
|
476 |
+
"loss": 0.2676,
|
477 |
+
"step": 5900
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 0.134390538906061,
|
481 |
+
"grad_norm": 15.861074447631836,
|
482 |
+
"learning_rate": 9.573370088957712e-05,
|
483 |
+
"loss": 0.2953,
|
484 |
+
"step": 6000
|
485 |
+
},
|
486 |
+
{
|
487 |
+
"epoch": 0.134390538906061,
|
488 |
+
"eval_avg_non_pair_similarity": -0.00021786548842126337,
|
489 |
+
"eval_avg_pair_similarity": -0.001965825233142823,
|
490 |
+
"eval_loss": 0.31484127044677734,
|
491 |
+
"eval_runtime": 19.5134,
|
492 |
+
"eval_samples_per_second": 25.623,
|
493 |
+
"eval_similarity_ratio": 9.023114433533939,
|
494 |
+
"eval_steps_per_second": 0.82,
|
495 |
+
"step": 6000
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 0.13663038122116203,
|
499 |
+
"grad_norm": 14.910375595092773,
|
500 |
+
"learning_rate": 9.559003705402737e-05,
|
501 |
+
"loss": 0.2846,
|
502 |
+
"step": 6100
|
503 |
+
},
|
504 |
+
{
|
505 |
+
"epoch": 0.13887022353626305,
|
506 |
+
"grad_norm": 17.768138885498047,
|
507 |
+
"learning_rate": 9.544410570193014e-05,
|
508 |
+
"loss": 0.3031,
|
509 |
+
"step": 6200
|
510 |
+
},
|
511 |
+
{
|
512 |
+
"epoch": 0.14111006585136407,
|
513 |
+
"grad_norm": 13.620220184326172,
|
514 |
+
"learning_rate": 9.529591409148874e-05,
|
515 |
+
"loss": 0.2716,
|
516 |
+
"step": 6300
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"epoch": 0.1433499081664651,
|
520 |
+
"grad_norm": 14.697577476501465,
|
521 |
+
"learning_rate": 9.514546959332509e-05,
|
522 |
+
"loss": 0.2759,
|
523 |
+
"step": 6400
|
524 |
+
},
|
525 |
+
{
|
526 |
+
"epoch": 0.1455897504815661,
|
527 |
+
"grad_norm": 21.234310150146484,
|
528 |
+
"learning_rate": 9.499277969011334e-05,
|
529 |
+
"loss": 0.2845,
|
530 |
+
"step": 6500
|
531 |
+
},
|
532 |
+
{
|
533 |
+
"epoch": 0.1478295927966671,
|
534 |
+
"grad_norm": 16.50038719177246,
|
535 |
+
"learning_rate": 9.483785197620747e-05,
|
536 |
+
"loss": 0.2471,
|
537 |
+
"step": 6600
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 0.15006943511176812,
|
541 |
+
"grad_norm": 16.261537551879883,
|
542 |
+
"learning_rate": 9.468069415726377e-05,
|
543 |
+
"loss": 0.2978,
|
544 |
+
"step": 6700
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"epoch": 0.15230927742686914,
|
548 |
+
"grad_norm": 16.553571701049805,
|
549 |
+
"learning_rate": 9.452131404985752e-05,
|
550 |
+
"loss": 0.311,
|
551 |
+
"step": 6800
|
552 |
+
},
|
553 |
+
{
|
554 |
+
"epoch": 0.15454911974197016,
|
555 |
+
"grad_norm": 18.445993423461914,
|
556 |
+
"learning_rate": 9.43597195810941e-05,
|
557 |
+
"loss": 0.2777,
|
558 |
+
"step": 6900
|
559 |
+
},
|
560 |
+
{
|
561 |
+
"epoch": 0.15678896205707119,
|
562 |
+
"grad_norm": 13.675585746765137,
|
563 |
+
"learning_rate": 9.419591878821496e-05,
|
564 |
+
"loss": 0.2417,
|
565 |
+
"step": 7000
|
566 |
+
},
|
567 |
+
{
|
568 |
+
"epoch": 0.15678896205707119,
|
569 |
+
"eval_avg_non_pair_similarity": 0.0003613333898595999,
|
570 |
+
"eval_avg_pair_similarity": 0.005676110625499859,
|
571 |
+
"eval_loss": 0.3019685447216034,
|
572 |
+
"eval_runtime": 19.5219,
|
573 |
+
"eval_samples_per_second": 25.612,
|
574 |
+
"eval_similarity_ratio": 15.708790786551377,
|
575 |
+
"eval_steps_per_second": 0.82,
|
576 |
+
"step": 7000
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.1590288043721722,
|
580 |
+
"grad_norm": 15.780075073242188,
|
581 |
+
"learning_rate": 9.402991981819758e-05,
|
582 |
+
"loss": 0.2866,
|
583 |
+
"step": 7100
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.16126864668727323,
|
587 |
+
"grad_norm": 17.34490966796875,
|
588 |
+
"learning_rate": 9.386173092735051e-05,
|
589 |
+
"loss": 0.2526,
|
590 |
+
"step": 7200
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.16350848900237425,
|
594 |
+
"grad_norm": 14.431266784667969,
|
595 |
+
"learning_rate": 9.36913604809026e-05,
|
596 |
+
"loss": 0.2646,
|
597 |
+
"step": 7300
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.16574833131747524,
|
601 |
+
"grad_norm": 11.20712661743164,
|
602 |
+
"learning_rate": 9.351881695258693e-05,
|
603 |
+
"loss": 0.2279,
|
604 |
+
"step": 7400
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.16798817363257626,
|
608 |
+
"grad_norm": 15.549891471862793,
|
609 |
+
"learning_rate": 9.334410892421945e-05,
|
610 |
+
"loss": 0.2884,
|
611 |
+
"step": 7500
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.17022801594767728,
|
615 |
+
"grad_norm": 14.635927200317383,
|
616 |
+
"learning_rate": 9.316724508527205e-05,
|
617 |
+
"loss": 0.2729,
|
618 |
+
"step": 7600
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.1724678582627783,
|
622 |
+
"grad_norm": 12.841429710388184,
|
623 |
+
"learning_rate": 9.298823423244038e-05,
|
624 |
+
"loss": 0.271,
|
625 |
+
"step": 7700
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.17470770057787932,
|
629 |
+
"grad_norm": 13.053755760192871,
|
630 |
+
"learning_rate": 9.280708526920636e-05,
|
631 |
+
"loss": 0.3094,
|
632 |
+
"step": 7800
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.17694754289298034,
|
636 |
+
"grad_norm": 14.694463729858398,
|
637 |
+
"learning_rate": 9.262380720539536e-05,
|
638 |
+
"loss": 0.2687,
|
639 |
+
"step": 7900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.17918738520808136,
|
643 |
+
"grad_norm": 14.9842529296875,
|
644 |
+
"learning_rate": 9.243840915672804e-05,
|
645 |
+
"loss": 0.2539,
|
646 |
+
"step": 8000
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.17918738520808136,
|
650 |
+
"eval_avg_non_pair_similarity": -0.0009227607365296818,
|
651 |
+
"eval_avg_pair_similarity": -0.005276160409208387,
|
652 |
+
"eval_loss": 0.22878731787204742,
|
653 |
+
"eval_runtime": 19.5192,
|
654 |
+
"eval_samples_per_second": 25.616,
|
655 |
+
"eval_similarity_ratio": 5.717798992023619,
|
656 |
+
"eval_steps_per_second": 0.82,
|
657 |
+
"step": 8000
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 0.18142722752318236,
|
661 |
+
"grad_norm": 11.896175384521484,
|
662 |
+
"learning_rate": 9.225090034436697e-05,
|
663 |
+
"loss": 0.2739,
|
664 |
+
"step": 8100
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 0.18366706983828338,
|
668 |
+
"grad_norm": 12.229938507080078,
|
669 |
+
"learning_rate": 9.206129009445796e-05,
|
670 |
+
"loss": 0.2467,
|
671 |
+
"step": 8200
|
672 |
+
},
|
673 |
+
{
|
674 |
+
"epoch": 0.1859069121533844,
|
675 |
+
"grad_norm": 15.017464637756348,
|
676 |
+
"learning_rate": 9.186958783766633e-05,
|
677 |
+
"loss": 0.2668,
|
678 |
+
"step": 8300
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 0.18814675446848542,
|
682 |
+
"grad_norm": 15.509882926940918,
|
683 |
+
"learning_rate": 9.167580310870769e-05,
|
684 |
+
"loss": 0.2614,
|
685 |
+
"step": 8400
|
686 |
+
},
|
687 |
+
{
|
688 |
+
"epoch": 0.19038659678358644,
|
689 |
+
"grad_norm": 14.945472717285156,
|
690 |
+
"learning_rate": 9.147994554587385e-05,
|
691 |
+
"loss": 0.2266,
|
692 |
+
"step": 8500
|
693 |
+
},
|
694 |
+
{
|
695 |
+
"epoch": 0.19262643909868746,
|
696 |
+
"grad_norm": 16.0211238861084,
|
697 |
+
"learning_rate": 9.128202489055335e-05,
|
698 |
+
"loss": 0.2428,
|
699 |
+
"step": 8600
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 0.19486628141378848,
|
703 |
+
"grad_norm": 16.380727767944336,
|
704 |
+
"learning_rate": 9.108205098674698e-05,
|
705 |
+
"loss": 0.2578,
|
706 |
+
"step": 8700
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"epoch": 0.1971061237288895,
|
710 |
+
"grad_norm": 12.938665390014648,
|
711 |
+
"learning_rate": 9.088003378057816e-05,
|
712 |
+
"loss": 0.2445,
|
713 |
+
"step": 8800
|
714 |
+
},
|
715 |
+
{
|
716 |
+
"epoch": 0.1993459660439905,
|
717 |
+
"grad_norm": 7.832601070404053,
|
718 |
+
"learning_rate": 9.067598331979829e-05,
|
719 |
+
"loss": 0.2284,
|
720 |
+
"step": 8900
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 0.2015858083590915,
|
724 |
+
"grad_norm": 12.243515014648438,
|
725 |
+
"learning_rate": 9.046990975328694e-05,
|
726 |
+
"loss": 0.271,
|
727 |
+
"step": 9000
|
728 |
+
},
|
729 |
+
{
|
730 |
+
"epoch": 0.2015858083590915,
|
731 |
+
"eval_avg_non_pair_similarity": -0.0005964781935742025,
|
732 |
+
"eval_avg_pair_similarity": -0.0012636589542962611,
|
733 |
+
"eval_loss": 0.24378302693367004,
|
734 |
+
"eval_runtime": 19.5294,
|
735 |
+
"eval_samples_per_second": 25.602,
|
736 |
+
"eval_similarity_ratio": 2.1185333645211637,
|
737 |
+
"eval_steps_per_second": 0.819,
|
738 |
+
"step": 9000
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 0.20382565067419253,
|
742 |
+
"grad_norm": 14.057201385498047,
|
743 |
+
"learning_rate": 9.026182333054714e-05,
|
744 |
+
"loss": 0.2216,
|
745 |
+
"step": 9100
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 0.20606549298929355,
|
749 |
+
"grad_norm": 9.724617958068848,
|
750 |
+
"learning_rate": 9.005173440119555e-05,
|
751 |
+
"loss": 0.2625,
|
752 |
+
"step": 9200
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 0.20830533530439457,
|
756 |
+
"grad_norm": 13.209918975830078,
|
757 |
+
"learning_rate": 8.983965341444769e-05,
|
758 |
+
"loss": 0.2508,
|
759 |
+
"step": 9300
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.2105451776194956,
|
763 |
+
"grad_norm": 12.0099458694458,
|
764 |
+
"learning_rate": 8.962559091859828e-05,
|
765 |
+
"loss": 0.2263,
|
766 |
+
"step": 9400
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 0.21278501993459661,
|
770 |
+
"grad_norm": 11.473740577697754,
|
771 |
+
"learning_rate": 8.940955756049658e-05,
|
772 |
+
"loss": 0.256,
|
773 |
+
"step": 9500
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 0.21502486224969763,
|
777 |
+
"grad_norm": 13.63650131225586,
|
778 |
+
"learning_rate": 8.919156408501678e-05,
|
779 |
+
"loss": 0.2594,
|
780 |
+
"step": 9600
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 0.21726470456479863,
|
784 |
+
"grad_norm": 11.578638076782227,
|
785 |
+
"learning_rate": 8.897162133452375e-05,
|
786 |
+
"loss": 0.225,
|
787 |
+
"step": 9700
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 0.21950454687989965,
|
791 |
+
"grad_norm": 8.985605239868164,
|
792 |
+
"learning_rate": 8.874974024833351e-05,
|
793 |
+
"loss": 0.2525,
|
794 |
+
"step": 9800
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 0.22174438919500067,
|
798 |
+
"grad_norm": 15.842543601989746,
|
799 |
+
"learning_rate": 8.852593186216942e-05,
|
800 |
+
"loss": 0.239,
|
801 |
+
"step": 9900
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 0.2239842315101017,
|
805 |
+
"grad_norm": 13.797926902770996,
|
806 |
+
"learning_rate": 8.830020730761308e-05,
|
807 |
+
"loss": 0.2133,
|
808 |
+
"step": 10000
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 0.2239842315101017,
|
812 |
+
"eval_avg_non_pair_similarity": 3.824399728542623e-05,
|
813 |
+
"eval_avg_pair_similarity": -0.010851175129879266,
|
814 |
+
"eval_loss": 0.2623503506183624,
|
815 |
+
"eval_runtime": 19.4973,
|
816 |
+
"eval_samples_per_second": 25.645,
|
817 |
+
"eval_similarity_ratio": -283.7353807159264,
|
818 |
+
"eval_steps_per_second": 0.821,
|
819 |
+
"step": 10000
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 0.2262240738252027,
|
823 |
+
"grad_norm": 13.315906524658203,
|
824 |
+
"learning_rate": 8.807257781155081e-05,
|
825 |
+
"loss": 0.2192,
|
826 |
+
"step": 10100
|
827 |
+
},
|
828 |
+
{
|
829 |
+
"epoch": 0.22846391614030373,
|
830 |
+
"grad_norm": 11.403688430786133,
|
831 |
+
"learning_rate": 8.784305469561519e-05,
|
832 |
+
"loss": 0.2359,
|
833 |
+
"step": 10200
|
834 |
+
},
|
835 |
+
{
|
836 |
+
"epoch": 0.23070375845540475,
|
837 |
+
"grad_norm": 11.676390647888184,
|
838 |
+
"learning_rate": 8.761164937562199e-05,
|
839 |
+
"loss": 0.224,
|
840 |
+
"step": 10300
|
841 |
+
},
|
842 |
+
{
|
843 |
+
"epoch": 0.23294360077050574,
|
844 |
+
"grad_norm": 11.526522636413574,
|
845 |
+
"learning_rate": 8.737837336100233e-05,
|
846 |
+
"loss": 0.2288,
|
847 |
+
"step": 10400
|
848 |
+
},
|
849 |
+
{
|
850 |
+
"epoch": 0.23518344308560676,
|
851 |
+
"grad_norm": 15.950361251831055,
|
852 |
+
"learning_rate": 8.714323825423031e-05,
|
853 |
+
"loss": 0.2388,
|
854 |
+
"step": 10500
|
855 |
+
},
|
856 |
+
{
|
857 |
+
"epoch": 0.23742328540070778,
|
858 |
+
"grad_norm": 11.149569511413574,
|
859 |
+
"learning_rate": 8.690625575024586e-05,
|
860 |
+
"loss": 0.2067,
|
861 |
+
"step": 10600
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 0.2396631277158088,
|
865 |
+
"grad_norm": 10.904093742370605,
|
866 |
+
"learning_rate": 8.666743763587315e-05,
|
867 |
+
"loss": 0.2354,
|
868 |
+
"step": 10700
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"epoch": 0.24190297003090983,
|
872 |
+
"grad_norm": 10.213909149169922,
|
873 |
+
"learning_rate": 8.642679578923422e-05,
|
874 |
+
"loss": 0.2236,
|
875 |
+
"step": 10800
|
876 |
+
},
|
877 |
+
{
|
878 |
+
"epoch": 0.24414281234601085,
|
879 |
+
"grad_norm": 10.054357528686523,
|
880 |
+
"learning_rate": 8.618434217915838e-05,
|
881 |
+
"loss": 0.1987,
|
882 |
+
"step": 10900
|
883 |
+
},
|
884 |
+
{
|
885 |
+
"epoch": 0.24638265466111187,
|
886 |
+
"grad_norm": 11.040483474731445,
|
887 |
+
"learning_rate": 8.59400888645867e-05,
|
888 |
+
"loss": 0.2471,
|
889 |
+
"step": 11000
|
890 |
+
},
|
891 |
+
{
|
892 |
+
"epoch": 0.24638265466111187,
|
893 |
+
"eval_avg_non_pair_similarity": 0.00023486310084357872,
|
894 |
+
"eval_avg_pair_similarity": 0.00031724373530596497,
|
895 |
+
"eval_loss": 0.17524191737174988,
|
896 |
+
"eval_runtime": 19.4456,
|
897 |
+
"eval_samples_per_second": 25.713,
|
898 |
+
"eval_similarity_ratio": 1.3507602265596101,
|
899 |
+
"eval_steps_per_second": 0.823,
|
900 |
+
"step": 11000
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"epoch": 0.2486224969762129,
|
904 |
+
"grad_norm": 12.810904502868652,
|
905 |
+
"learning_rate": 8.569404799397247e-05,
|
906 |
+
"loss": 0.1974,
|
907 |
+
"step": 11100
|
908 |
+
},
|
909 |
+
{
|
910 |
+
"epoch": 0.2508623392913139,
|
911 |
+
"grad_norm": 15.360380172729492,
|
912 |
+
"learning_rate": 8.544623180467673e-05,
|
913 |
+
"loss": 0.2204,
|
914 |
+
"step": 11200
|
915 |
+
},
|
916 |
+
{
|
917 |
+
"epoch": 0.2531021816064149,
|
918 |
+
"grad_norm": 13.429972648620605,
|
919 |
+
"learning_rate": 8.519665262235979e-05,
|
920 |
+
"loss": 0.2479,
|
921 |
+
"step": 11300
|
922 |
+
},
|
923 |
+
{
|
924 |
+
"epoch": 0.25534202392151595,
|
925 |
+
"grad_norm": 10.252517700195312,
|
926 |
+
"learning_rate": 8.494532286036813e-05,
|
927 |
+
"loss": 0.1886,
|
928 |
+
"step": 11400
|
929 |
+
},
|
930 |
+
{
|
931 |
+
"epoch": 0.25758186623661694,
|
932 |
+
"grad_norm": 11.51858901977539,
|
933 |
+
"learning_rate": 8.4692255019117e-05,
|
934 |
+
"loss": 0.2191,
|
935 |
+
"step": 11500
|
936 |
+
},
|
937 |
+
{
|
938 |
+
"epoch": 0.25982170855171793,
|
939 |
+
"grad_norm": 8.529491424560547,
|
940 |
+
"learning_rate": 8.443746168546867e-05,
|
941 |
+
"loss": 0.1856,
|
942 |
+
"step": 11600
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 0.262061550866819,
|
946 |
+
"grad_norm": 7.846275806427002,
|
947 |
+
"learning_rate": 8.418095553210641e-05,
|
948 |
+
"loss": 0.1879,
|
949 |
+
"step": 11700
|
950 |
+
},
|
951 |
+
{
|
952 |
+
"epoch": 0.26430139318192,
|
953 |
+
"grad_norm": 10.658947944641113,
|
954 |
+
"learning_rate": 8.392274931690422e-05,
|
955 |
+
"loss": 0.1998,
|
956 |
+
"step": 11800
|
957 |
+
},
|
958 |
+
{
|
959 |
+
"epoch": 0.266541235497021,
|
960 |
+
"grad_norm": 8.623246192932129,
|
961 |
+
"learning_rate": 8.366285588229223e-05,
|
962 |
+
"loss": 0.2073,
|
963 |
+
"step": 11900
|
964 |
+
},
|
965 |
+
{
|
966 |
+
"epoch": 0.268781077812122,
|
967 |
+
"grad_norm": 8.197981834411621,
|
968 |
+
"learning_rate": 8.340128815461797e-05,
|
969 |
+
"loss": 0.179,
|
970 |
+
"step": 12000
|
971 |
+
},
|
972 |
+
{
|
973 |
+
"epoch": 0.268781077812122,
|
974 |
+
"eval_avg_non_pair_similarity": 5.694422117604282e-05,
|
975 |
+
"eval_avg_pair_similarity": -0.00526782719604671,
|
976 |
+
"eval_loss": 0.19077710807323456,
|
977 |
+
"eval_runtime": 19.459,
|
978 |
+
"eval_samples_per_second": 25.695,
|
979 |
+
"eval_similarity_ratio": -92.50854761471308,
|
980 |
+
"eval_steps_per_second": 0.822,
|
981 |
+
"step": 12000
|
982 |
+
},
|
983 |
+
{
|
984 |
+
"epoch": 0.27102092012722306,
|
985 |
+
"grad_norm": 9.695572853088379,
|
986 |
+
"learning_rate": 8.313805914350349e-05,
|
987 |
+
"loss": 0.2022,
|
988 |
+
"step": 12100
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"epoch": 0.27326076244232406,
|
992 |
+
"grad_norm": 8.658490180969238,
|
993 |
+
"learning_rate": 8.287318194119825e-05,
|
994 |
+
"loss": 0.2011,
|
995 |
+
"step": 12200
|
996 |
+
},
|
997 |
+
{
|
998 |
+
"epoch": 0.2755006047574251,
|
999 |
+
"grad_norm": 9.974045753479004,
|
1000 |
+
"learning_rate": 8.260666972192798e-05,
|
1001 |
+
"loss": 0.2163,
|
1002 |
+
"step": 12300
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"epoch": 0.2777404470725261,
|
1006 |
+
"grad_norm": 8.445626258850098,
|
1007 |
+
"learning_rate": 8.233853574123944e-05,
|
1008 |
+
"loss": 0.1947,
|
1009 |
+
"step": 12400
|
1010 |
+
},
|
1011 |
+
{
|
1012 |
+
"epoch": 0.2799802893876271,
|
1013 |
+
"grad_norm": 7.46968412399292,
|
1014 |
+
"learning_rate": 8.206879333534109e-05,
|
1015 |
+
"loss": 0.1709,
|
1016 |
+
"step": 12500
|
1017 |
+
},
|
1018 |
+
{
|
1019 |
+
"epoch": 0.28222013170272814,
|
1020 |
+
"grad_norm": 12.895581245422363,
|
1021 |
+
"learning_rate": 8.179745592043982e-05,
|
1022 |
+
"loss": 0.2281,
|
1023 |
+
"step": 12600
|
1024 |
+
},
|
1025 |
+
{
|
1026 |
+
"epoch": 0.28445997401782913,
|
1027 |
+
"grad_norm": 10.54282283782959,
|
1028 |
+
"learning_rate": 8.152453699207362e-05,
|
1029 |
+
"loss": 0.184,
|
1030 |
+
"step": 12700
|
1031 |
+
},
|
1032 |
+
{
|
1033 |
+
"epoch": 0.2866998163329302,
|
1034 |
+
"grad_norm": 7.9730634689331055,
|
1035 |
+
"learning_rate": 8.125005012444044e-05,
|
1036 |
+
"loss": 0.2525,
|
1037 |
+
"step": 12800
|
1038 |
+
},
|
1039 |
+
{
|
1040 |
+
"epoch": 0.2889396586480312,
|
1041 |
+
"grad_norm": 9.222777366638184,
|
1042 |
+
"learning_rate": 8.097400896972296e-05,
|
1043 |
+
"loss": 0.1912,
|
1044 |
+
"step": 12900
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 0.2911795009631322,
|
1048 |
+
"grad_norm": 10.500025749206543,
|
1049 |
+
"learning_rate": 8.069642725740957e-05,
|
1050 |
+
"loss": 0.1948,
|
1051 |
+
"step": 13000
|
1052 |
+
},
|
1053 |
+
{
|
1054 |
+
"epoch": 0.2911795009631322,
|
1055 |
+
"eval_avg_non_pair_similarity": 0.0007876502425145761,
|
1056 |
+
"eval_avg_pair_similarity": -0.007292054696008563,
|
1057 |
+
"eval_loss": 0.18571369349956512,
|
1058 |
+
"eval_runtime": 19.4919,
|
1059 |
+
"eval_samples_per_second": 25.652,
|
1060 |
+
"eval_similarity_ratio": -9.257985718037302,
|
1061 |
+
"eval_steps_per_second": 0.821,
|
1062 |
+
"step": 13000
|
1063 |
+
},
|
1064 |
+
{
|
1065 |
+
"epoch": 0.2934193432782332,
|
1066 |
+
"grad_norm": 7.784255504608154,
|
1067 |
+
"learning_rate": 8.041731879361164e-05,
|
1068 |
+
"loss": 0.214,
|
1069 |
+
"step": 13100
|
1070 |
+
},
|
1071 |
+
{
|
1072 |
+
"epoch": 0.2956591855933342,
|
1073 |
+
"grad_norm": 9.356244087219238,
|
1074 |
+
"learning_rate": 8.013669746037662e-05,
|
1075 |
+
"loss": 0.168,
|
1076 |
+
"step": 13200
|
1077 |
+
},
|
1078 |
+
{
|
1079 |
+
"epoch": 0.29789902790843525,
|
1080 |
+
"grad_norm": 8.42744255065918,
|
1081 |
+
"learning_rate": 7.985457721499779e-05,
|
1082 |
+
"loss": 0.1759,
|
1083 |
+
"step": 13300
|
1084 |
+
},
|
1085 |
+
{
|
1086 |
+
"epoch": 0.30013887022353625,
|
1087 |
+
"grad_norm": 8.323335647583008,
|
1088 |
+
"learning_rate": 7.957097208931997e-05,
|
1089 |
+
"loss": 0.193,
|
1090 |
+
"step": 13400
|
1091 |
+
},
|
1092 |
+
{
|
1093 |
+
"epoch": 0.3023787125386373,
|
1094 |
+
"grad_norm": 10.35467529296875,
|
1095 |
+
"learning_rate": 7.928589618904158e-05,
|
1096 |
+
"loss": 0.244,
|
1097 |
+
"step": 13500
|
1098 |
+
},
|
1099 |
+
{
|
1100 |
+
"epoch": 0.3046185548537383,
|
1101 |
+
"grad_norm": 9.027731895446777,
|
1102 |
+
"learning_rate": 7.899936369301319e-05,
|
1103 |
+
"loss": 0.1878,
|
1104 |
+
"step": 13600
|
1105 |
+
},
|
1106 |
+
{
|
1107 |
+
"epoch": 0.30685839716883934,
|
1108 |
+
"grad_norm": 11.577035903930664,
|
1109 |
+
"learning_rate": 7.871138885253216e-05,
|
1110 |
+
"loss": 0.2033,
|
1111 |
+
"step": 13700
|
1112 |
+
},
|
1113 |
+
{
|
1114 |
+
"epoch": 0.30909823948394033,
|
1115 |
+
"grad_norm": 6.524970054626465,
|
1116 |
+
"learning_rate": 7.842198599063395e-05,
|
1117 |
+
"loss": 0.1792,
|
1118 |
+
"step": 13800
|
1119 |
+
},
|
1120 |
+
{
|
1121 |
+
"epoch": 0.3113380817990413,
|
1122 |
+
"grad_norm": 8.987520217895508,
|
1123 |
+
"learning_rate": 7.813116950137964e-05,
|
1124 |
+
"loss": 0.1936,
|
1125 |
+
"step": 13900
|
1126 |
+
},
|
1127 |
+
{
|
1128 |
+
"epoch": 0.31357792411414237,
|
1129 |
+
"grad_norm": 8.995404243469238,
|
1130 |
+
"learning_rate": 7.783895384914004e-05,
|
1131 |
+
"loss": 0.1831,
|
1132 |
+
"step": 14000
|
1133 |
+
},
|
1134 |
+
{
|
1135 |
+
"epoch": 0.31357792411414237,
|
1136 |
+
"eval_avg_non_pair_similarity": 0.000529102375272793,
|
1137 |
+
"eval_avg_pair_similarity": -0.001620051197707653,
|
1138 |
+
"eval_loss": 0.18957822024822235,
|
1139 |
+
"eval_runtime": 19.4615,
|
1140 |
+
"eval_samples_per_second": 25.692,
|
1141 |
+
"eval_similarity_ratio": -3.061886079933759,
|
1142 |
+
"eval_steps_per_second": 0.822,
|
1143 |
+
"step": 14000
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 0.31581776642924336,
|
1147 |
+
"grad_norm": 6.4864068031311035,
|
1148 |
+
"learning_rate": 7.754535356787629e-05,
|
1149 |
+
"loss": 0.1976,
|
1150 |
+
"step": 14100
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 0.3180576087443444,
|
1154 |
+
"grad_norm": 7.537520885467529,
|
1155 |
+
"learning_rate": 7.725038326041698e-05,
|
1156 |
+
"loss": 0.1935,
|
1157 |
+
"step": 14200
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 0.3202974510594454,
|
1161 |
+
"grad_norm": 7.547581195831299,
|
1162 |
+
"learning_rate": 7.695405759773182e-05,
|
1163 |
+
"loss": 0.1706,
|
1164 |
+
"step": 14300
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 0.32253729337454645,
|
1168 |
+
"grad_norm": 6.900126934051514,
|
1169 |
+
"learning_rate": 7.665639131820198e-05,
|
1170 |
+
"loss": 0.1479,
|
1171 |
+
"step": 14400
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 0.32477713568964744,
|
1175 |
+
"grad_norm": 9.628849983215332,
|
1176 |
+
"learning_rate": 7.635739922688702e-05,
|
1177 |
+
"loss": 0.1649,
|
1178 |
+
"step": 14500
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 0.3270169780047485,
|
1182 |
+
"grad_norm": 9.103193283081055,
|
1183 |
+
"learning_rate": 7.60570961947886e-05,
|
1184 |
+
"loss": 0.1672,
|
1185 |
+
"step": 14600
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 0.3292568203198495,
|
1189 |
+
"grad_norm": 9.142544746398926,
|
1190 |
+
"learning_rate": 7.57554971581107e-05,
|
1191 |
+
"loss": 0.1831,
|
1192 |
+
"step": 14700
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 0.3314966626349505,
|
1196 |
+
"grad_norm": 8.968326568603516,
|
1197 |
+
"learning_rate": 7.545261711751685e-05,
|
1198 |
+
"loss": 0.1566,
|
1199 |
+
"step": 14800
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 0.3337365049500515,
|
1203 |
+
"grad_norm": 9.484015464782715,
|
1204 |
+
"learning_rate": 7.514847113738401e-05,
|
1205 |
+
"loss": 0.1949,
|
1206 |
+
"step": 14900
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 0.3359763472651525,
|
1210 |
+
"grad_norm": 7.016854763031006,
|
1211 |
+
"learning_rate": 7.484307434505335e-05,
|
1212 |
+
"loss": 0.1419,
|
1213 |
+
"step": 15000
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 0.3359763472651525,
|
1217 |
+
"eval_avg_non_pair_similarity": 8.844231431561901e-05,
|
1218 |
+
"eval_avg_pair_similarity": -0.0006939181396737695,
|
1219 |
+
"eval_loss": 0.1548442244529724,
|
1220 |
+
"eval_runtime": 19.513,
|
1221 |
+
"eval_samples_per_second": 25.624,
|
1222 |
+
"eval_similarity_ratio": -7.845997077794952,
|
1223 |
+
"eval_steps_per_second": 0.82,
|
1224 |
+
"step": 15000
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 0.33821618958025357,
|
1228 |
+
"grad_norm": 7.563048362731934,
|
1229 |
+
"learning_rate": 7.453644193007774e-05,
|
1230 |
+
"loss": 0.1891,
|
1231 |
+
"step": 15100
|
1232 |
+
},
|
1233 |
+
{
|
1234 |
+
"epoch": 0.34045603189535456,
|
1235 |
+
"grad_norm": 7.630728244781494,
|
1236 |
+
"learning_rate": 7.422858914346641e-05,
|
1237 |
+
"loss": 0.1896,
|
1238 |
+
"step": 15200
|
1239 |
+
},
|
1240 |
+
{
|
1241 |
+
"epoch": 0.3426958742104556,
|
1242 |
+
"grad_norm": 7.264196395874023,
|
1243 |
+
"learning_rate": 7.391953129692633e-05,
|
1244 |
+
"loss": 0.1608,
|
1245 |
+
"step": 15300
|
1246 |
+
},
|
1247 |
+
{
|
1248 |
+
"epoch": 0.3449357165255566,
|
1249 |
+
"grad_norm": 10.433218955993652,
|
1250 |
+
"learning_rate": 7.360928376210064e-05,
|
1251 |
+
"loss": 0.1668,
|
1252 |
+
"step": 15400
|
1253 |
+
},
|
1254 |
+
{
|
1255 |
+
"epoch": 0.3471755588406576,
|
1256 |
+
"grad_norm": 7.074268817901611,
|
1257 |
+
"learning_rate": 7.329786196980417e-05,
|
1258 |
+
"loss": 0.1882,
|
1259 |
+
"step": 15500
|
1260 |
+
},
|
1261 |
+
{
|
1262 |
+
"epoch": 0.34941540115575864,
|
1263 |
+
"grad_norm": 8.925777435302734,
|
1264 |
+
"learning_rate": 7.298528140925589e-05,
|
1265 |
+
"loss": 0.1697,
|
1266 |
+
"step": 15600
|
1267 |
+
},
|
1268 |
+
{
|
1269 |
+
"epoch": 0.35165524347085964,
|
1270 |
+
"grad_norm": 9.438268661499023,
|
1271 |
+
"learning_rate": 7.267155762730855e-05,
|
1272 |
+
"loss": 0.1566,
|
1273 |
+
"step": 15700
|
1274 |
+
},
|
1275 |
+
{
|
1276 |
+
"epoch": 0.3538950857859607,
|
1277 |
+
"grad_norm": 8.304781913757324,
|
1278 |
+
"learning_rate": 7.235670622767546e-05,
|
1279 |
+
"loss": 0.1624,
|
1280 |
+
"step": 15800
|
1281 |
+
},
|
1282 |
+
{
|
1283 |
+
"epoch": 0.3561349281010617,
|
1284 |
+
"grad_norm": 6.194468975067139,
|
1285 |
+
"learning_rate": 7.204074287015433e-05,
|
1286 |
+
"loss": 0.1423,
|
1287 |
+
"step": 15900
|
1288 |
+
},
|
1289 |
+
{
|
1290 |
+
"epoch": 0.3583747704161627,
|
1291 |
+
"grad_norm": 10.31971549987793,
|
1292 |
+
"learning_rate": 7.172368326984846e-05,
|
1293 |
+
"loss": 0.15,
|
1294 |
+
"step": 16000
|
1295 |
+
},
|
1296 |
+
{
|
1297 |
+
"epoch": 0.3583747704161627,
|
1298 |
+
"eval_avg_non_pair_similarity": 0.0005423984809662916,
|
1299 |
+
"eval_avg_pair_similarity": -0.004194568347651512,
|
1300 |
+
"eval_loss": 0.16588900983333588,
|
1301 |
+
"eval_runtime": 19.5058,
|
1302 |
+
"eval_samples_per_second": 25.633,
|
1303 |
+
"eval_similarity_ratio": -7.7333703814561225,
|
1304 |
+
"eval_steps_per_second": 0.82,
|
1305 |
+
"step": 16000
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 0.3606146127312637,
|
1309 |
+
"grad_norm": 6.6040449142456055,
|
1310 |
+
"learning_rate": 7.14055431963851e-05,
|
1311 |
+
"loss": 0.1929,
|
1312 |
+
"step": 16100
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 0.3628544550463647,
|
1316 |
+
"grad_norm": 6.839012622833252,
|
1317 |
+
"learning_rate": 7.108633847313109e-05,
|
1318 |
+
"loss": 0.169,
|
1319 |
+
"step": 16200
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 0.36509429736146576,
|
1323 |
+
"grad_norm": 8.55441665649414,
|
1324 |
+
"learning_rate": 7.076608497640588e-05,
|
1325 |
+
"loss": 0.1772,
|
1326 |
+
"step": 16300
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 0.36733413967656675,
|
1330 |
+
"grad_norm": 7.641045093536377,
|
1331 |
+
"learning_rate": 7.044479863469189e-05,
|
1332 |
+
"loss": 0.1544,
|
1333 |
+
"step": 16400
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 0.3695739819916678,
|
1337 |
+
"grad_norm": 9.20849895477295,
|
1338 |
+
"learning_rate": 7.012249542784223e-05,
|
1339 |
+
"loss": 0.1485,
|
1340 |
+
"step": 16500
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 0.3718138243067688,
|
1344 |
+
"grad_norm": 6.943836212158203,
|
1345 |
+
"learning_rate": 6.979919138628598e-05,
|
1346 |
+
"loss": 0.1611,
|
1347 |
+
"step": 16600
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 0.37405366662186984,
|
1351 |
+
"grad_norm": 8.36039924621582,
|
1352 |
+
"learning_rate": 6.94749025902308e-05,
|
1353 |
+
"loss": 0.1646,
|
1354 |
+
"step": 16700
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 0.37629350893697083,
|
1358 |
+
"grad_norm": 7.000893592834473,
|
1359 |
+
"learning_rate": 6.914964516886324e-05,
|
1360 |
+
"loss": 0.1719,
|
1361 |
+
"step": 16800
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 0.3785333512520719,
|
1365 |
+
"grad_norm": 8.178186416625977,
|
1366 |
+
"learning_rate": 6.882343529954641e-05,
|
1367 |
+
"loss": 0.1528,
|
1368 |
+
"step": 16900
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 0.3807731935671729,
|
1372 |
+
"grad_norm": 7.016880512237549,
|
1373 |
+
"learning_rate": 6.849628920701549e-05,
|
1374 |
+
"loss": 0.1609,
|
1375 |
+
"step": 17000
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 0.3807731935671729,
|
1379 |
+
"eval_avg_non_pair_similarity": -0.0001115500947464265,
|
1380 |
+
"eval_avg_pair_similarity": 0.00017968433955684303,
|
1381 |
+
"eval_loss": 0.13874633610248566,
|
1382 |
+
"eval_runtime": 19.4966,
|
1383 |
+
"eval_samples_per_second": 25.646,
|
1384 |
+
"eval_similarity_ratio": -1.6107950420418555,
|
1385 |
+
"eval_steps_per_second": 0.821,
|
1386 |
+
"step": 17000
|
1387 |
+
},
|
1388 |
+
{
|
1389 |
+
"epoch": 0.38301303588227387,
|
1390 |
+
"grad_norm": 6.156219005584717,
|
1391 |
+
"learning_rate": 6.816822316257066e-05,
|
1392 |
+
"loss": 0.1405,
|
1393 |
+
"step": 17100
|
1394 |
+
},
|
1395 |
+
{
|
1396 |
+
"epoch": 0.3852528781973749,
|
1397 |
+
"grad_norm": 6.131289482116699,
|
1398 |
+
"learning_rate": 6.783925348326782e-05,
|
1399 |
+
"loss": 0.167,
|
1400 |
+
"step": 17200
|
1401 |
+
},
|
1402 |
+
{
|
1403 |
+
"epoch": 0.3874927205124759,
|
1404 |
+
"grad_norm": 6.05889892578125,
|
1405 |
+
"learning_rate": 6.750939653110711e-05,
|
1406 |
+
"loss": 0.1819,
|
1407 |
+
"step": 17300
|
1408 |
+
},
|
1409 |
+
{
|
1410 |
+
"epoch": 0.38973256282757696,
|
1411 |
+
"grad_norm": 6.4408488273620605,
|
1412 |
+
"learning_rate": 6.717866871221901e-05,
|
1413 |
+
"loss": 0.1518,
|
1414 |
+
"step": 17400
|
1415 |
+
},
|
1416 |
+
{
|
1417 |
+
"epoch": 0.39197240514267795,
|
1418 |
+
"grad_norm": 7.979753494262695,
|
1419 |
+
"learning_rate": 6.684708647604843e-05,
|
1420 |
+
"loss": 0.1812,
|
1421 |
+
"step": 17500
|
1422 |
+
},
|
1423 |
+
{
|
1424 |
+
"epoch": 0.394212247457779,
|
1425 |
+
"grad_norm": 6.986364364624023,
|
1426 |
+
"learning_rate": 6.65146663145365e-05,
|
1427 |
+
"loss": 0.1447,
|
1428 |
+
"step": 17600
|
1429 |
+
},
|
1430 |
+
{
|
1431 |
+
"epoch": 0.39645208977288,
|
1432 |
+
"grad_norm": 6.019850730895996,
|
1433 |
+
"learning_rate": 6.618142476130032e-05,
|
1434 |
+
"loss": 0.1893,
|
1435 |
+
"step": 17700
|
1436 |
+
},
|
1437 |
+
{
|
1438 |
+
"epoch": 0.398691932087981,
|
1439 |
+
"grad_norm": 6.467818260192871,
|
1440 |
+
"learning_rate": 6.584737839081066e-05,
|
1441 |
+
"loss": 0.1751,
|
1442 |
+
"step": 17800
|
1443 |
+
},
|
1444 |
+
{
|
1445 |
+
"epoch": 0.40093177440308203,
|
1446 |
+
"grad_norm": 9.779342651367188,
|
1447 |
+
"learning_rate": 6.551254381756758e-05,
|
1448 |
+
"loss": 0.1524,
|
1449 |
+
"step": 17900
|
1450 |
+
},
|
1451 |
+
{
|
1452 |
+
"epoch": 0.403171616718183,
|
1453 |
+
"grad_norm": 6.93033504486084,
|
1454 |
+
"learning_rate": 6.517693769527402e-05,
|
1455 |
+
"loss": 0.1327,
|
1456 |
+
"step": 18000
|
1457 |
+
},
|
1458 |
+
{
|
1459 |
+
"epoch": 0.403171616718183,
|
1460 |
+
"eval_avg_non_pair_similarity": 0.0004680827924466593,
|
1461 |
+
"eval_avg_pair_similarity": 0.0004896340477280319,
|
1462 |
+
"eval_loss": 0.11614304035902023,
|
1463 |
+
"eval_runtime": 19.4704,
|
1464 |
+
"eval_samples_per_second": 25.68,
|
1465 |
+
"eval_similarity_ratio": 1.0460415457033243,
|
1466 |
+
"eval_steps_per_second": 0.822,
|
1467 |
+
"step": 18000
|
1468 |
+
}
|
1469 |
+
],
|
1470 |
+
"logging_steps": 100,
|
1471 |
+
"max_steps": 44646,
|
1472 |
+
"num_input_tokens_seen": 0,
|
1473 |
+
"num_train_epochs": 1,
|
1474 |
+
"save_steps": 1000,
|
1475 |
+
"stateful_callbacks": {
|
1476 |
+
"TrainerControl": {
|
1477 |
+
"args": {
|
1478 |
+
"should_epoch_stop": false,
|
1479 |
+
"should_evaluate": false,
|
1480 |
+
"should_log": false,
|
1481 |
+
"should_save": true,
|
1482 |
+
"should_training_stop": false
|
1483 |
+
},
|
1484 |
+
"attributes": {}
|
1485 |
+
}
|
1486 |
+
},
|
1487 |
+
"total_flos": 0.0,
|
1488 |
+
"train_batch_size": 32,
|
1489 |
+
"trial_name": null,
|
1490 |
+
"trial_params": null
|
1491 |
+
}
|
checkpoint-18000/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7c6bbe0930b5977d5ef83c9c6a2872aea787e9f64c3d133a00319e43c833ad34
|
3 |
+
size 5112
|