Training in progress, step 15000, checkpoint
Browse files
checkpoint-15000/config.json
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"annotation_transformer": true,
|
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": "lhallee/annotation_transformer_uniref90",
|
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-15000/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:44d82d9811b5f5ec42b8564dde567436c54e377d991353fcdc17ff7f1841da54
|
3 |
+
size 1384137156
|
checkpoint-15000/trainer_state.json
ADDED
@@ -0,0 +1,1248 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.9307520476545048,
|
5 |
+
"eval_steps": 1000,
|
6 |
+
"global_step": 15000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.0062050136510300325,
|
13 |
+
"grad_norm": 101.88961029052734,
|
14 |
+
"learning_rate": 0.0001,
|
15 |
+
"loss": 4.5088,
|
16 |
+
"step": 100
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.012410027302060065,
|
20 |
+
"grad_norm": 99.7363510131836,
|
21 |
+
"learning_rate": 9.999038127056248e-05,
|
22 |
+
"loss": 0.8644,
|
23 |
+
"step": 200
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.018615040953090096,
|
27 |
+
"grad_norm": 59.41786575317383,
|
28 |
+
"learning_rate": 9.996152878304816e-05,
|
29 |
+
"loss": 0.7189,
|
30 |
+
"step": 300
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.02482005460412013,
|
34 |
+
"grad_norm": 61.05416488647461,
|
35 |
+
"learning_rate": 9.991345363842789e-05,
|
36 |
+
"loss": 0.6253,
|
37 |
+
"step": 400
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.03102506825515016,
|
41 |
+
"grad_norm": 53.87808609008789,
|
42 |
+
"learning_rate": 9.9846174333574e-05,
|
43 |
+
"loss": 0.5863,
|
44 |
+
"step": 500
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.03723008190618019,
|
48 |
+
"grad_norm": 48.58293914794922,
|
49 |
+
"learning_rate": 9.975971675414371e-05,
|
50 |
+
"loss": 0.5555,
|
51 |
+
"step": 600
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.04343509555721023,
|
55 |
+
"grad_norm": 42.353153228759766,
|
56 |
+
"learning_rate": 9.965411416461959e-05,
|
57 |
+
"loss": 0.5566,
|
58 |
+
"step": 700
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.04964010920824026,
|
62 |
+
"grad_norm": 46.06748580932617,
|
63 |
+
"learning_rate": 9.952940719551112e-05,
|
64 |
+
"loss": 0.5425,
|
65 |
+
"step": 800
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.05584512285927029,
|
69 |
+
"grad_norm": 46.486637115478516,
|
70 |
+
"learning_rate": 9.938564382772205e-05,
|
71 |
+
"loss": 0.5004,
|
72 |
+
"step": 900
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.06205013651030032,
|
76 |
+
"grad_norm": 43.99421691894531,
|
77 |
+
"learning_rate": 9.922287937408994e-05,
|
78 |
+
"loss": 0.4722,
|
79 |
+
"step": 1000
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.06205013651030032,
|
83 |
+
"eval_avg_non_pair_similarity": -0.00011276310920446573,
|
84 |
+
"eval_avg_pair_similarity": 0.0020758428336121143,
|
85 |
+
"eval_loss": 0.44835320115089417,
|
86 |
+
"eval_runtime": 52.6112,
|
87 |
+
"eval_samples_per_second": 9.504,
|
88 |
+
"eval_similarity_ratio": -18.408882552609725,
|
89 |
+
"eval_steps_per_second": 0.304,
|
90 |
+
"step": 1000
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"epoch": 0.06825515016133035,
|
94 |
+
"grad_norm": 38.14698791503906,
|
95 |
+
"learning_rate": 9.904117645810441e-05,
|
96 |
+
"loss": 0.4444,
|
97 |
+
"step": 1100
|
98 |
+
},
|
99 |
+
{
|
100 |
+
"epoch": 0.07446016381236038,
|
101 |
+
"grad_norm": 29.508113861083984,
|
102 |
+
"learning_rate": 9.884060498981296e-05,
|
103 |
+
"loss": 0.4615,
|
104 |
+
"step": 1200
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 0.08066517746339041,
|
108 |
+
"grad_norm": 34.64622497558594,
|
109 |
+
"learning_rate": 9.862124213892304e-05,
|
110 |
+
"loss": 0.4665,
|
111 |
+
"step": 1300
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 0.08687019111442046,
|
115 |
+
"grad_norm": 40.75010299682617,
|
116 |
+
"learning_rate": 9.838317230511112e-05,
|
117 |
+
"loss": 0.4275,
|
118 |
+
"step": 1400
|
119 |
+
},
|
120 |
+
{
|
121 |
+
"epoch": 0.09307520476545049,
|
122 |
+
"grad_norm": 28.284589767456055,
|
123 |
+
"learning_rate": 9.81264870855499e-05,
|
124 |
+
"loss": 0.4129,
|
125 |
+
"step": 1500
|
126 |
+
},
|
127 |
+
{
|
128 |
+
"epoch": 0.09928021841648052,
|
129 |
+
"grad_norm": 34.2197151184082,
|
130 |
+
"learning_rate": 9.785128523966653e-05,
|
131 |
+
"loss": 0.4389,
|
132 |
+
"step": 1600
|
133 |
+
},
|
134 |
+
{
|
135 |
+
"epoch": 0.10548523206751055,
|
136 |
+
"grad_norm": 27.808895111083984,
|
137 |
+
"learning_rate": 9.755767265114484e-05,
|
138 |
+
"loss": 0.4184,
|
139 |
+
"step": 1700
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"epoch": 0.11169024571854058,
|
143 |
+
"grad_norm": 35.00907897949219,
|
144 |
+
"learning_rate": 9.724576228718678e-05,
|
145 |
+
"loss": 0.3995,
|
146 |
+
"step": 1800
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 0.11789525936957061,
|
150 |
+
"grad_norm": 30.57769203186035,
|
151 |
+
"learning_rate": 9.691567415504832e-05,
|
152 |
+
"loss": 0.415,
|
153 |
+
"step": 1900
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 0.12410027302060064,
|
157 |
+
"grad_norm": 26.989404678344727,
|
158 |
+
"learning_rate": 9.656753525586681e-05,
|
159 |
+
"loss": 0.4052,
|
160 |
+
"step": 2000
|
161 |
+
},
|
162 |
+
{
|
163 |
+
"epoch": 0.12410027302060064,
|
164 |
+
"eval_avg_non_pair_similarity": 0.0006865022985091086,
|
165 |
+
"eval_avg_pair_similarity": 0.007665629971772433,
|
166 |
+
"eval_loss": 0.3601702153682709,
|
167 |
+
"eval_runtime": 52.565,
|
168 |
+
"eval_samples_per_second": 9.512,
|
169 |
+
"eval_similarity_ratio": 11.166211662247951,
|
170 |
+
"eval_steps_per_second": 0.304,
|
171 |
+
"step": 2000
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.1303052866716307,
|
175 |
+
"grad_norm": 32.37137222290039,
|
176 |
+
"learning_rate": 9.620147953579737e-05,
|
177 |
+
"loss": 0.3979,
|
178 |
+
"step": 2100
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.1365103003226607,
|
182 |
+
"grad_norm": 22.6475772857666,
|
183 |
+
"learning_rate": 9.581764783447719e-05,
|
184 |
+
"loss": 0.3807,
|
185 |
+
"step": 2200
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.14271531397369075,
|
189 |
+
"grad_norm": 22.612077713012695,
|
190 |
+
"learning_rate": 9.54161878308377e-05,
|
191 |
+
"loss": 0.3711,
|
192 |
+
"step": 2300
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.14892032762472077,
|
196 |
+
"grad_norm": 30.973310470581055,
|
197 |
+
"learning_rate": 9.499725398628507e-05,
|
198 |
+
"loss": 0.3658,
|
199 |
+
"step": 2400
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.1551253412757508,
|
203 |
+
"grad_norm": 24.443492889404297,
|
204 |
+
"learning_rate": 9.456100748527143e-05,
|
205 |
+
"loss": 0.37,
|
206 |
+
"step": 2500
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.16133035492678083,
|
210 |
+
"grad_norm": 33.75666809082031,
|
211 |
+
"learning_rate": 9.410761617327921e-05,
|
212 |
+
"loss": 0.3512,
|
213 |
+
"step": 2600
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.16753536857781087,
|
217 |
+
"grad_norm": 23.727365493774414,
|
218 |
+
"learning_rate": 9.363725449224282e-05,
|
219 |
+
"loss": 0.3318,
|
220 |
+
"step": 2700
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.17374038222884092,
|
224 |
+
"grad_norm": 28.48720359802246,
|
225 |
+
"learning_rate": 9.315010341343213e-05,
|
226 |
+
"loss": 0.3729,
|
227 |
+
"step": 2800
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.17994539587987093,
|
231 |
+
"grad_norm": 28.53714370727539,
|
232 |
+
"learning_rate": 9.264635036782405e-05,
|
233 |
+
"loss": 0.3369,
|
234 |
+
"step": 2900
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.18615040953090098,
|
238 |
+
"grad_norm": 21.766740798950195,
|
239 |
+
"learning_rate": 9.212618917398855e-05,
|
240 |
+
"loss": 0.317,
|
241 |
+
"step": 3000
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.18615040953090098,
|
245 |
+
"eval_avg_non_pair_similarity": 0.00025785160030682527,
|
246 |
+
"eval_avg_pair_similarity": 0.0073178326906636355,
|
247 |
+
"eval_loss": 0.323689728975296,
|
248 |
+
"eval_runtime": 52.8381,
|
249 |
+
"eval_samples_per_second": 9.463,
|
250 |
+
"eval_similarity_ratio": 28.38001657525464,
|
251 |
+
"eval_steps_per_second": 0.303,
|
252 |
+
"step": 3000
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 0.192355423181931,
|
256 |
+
"grad_norm": 25.073217391967773,
|
257 |
+
"learning_rate": 9.158981996351711e-05,
|
258 |
+
"loss": 0.3092,
|
259 |
+
"step": 3100
|
260 |
+
},
|
261 |
+
{
|
262 |
+
"epoch": 0.19856043683296104,
|
263 |
+
"grad_norm": 25.02987289428711,
|
264 |
+
"learning_rate": 9.10374491040223e-05,
|
265 |
+
"loss": 0.3177,
|
266 |
+
"step": 3200
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 0.20476545048399106,
|
270 |
+
"grad_norm": 23.147781372070312,
|
271 |
+
"learning_rate": 9.046928911973799e-05,
|
272 |
+
"loss": 0.3357,
|
273 |
+
"step": 3300
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"epoch": 0.2109704641350211,
|
277 |
+
"grad_norm": 18.975126266479492,
|
278 |
+
"learning_rate": 8.988555860975082e-05,
|
279 |
+
"loss": 0.2946,
|
280 |
+
"step": 3400
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"epoch": 0.21717547778605112,
|
284 |
+
"grad_norm": 22.149898529052734,
|
285 |
+
"learning_rate": 8.92864821638944e-05,
|
286 |
+
"loss": 0.3184,
|
287 |
+
"step": 3500
|
288 |
+
},
|
289 |
+
{
|
290 |
+
"epoch": 0.22338049143708116,
|
291 |
+
"grad_norm": 31.714447021484375,
|
292 |
+
"learning_rate": 8.86722902763385e-05,
|
293 |
+
"loss": 0.315,
|
294 |
+
"step": 3600
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"epoch": 0.22958550508811118,
|
298 |
+
"grad_norm": 22.137062072753906,
|
299 |
+
"learning_rate": 8.804321925690672e-05,
|
300 |
+
"loss": 0.3007,
|
301 |
+
"step": 3700
|
302 |
+
},
|
303 |
+
{
|
304 |
+
"epoch": 0.23579051873914122,
|
305 |
+
"grad_norm": 27.34349250793457,
|
306 |
+
"learning_rate": 8.739951114015636e-05,
|
307 |
+
"loss": 0.3297,
|
308 |
+
"step": 3800
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 0.24199553239017127,
|
312 |
+
"grad_norm": 20.715852737426758,
|
313 |
+
"learning_rate": 8.674141359225589e-05,
|
314 |
+
"loss": 0.2943,
|
315 |
+
"step": 3900
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 0.2482005460412013,
|
319 |
+
"grad_norm": 21.26046371459961,
|
320 |
+
"learning_rate": 8.60691798156956e-05,
|
321 |
+
"loss": 0.284,
|
322 |
+
"step": 4000
|
323 |
+
},
|
324 |
+
{
|
325 |
+
"epoch": 0.2482005460412013,
|
326 |
+
"eval_avg_non_pair_similarity": 0.00036512027781798257,
|
327 |
+
"eval_avg_pair_similarity": 0.002839699084404856,
|
328 |
+
"eval_loss": 0.26940277218818665,
|
329 |
+
"eval_runtime": 52.6691,
|
330 |
+
"eval_samples_per_second": 9.493,
|
331 |
+
"eval_similarity_ratio": 7.777434606961175,
|
332 |
+
"eval_steps_per_second": 0.304,
|
333 |
+
"step": 4000
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 0.25440555969223133,
|
337 |
+
"grad_norm": 22.599706649780273,
|
338 |
+
"learning_rate": 8.538306845186808e-05,
|
339 |
+
"loss": 0.2942,
|
340 |
+
"step": 4100
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 0.2606105733432614,
|
344 |
+
"grad_norm": 16.177858352661133,
|
345 |
+
"learning_rate": 8.468334348155624e-05,
|
346 |
+
"loss": 0.2874,
|
347 |
+
"step": 4200
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 0.26681558699429136,
|
351 |
+
"grad_norm": 15.479412078857422,
|
352 |
+
"learning_rate": 8.39702741233669e-05,
|
353 |
+
"loss": 0.2758,
|
354 |
+
"step": 4300
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 0.2730206006453214,
|
358 |
+
"grad_norm": 20.40460777282715,
|
359 |
+
"learning_rate": 8.324413473014913e-05,
|
360 |
+
"loss": 0.2788,
|
361 |
+
"step": 4400
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 0.27922561429635145,
|
365 |
+
"grad_norm": 15.524222373962402,
|
366 |
+
"learning_rate": 8.250520468343722e-05,
|
367 |
+
"loss": 0.2667,
|
368 |
+
"step": 4500
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 0.2854306279473815,
|
372 |
+
"grad_norm": 19.124486923217773,
|
373 |
+
"learning_rate": 8.175376828595885e-05,
|
374 |
+
"loss": 0.2828,
|
375 |
+
"step": 4600
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 0.29163564159841154,
|
379 |
+
"grad_norm": 14.729342460632324,
|
380 |
+
"learning_rate": 8.09901146522499e-05,
|
381 |
+
"loss": 0.2757,
|
382 |
+
"step": 4700
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 0.29784065524944153,
|
386 |
+
"grad_norm": 19.547435760498047,
|
387 |
+
"learning_rate": 8.021453759741784e-05,
|
388 |
+
"loss": 0.3102,
|
389 |
+
"step": 4800
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 0.3040456689004716,
|
393 |
+
"grad_norm": 16.095699310302734,
|
394 |
+
"learning_rate": 7.942733552409662e-05,
|
395 |
+
"loss": 0.2791,
|
396 |
+
"step": 4900
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 0.3102506825515016,
|
400 |
+
"grad_norm": 25.538352966308594,
|
401 |
+
"learning_rate": 7.862881130763646e-05,
|
402 |
+
"loss": 0.2487,
|
403 |
+
"step": 5000
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 0.3102506825515016,
|
407 |
+
"eval_avg_non_pair_similarity": -2.6742616108896766e-05,
|
408 |
+
"eval_avg_pair_similarity": 0.007982193220406771,
|
409 |
+
"eval_loss": 0.3381979465484619,
|
410 |
+
"eval_runtime": 52.9305,
|
411 |
+
"eval_samples_per_second": 9.446,
|
412 |
+
"eval_similarity_ratio": -298.4821375703496,
|
413 |
+
"eval_steps_per_second": 0.302,
|
414 |
+
"step": 5000
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"epoch": 0.31645569620253167,
|
418 |
+
"grad_norm": 19.506053924560547,
|
419 |
+
"learning_rate": 7.781927217957286e-05,
|
420 |
+
"loss": 0.2534,
|
421 |
+
"step": 5100
|
422 |
+
},
|
423 |
+
{
|
424 |
+
"epoch": 0.32266070985356166,
|
425 |
+
"grad_norm": 17.35991668701172,
|
426 |
+
"learning_rate": 7.699902960941952e-05,
|
427 |
+
"loss": 0.2591,
|
428 |
+
"step": 5200
|
429 |
+
},
|
430 |
+
{
|
431 |
+
"epoch": 0.3288657235045917,
|
432 |
+
"grad_norm": 18.530628204345703,
|
433 |
+
"learning_rate": 7.616839918483061e-05,
|
434 |
+
"loss": 0.2504,
|
435 |
+
"step": 5300
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 0.33507073715562175,
|
439 |
+
"grad_norm": 16.918973922729492,
|
440 |
+
"learning_rate": 7.532770049017883e-05,
|
441 |
+
"loss": 0.2697,
|
442 |
+
"step": 5400
|
443 |
+
},
|
444 |
+
{
|
445 |
+
"epoch": 0.3412757508066518,
|
446 |
+
"grad_norm": 15.549245834350586,
|
447 |
+
"learning_rate": 7.447725698359547e-05,
|
448 |
+
"loss": 0.269,
|
449 |
+
"step": 5500
|
450 |
+
},
|
451 |
+
{
|
452 |
+
"epoch": 0.34748076445768183,
|
453 |
+
"grad_norm": 14.331618309020996,
|
454 |
+
"learning_rate": 7.361739587252019e-05,
|
455 |
+
"loss": 0.2466,
|
456 |
+
"step": 5600
|
457 |
+
},
|
458 |
+
{
|
459 |
+
"epoch": 0.3536857781087118,
|
460 |
+
"grad_norm": 13.95034122467041,
|
461 |
+
"learning_rate": 7.274844798780826e-05,
|
462 |
+
"loss": 0.2491,
|
463 |
+
"step": 5700
|
464 |
+
},
|
465 |
+
{
|
466 |
+
"epoch": 0.35989079175974187,
|
467 |
+
"grad_norm": 13.919034004211426,
|
468 |
+
"learning_rate": 7.18707476564436e-05,
|
469 |
+
"loss": 0.2189,
|
470 |
+
"step": 5800
|
471 |
+
},
|
472 |
+
{
|
473 |
+
"epoch": 0.3660958054107719,
|
474 |
+
"grad_norm": 15.497775077819824,
|
475 |
+
"learning_rate": 7.098463257290679e-05,
|
476 |
+
"loss": 0.1965,
|
477 |
+
"step": 5900
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 0.37230081906180196,
|
481 |
+
"grad_norm": 17.710037231445312,
|
482 |
+
"learning_rate": 7.009044366924739e-05,
|
483 |
+
"loss": 0.2509,
|
484 |
+
"step": 6000
|
485 |
+
},
|
486 |
+
{
|
487 |
+
"epoch": 0.37230081906180196,
|
488 |
+
"eval_avg_non_pair_similarity": 2.055158477392279e-05,
|
489 |
+
"eval_avg_pair_similarity": 0.006798229112289846,
|
490 |
+
"eval_loss": 0.24542327225208282,
|
491 |
+
"eval_runtime": 52.6713,
|
492 |
+
"eval_samples_per_second": 9.493,
|
493 |
+
"eval_similarity_ratio": 330.78855898820456,
|
494 |
+
"eval_steps_per_second": 0.304,
|
495 |
+
"step": 6000
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 0.37850583271283195,
|
499 |
+
"grad_norm": 13.973624229431152,
|
500 |
+
"learning_rate": 6.918852498391063e-05,
|
501 |
+
"loss": 0.2343,
|
502 |
+
"step": 6100
|
503 |
+
},
|
504 |
+
{
|
505 |
+
"epoch": 0.384710846363862,
|
506 |
+
"grad_norm": 14.253776550292969,
|
507 |
+
"learning_rate": 6.827922352936886e-05,
|
508 |
+
"loss": 0.2131,
|
509 |
+
"step": 6200
|
510 |
+
},
|
511 |
+
{
|
512 |
+
"epoch": 0.39091586001489204,
|
513 |
+
"grad_norm": 22.384563446044922,
|
514 |
+
"learning_rate": 6.736288915860883e-05,
|
515 |
+
"loss": 0.25,
|
516 |
+
"step": 6300
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"epoch": 0.3971208736659221,
|
520 |
+
"grad_norm": 14.788619041442871,
|
521 |
+
"learning_rate": 6.643987443052595e-05,
|
522 |
+
"loss": 0.2296,
|
523 |
+
"step": 6400
|
524 |
+
},
|
525 |
+
{
|
526 |
+
"epoch": 0.40332588731695207,
|
527 |
+
"grad_norm": 14.269457817077637,
|
528 |
+
"learning_rate": 6.551053447427773e-05,
|
529 |
+
"loss": 0.2208,
|
530 |
+
"step": 6500
|
531 |
+
},
|
532 |
+
{
|
533 |
+
"epoch": 0.4095309009679821,
|
534 |
+
"grad_norm": 14.065217971801758,
|
535 |
+
"learning_rate": 6.457522685264792e-05,
|
536 |
+
"loss": 0.2262,
|
537 |
+
"step": 6600
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 0.41573591461901216,
|
541 |
+
"grad_norm": 16.55682945251465,
|
542 |
+
"learning_rate": 6.363431142447469e-05,
|
543 |
+
"loss": 0.233,
|
544 |
+
"step": 6700
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"epoch": 0.4219409282700422,
|
548 |
+
"grad_norm": 11.659428596496582,
|
549 |
+
"learning_rate": 6.268815020619509e-05,
|
550 |
+
"loss": 0.197,
|
551 |
+
"step": 6800
|
552 |
+
},
|
553 |
+
{
|
554 |
+
"epoch": 0.42814594192107225,
|
555 |
+
"grad_norm": 20.64366912841797,
|
556 |
+
"learning_rate": 6.173710723255966e-05,
|
557 |
+
"loss": 0.2127,
|
558 |
+
"step": 6900
|
559 |
+
},
|
560 |
+
{
|
561 |
+
"epoch": 0.43435095557210224,
|
562 |
+
"grad_norm": 13.794021606445312,
|
563 |
+
"learning_rate": 6.078154841657027e-05,
|
564 |
+
"loss": 0.2114,
|
565 |
+
"step": 7000
|
566 |
+
},
|
567 |
+
{
|
568 |
+
"epoch": 0.43435095557210224,
|
569 |
+
"eval_avg_non_pair_similarity": 0.0004388350229078105,
|
570 |
+
"eval_avg_pair_similarity": 0.010316268266644328,
|
571 |
+
"eval_loss": 0.200907900929451,
|
572 |
+
"eval_runtime": 52.7763,
|
573 |
+
"eval_samples_per_second": 9.474,
|
574 |
+
"eval_similarity_ratio": 23.508306602984025,
|
575 |
+
"eval_steps_per_second": 0.303,
|
576 |
+
"step": 7000
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.4405559692231323,
|
580 |
+
"grad_norm": 13.422426223754883,
|
581 |
+
"learning_rate": 5.982184140869539e-05,
|
582 |
+
"loss": 0.1997,
|
583 |
+
"step": 7100
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.4467609828741623,
|
587 |
+
"grad_norm": 20.181900024414062,
|
588 |
+
"learning_rate": 5.885835545541699e-05,
|
589 |
+
"loss": 0.1893,
|
590 |
+
"step": 7200
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.45296599652519237,
|
594 |
+
"grad_norm": 18.431699752807617,
|
595 |
+
"learning_rate": 5.7891461257163105e-05,
|
596 |
+
"loss": 0.2062,
|
597 |
+
"step": 7300
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.45917101017622236,
|
601 |
+
"grad_norm": 16.194856643676758,
|
602 |
+
"learning_rate": 5.6921530825681255e-05,
|
603 |
+
"loss": 0.2087,
|
604 |
+
"step": 7400
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.4653760238272524,
|
608 |
+
"grad_norm": 18.078935623168945,
|
609 |
+
"learning_rate": 5.5948937340907156e-05,
|
610 |
+
"loss": 0.1978,
|
611 |
+
"step": 7500
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.47158103747828245,
|
615 |
+
"grad_norm": 20.49361801147461,
|
616 |
+
"learning_rate": 5.497405500738417e-05,
|
617 |
+
"loss": 0.1883,
|
618 |
+
"step": 7600
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.4777860511293125,
|
622 |
+
"grad_norm": 14.588258743286133,
|
623 |
+
"learning_rate": 5.399725891028824e-05,
|
624 |
+
"loss": 0.1829,
|
625 |
+
"step": 7700
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.48399106478034254,
|
629 |
+
"grad_norm": 14.363248825073242,
|
630 |
+
"learning_rate": 5.3018924871114305e-05,
|
631 |
+
"loss": 0.211,
|
632 |
+
"step": 7800
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.49019607843137253,
|
636 |
+
"grad_norm": 12.266989707946777,
|
637 |
+
"learning_rate": 5.203942930307929e-05,
|
638 |
+
"loss": 0.1853,
|
639 |
+
"step": 7900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.4964010920824026,
|
643 |
+
"grad_norm": 13.93957805633545,
|
644 |
+
"learning_rate": 5.105914906629734e-05,
|
645 |
+
"loss": 0.1616,
|
646 |
+
"step": 8000
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.4964010920824026,
|
650 |
+
"eval_avg_non_pair_similarity": 0.00028632804341306294,
|
651 |
+
"eval_avg_pair_similarity": 0.009248366304673255,
|
652 |
+
"eval_loss": 0.17835575342178345,
|
653 |
+
"eval_runtime": 52.8775,
|
654 |
+
"eval_samples_per_second": 9.456,
|
655 |
+
"eval_similarity_ratio": 32.299896979812644,
|
656 |
+
"eval_steps_per_second": 0.303,
|
657 |
+
"step": 8000
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 0.5026061057334326,
|
661 |
+
"grad_norm": 12.441362380981445,
|
662 |
+
"learning_rate": 5.007846132278327e-05,
|
663 |
+
"loss": 0.1687,
|
664 |
+
"step": 8100
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 0.5088111193844627,
|
668 |
+
"grad_norm": 17.834318161010742,
|
669 |
+
"learning_rate": 4.90977433913398e-05,
|
670 |
+
"loss": 0.1674,
|
671 |
+
"step": 8200
|
672 |
+
},
|
673 |
+
{
|
674 |
+
"epoch": 0.5150161330354927,
|
675 |
+
"grad_norm": 15.299223899841309,
|
676 |
+
"learning_rate": 4.8117372602384405e-05,
|
677 |
+
"loss": 0.1649,
|
678 |
+
"step": 8300
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 0.5212211466865228,
|
682 |
+
"grad_norm": 13.506607055664062,
|
683 |
+
"learning_rate": 4.713772615277178e-05,
|
684 |
+
"loss": 0.1833,
|
685 |
+
"step": 8400
|
686 |
+
},
|
687 |
+
{
|
688 |
+
"epoch": 0.5274261603375527,
|
689 |
+
"grad_norm": 12.79785442352295,
|
690 |
+
"learning_rate": 4.615918096066766e-05,
|
691 |
+
"loss": 0.1669,
|
692 |
+
"step": 8500
|
693 |
+
},
|
694 |
+
{
|
695 |
+
"epoch": 0.5336311739885827,
|
696 |
+
"grad_norm": 18.04769515991211,
|
697 |
+
"learning_rate": 4.518211352052985e-05,
|
698 |
+
"loss": 0.1611,
|
699 |
+
"step": 8600
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 0.5398361876396128,
|
703 |
+
"grad_norm": 14.453876495361328,
|
704 |
+
"learning_rate": 4.420689975825233e-05,
|
705 |
+
"loss": 0.1657,
|
706 |
+
"step": 8700
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"epoch": 0.5460412012906428,
|
710 |
+
"grad_norm": 15.206610679626465,
|
711 |
+
"learning_rate": 4.323391488652798e-05,
|
712 |
+
"loss": 0.1776,
|
713 |
+
"step": 8800
|
714 |
+
},
|
715 |
+
{
|
716 |
+
"epoch": 0.5522462149416729,
|
717 |
+
"grad_norm": 13.860987663269043,
|
718 |
+
"learning_rate": 4.226353326048593e-05,
|
719 |
+
"loss": 0.1491,
|
720 |
+
"step": 8900
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 0.5584512285927029,
|
724 |
+
"grad_norm": 13.18816089630127,
|
725 |
+
"learning_rate": 4.12961282336587e-05,
|
726 |
+
"loss": 0.1426,
|
727 |
+
"step": 9000
|
728 |
+
},
|
729 |
+
{
|
730 |
+
"epoch": 0.5584512285927029,
|
731 |
+
"eval_avg_non_pair_similarity": 0.00016228582146754844,
|
732 |
+
"eval_avg_pair_similarity": 0.00711837838916108,
|
733 |
+
"eval_loss": 0.16373609006404877,
|
734 |
+
"eval_runtime": 52.9511,
|
735 |
+
"eval_samples_per_second": 9.443,
|
736 |
+
"eval_similarity_ratio": 43.86321814678376,
|
737 |
+
"eval_steps_per_second": 0.302,
|
738 |
+
"step": 9000
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 0.5646562422437329,
|
742 |
+
"grad_norm": 12.775832176208496,
|
743 |
+
"learning_rate": 4.0332072014334626e-05,
|
744 |
+
"loss": 0.1713,
|
745 |
+
"step": 9100
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 0.570861255894763,
|
749 |
+
"grad_norm": 15.302083015441895,
|
750 |
+
"learning_rate": 3.937173552235117e-05,
|
751 |
+
"loss": 0.1613,
|
752 |
+
"step": 9200
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 0.577066269545793,
|
756 |
+
"grad_norm": 11.7548189163208,
|
757 |
+
"learning_rate": 3.841548824638376e-05,
|
758 |
+
"loss": 0.1684,
|
759 |
+
"step": 9300
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.5832712831968231,
|
763 |
+
"grad_norm": 10.636964797973633,
|
764 |
+
"learning_rate": 3.7463698101785286e-05,
|
765 |
+
"loss": 0.1539,
|
766 |
+
"step": 9400
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 0.5894762968478531,
|
770 |
+
"grad_norm": 11.79915714263916,
|
771 |
+
"learning_rate": 3.6516731289031054e-05,
|
772 |
+
"loss": 0.1624,
|
773 |
+
"step": 9500
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 0.5956813104988831,
|
777 |
+
"grad_norm": 14.726311683654785,
|
778 |
+
"learning_rate": 3.557495215282342e-05,
|
779 |
+
"loss": 0.1534,
|
780 |
+
"step": 9600
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 0.6018863241499132,
|
784 |
+
"grad_norm": 13.101693153381348,
|
785 |
+
"learning_rate": 3.463872304191036e-05,
|
786 |
+
"loss": 0.1591,
|
787 |
+
"step": 9700
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 0.6080913378009432,
|
791 |
+
"grad_norm": 10.77004337310791,
|
792 |
+
"learning_rate": 3.3708404169672316e-05,
|
793 |
+
"loss": 0.1523,
|
794 |
+
"step": 9800
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 0.6142963514519731,
|
798 |
+
"grad_norm": 11.254857063293457,
|
799 |
+
"learning_rate": 3.278435347553014e-05,
|
800 |
+
"loss": 0.1494,
|
801 |
+
"step": 9900
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 0.6205013651030032,
|
805 |
+
"grad_norm": 12.389632225036621,
|
806 |
+
"learning_rate": 3.18669264872284e-05,
|
807 |
+
"loss": 0.1438,
|
808 |
+
"step": 10000
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 0.6205013651030032,
|
812 |
+
"eval_avg_non_pair_similarity": 0.00022872556135607118,
|
813 |
+
"eval_avg_pair_similarity": 0.008581261232029647,
|
814 |
+
"eval_loss": 0.17591074109077454,
|
815 |
+
"eval_runtime": 52.6015,
|
816 |
+
"eval_samples_per_second": 9.505,
|
817 |
+
"eval_similarity_ratio": 37.51771853199507,
|
818 |
+
"eval_steps_per_second": 0.304,
|
819 |
+
"step": 10000
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 0.6267063787540332,
|
823 |
+
"grad_norm": 13.049078941345215,
|
824 |
+
"learning_rate": 3.0956476184046275e-05,
|
825 |
+
"loss": 0.1395,
|
826 |
+
"step": 10100
|
827 |
+
},
|
828 |
+
{
|
829 |
+
"epoch": 0.6329113924050633,
|
830 |
+
"grad_norm": 14.119858741760254,
|
831 |
+
"learning_rate": 3.0053352860989036e-05,
|
832 |
+
"loss": 0.1404,
|
833 |
+
"step": 10200
|
834 |
+
},
|
835 |
+
{
|
836 |
+
"epoch": 0.6391164060560933,
|
837 |
+
"grad_norm": 11.724855422973633,
|
838 |
+
"learning_rate": 2.9157903994012448e-05,
|
839 |
+
"loss": 0.1475,
|
840 |
+
"step": 10300
|
841 |
+
},
|
842 |
+
{
|
843 |
+
"epoch": 0.6453214197071233,
|
844 |
+
"grad_norm": 16.469905853271484,
|
845 |
+
"learning_rate": 2.827047410633156e-05,
|
846 |
+
"loss": 0.1403,
|
847 |
+
"step": 10400
|
848 |
+
},
|
849 |
+
{
|
850 |
+
"epoch": 0.6515264333581534,
|
851 |
+
"grad_norm": 11.535907745361328,
|
852 |
+
"learning_rate": 2.7391404635865724e-05,
|
853 |
+
"loss": 0.1324,
|
854 |
+
"step": 10500
|
855 |
+
},
|
856 |
+
{
|
857 |
+
"epoch": 0.6577314470091834,
|
858 |
+
"grad_norm": 10.59726619720459,
|
859 |
+
"learning_rate": 2.6521033803870692e-05,
|
860 |
+
"loss": 0.1212,
|
861 |
+
"step": 10600
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 0.6639364606602135,
|
865 |
+
"grad_norm": 20.87034034729004,
|
866 |
+
"learning_rate": 2.565969648480818e-05,
|
867 |
+
"loss": 0.1552,
|
868 |
+
"step": 10700
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"epoch": 0.6701414743112435,
|
872 |
+
"grad_norm": 9.909876823425293,
|
873 |
+
"learning_rate": 2.480772407750327e-05,
|
874 |
+
"loss": 0.1367,
|
875 |
+
"step": 10800
|
876 |
+
},
|
877 |
+
{
|
878 |
+
"epoch": 0.6763464879622735,
|
879 |
+
"grad_norm": 13.353315353393555,
|
880 |
+
"learning_rate": 2.3965444377638906e-05,
|
881 |
+
"loss": 0.1447,
|
882 |
+
"step": 10900
|
883 |
+
},
|
884 |
+
{
|
885 |
+
"epoch": 0.6825515016133036,
|
886 |
+
"grad_norm": 11.409924507141113,
|
887 |
+
"learning_rate": 2.3133181451636858e-05,
|
888 |
+
"loss": 0.1351,
|
889 |
+
"step": 11000
|
890 |
+
},
|
891 |
+
{
|
892 |
+
"epoch": 0.6825515016133036,
|
893 |
+
"eval_avg_non_pair_similarity": 2.0496868876971767e-05,
|
894 |
+
"eval_avg_pair_similarity": 0.007564989160746336,
|
895 |
+
"eval_loss": 0.1368183046579361,
|
896 |
+
"eval_runtime": 52.8241,
|
897 |
+
"eval_samples_per_second": 9.465,
|
898 |
+
"eval_similarity_ratio": 369.08023396908203,
|
899 |
+
"eval_steps_per_second": 0.303,
|
900 |
+
"step": 11000
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"epoch": 0.6887565152643336,
|
904 |
+
"grad_norm": 12.820631980895996,
|
905 |
+
"learning_rate": 2.2311255511973345e-05,
|
906 |
+
"loss": 0.1467,
|
907 |
+
"step": 11100
|
908 |
+
},
|
909 |
+
{
|
910 |
+
"epoch": 0.6949615289153637,
|
911 |
+
"grad_norm": 13.379746437072754,
|
912 |
+
"learning_rate": 2.1499982793977624e-05,
|
913 |
+
"loss": 0.1222,
|
914 |
+
"step": 11200
|
915 |
+
},
|
916 |
+
{
|
917 |
+
"epoch": 0.7011665425663937,
|
918 |
+
"grad_norm": 12.680821418762207,
|
919 |
+
"learning_rate": 2.06996754341607e-05,
|
920 |
+
"loss": 0.1309,
|
921 |
+
"step": 11300
|
922 |
+
},
|
923 |
+
{
|
924 |
+
"epoch": 0.7073715562174236,
|
925 |
+
"grad_norm": 12.36201286315918,
|
926 |
+
"learning_rate": 1.9910641350120968e-05,
|
927 |
+
"loss": 0.1231,
|
928 |
+
"step": 11400
|
929 |
+
},
|
930 |
+
{
|
931 |
+
"epoch": 0.7135765698684537,
|
932 |
+
"grad_norm": 13.760623931884766,
|
933 |
+
"learning_rate": 1.9133184122073305e-05,
|
934 |
+
"loss": 0.122,
|
935 |
+
"step": 11500
|
936 |
+
},
|
937 |
+
{
|
938 |
+
"epoch": 0.7197815835194837,
|
939 |
+
"grad_norm": 12.678217887878418,
|
940 |
+
"learning_rate": 1.836760287604676e-05,
|
941 |
+
"loss": 0.134,
|
942 |
+
"step": 11600
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 0.7259865971705137,
|
946 |
+
"grad_norm": 15.280141830444336,
|
947 |
+
"learning_rate": 1.761419216879601e-05,
|
948 |
+
"loss": 0.1407,
|
949 |
+
"step": 11700
|
950 |
+
},
|
951 |
+
{
|
952 |
+
"epoch": 0.7321916108215438,
|
953 |
+
"grad_norm": 11.139805793762207,
|
954 |
+
"learning_rate": 1.687324187447102e-05,
|
955 |
+
"loss": 0.1026,
|
956 |
+
"step": 11800
|
957 |
+
},
|
958 |
+
{
|
959 |
+
"epoch": 0.7383966244725738,
|
960 |
+
"grad_norm": 14.420808792114258,
|
961 |
+
"learning_rate": 1.6145037073088108e-05,
|
962 |
+
"loss": 0.1226,
|
963 |
+
"step": 11900
|
964 |
+
},
|
965 |
+
{
|
966 |
+
"epoch": 0.7446016381236039,
|
967 |
+
"grad_norm": 10.728633880615234,
|
968 |
+
"learning_rate": 1.5429857940845633e-05,
|
969 |
+
"loss": 0.123,
|
970 |
+
"step": 12000
|
971 |
+
},
|
972 |
+
{
|
973 |
+
"epoch": 0.7446016381236039,
|
974 |
+
"eval_avg_non_pair_similarity": 9.602570107098453e-05,
|
975 |
+
"eval_avg_pair_similarity": 0.007442979822866618,
|
976 |
+
"eval_loss": 0.13972054421901703,
|
977 |
+
"eval_runtime": 52.699,
|
978 |
+
"eval_samples_per_second": 9.488,
|
979 |
+
"eval_similarity_ratio": 77.510288806583,
|
980 |
+
"eval_steps_per_second": 0.304,
|
981 |
+
"step": 12000
|
982 |
+
},
|
983 |
+
{
|
984 |
+
"epoch": 0.7508066517746339,
|
985 |
+
"grad_norm": 13.222859382629395,
|
986 |
+
"learning_rate": 1.47279796423265e-05,
|
987 |
+
"loss": 0.1267,
|
988 |
+
"step": 12100
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"epoch": 0.7570116654256639,
|
992 |
+
"grad_norm": 14.91487979888916,
|
993 |
+
"learning_rate": 1.4039672224628786e-05,
|
994 |
+
"loss": 0.1176,
|
995 |
+
"step": 12200
|
996 |
+
},
|
997 |
+
{
|
998 |
+
"epoch": 0.763216679076694,
|
999 |
+
"grad_norm": 13.453108787536621,
|
1000 |
+
"learning_rate": 1.3365200513465293e-05,
|
1001 |
+
"loss": 0.1263,
|
1002 |
+
"step": 12300
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"epoch": 0.769421692727724,
|
1006 |
+
"grad_norm": 13.018562316894531,
|
1007 |
+
"learning_rate": 1.270482401127216e-05,
|
1008 |
+
"loss": 0.1042,
|
1009 |
+
"step": 12400
|
1010 |
+
},
|
1011 |
+
{
|
1012 |
+
"epoch": 0.7756267063787541,
|
1013 |
+
"grad_norm": 16.143644332885742,
|
1014 |
+
"learning_rate": 1.2058796797365423e-05,
|
1015 |
+
"loss": 0.1267,
|
1016 |
+
"step": 12500
|
1017 |
+
},
|
1018 |
+
{
|
1019 |
+
"epoch": 0.7818317200297841,
|
1020 |
+
"grad_norm": 13.684396743774414,
|
1021 |
+
"learning_rate": 1.1427367430184293e-05,
|
1022 |
+
"loss": 0.135,
|
1023 |
+
"step": 12600
|
1024 |
+
},
|
1025 |
+
{
|
1026 |
+
"epoch": 0.7880367336808141,
|
1027 |
+
"grad_norm": 12.124221801757812,
|
1028 |
+
"learning_rate": 1.0810778851658438e-05,
|
1029 |
+
"loss": 0.1181,
|
1030 |
+
"step": 12700
|
1031 |
+
},
|
1032 |
+
{
|
1033 |
+
"epoch": 0.7942417473318442,
|
1034 |
+
"grad_norm": 11.37338924407959,
|
1035 |
+
"learning_rate": 1.0209268293736308e-05,
|
1036 |
+
"loss": 0.1112,
|
1037 |
+
"step": 12800
|
1038 |
+
},
|
1039 |
+
{
|
1040 |
+
"epoch": 0.8004467609828742,
|
1041 |
+
"grad_norm": 13.746869087219238,
|
1042 |
+
"learning_rate": 9.623067187110307e-06,
|
1043 |
+
"loss": 0.1407,
|
1044 |
+
"step": 12900
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 0.8066517746339041,
|
1048 |
+
"grad_norm": 13.340392112731934,
|
1049 |
+
"learning_rate": 9.052401072174088e-06,
|
1050 |
+
"loss": 0.1013,
|
1051 |
+
"step": 13000
|
1052 |
+
},
|
1053 |
+
{
|
1054 |
+
"epoch": 0.8066517746339041,
|
1055 |
+
"eval_avg_non_pair_similarity": 0.00011364468113511786,
|
1056 |
+
"eval_avg_pair_similarity": 0.006324736045207828,
|
1057 |
+
"eval_loss": 0.12133920192718506,
|
1058 |
+
"eval_runtime": 52.921,
|
1059 |
+
"eval_samples_per_second": 9.448,
|
1060 |
+
"eval_similarity_ratio": 55.65360368857063,
|
1061 |
+
"eval_steps_per_second": 0.302,
|
1062 |
+
"step": 13000
|
1063 |
+
},
|
1064 |
+
{
|
1065 |
+
"epoch": 0.8128567882849342,
|
1066 |
+
"grad_norm": 16.114587783813477,
|
1067 |
+
"learning_rate": 8.497489512245971e-06,
|
1068 |
+
"loss": 0.1231,
|
1069 |
+
"step": 13100
|
1070 |
+
},
|
1071 |
+
{
|
1072 |
+
"epoch": 0.8190618019359642,
|
1073 |
+
"grad_norm": 12.396312713623047,
|
1074 |
+
"learning_rate": 7.958546009092232e-06,
|
1075 |
+
"loss": 0.1194,
|
1076 |
+
"step": 13200
|
1077 |
+
},
|
1078 |
+
{
|
1079 |
+
"epoch": 0.8252668155869943,
|
1080 |
+
"grad_norm": 13.97695541381836,
|
1081 |
+
"learning_rate": 7.435777920782444e-06,
|
1082 |
+
"loss": 0.1055,
|
1083 |
+
"step": 13300
|
1084 |
+
},
|
1085 |
+
{
|
1086 |
+
"epoch": 0.8314718292380243,
|
1087 |
+
"grad_norm": 13.231639862060547,
|
1088 |
+
"learning_rate": 6.9293863819085865e-06,
|
1089 |
+
"loss": 0.1056,
|
1090 |
+
"step": 13400
|
1091 |
+
},
|
1092 |
+
{
|
1093 |
+
"epoch": 0.8376768428890543,
|
1094 |
+
"grad_norm": 14.493330955505371,
|
1095 |
+
"learning_rate": 6.43956622619874e-06,
|
1096 |
+
"loss": 0.1165,
|
1097 |
+
"step": 13500
|
1098 |
+
},
|
1099 |
+
{
|
1100 |
+
"epoch": 0.8438818565400844,
|
1101 |
+
"grad_norm": 14.159026145935059,
|
1102 |
+
"learning_rate": 5.9665059115549535e-06,
|
1103 |
+
"loss": 0.1158,
|
1104 |
+
"step": 13600
|
1105 |
+
},
|
1106 |
+
{
|
1107 |
+
"epoch": 0.8500868701911144,
|
1108 |
+
"grad_norm": 16.65438461303711,
|
1109 |
+
"learning_rate": 5.510387447544168e-06,
|
1110 |
+
"loss": 0.112,
|
1111 |
+
"step": 13700
|
1112 |
+
},
|
1113 |
+
{
|
1114 |
+
"epoch": 0.8562918838421445,
|
1115 |
+
"grad_norm": 10.922910690307617,
|
1116 |
+
"learning_rate": 5.07138632537027e-06,
|
1117 |
+
"loss": 0.1129,
|
1118 |
+
"step": 13800
|
1119 |
+
},
|
1120 |
+
{
|
1121 |
+
"epoch": 0.8624968974931745,
|
1122 |
+
"grad_norm": 11.508538246154785,
|
1123 |
+
"learning_rate": 4.649671450353948e-06,
|
1124 |
+
"loss": 0.1079,
|
1125 |
+
"step": 13900
|
1126 |
+
},
|
1127 |
+
{
|
1128 |
+
"epoch": 0.8687019111442045,
|
1129 |
+
"grad_norm": 13.64457893371582,
|
1130 |
+
"learning_rate": 4.245405076946479e-06,
|
1131 |
+
"loss": 0.0866,
|
1132 |
+
"step": 14000
|
1133 |
+
},
|
1134 |
+
{
|
1135 |
+
"epoch": 0.8687019111442045,
|
1136 |
+
"eval_avg_non_pair_similarity": 0.00020547029043721383,
|
1137 |
+
"eval_avg_pair_similarity": 0.00815738959936425,
|
1138 |
+
"eval_loss": 0.1222657635807991,
|
1139 |
+
"eval_runtime": 52.7731,
|
1140 |
+
"eval_samples_per_second": 9.475,
|
1141 |
+
"eval_similarity_ratio": 39.70106618337082,
|
1142 |
+
"eval_steps_per_second": 0.303,
|
1143 |
+
"step": 14000
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 0.8749069247952346,
|
1147 |
+
"grad_norm": 15.069809913635254,
|
1148 |
+
"learning_rate": 3.858742746302535e-06,
|
1149 |
+
"loss": 0.1057,
|
1150 |
+
"step": 14100
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 0.8811119384462646,
|
1154 |
+
"grad_norm": 12.91094970703125,
|
1155 |
+
"learning_rate": 3.489833226435818e-06,
|
1156 |
+
"loss": 0.1103,
|
1157 |
+
"step": 14200
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 0.8873169520972947,
|
1161 |
+
"grad_norm": 12.701448440551758,
|
1162 |
+
"learning_rate": 3.138818454980641e-06,
|
1163 |
+
"loss": 0.1108,
|
1164 |
+
"step": 14300
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 0.8935219657483247,
|
1168 |
+
"grad_norm": 12.398956298828125,
|
1169 |
+
"learning_rate": 2.8058334845816213e-06,
|
1170 |
+
"loss": 0.1111,
|
1171 |
+
"step": 14400
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 0.8997269793993546,
|
1175 |
+
"grad_norm": 11.217853546142578,
|
1176 |
+
"learning_rate": 2.4910064309322523e-06,
|
1177 |
+
"loss": 0.0992,
|
1178 |
+
"step": 14500
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 0.9059319930503847,
|
1182 |
+
"grad_norm": 13.966708183288574,
|
1183 |
+
"learning_rate": 2.194458423482465e-06,
|
1184 |
+
"loss": 0.1228,
|
1185 |
+
"step": 14600
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 0.9121370067014147,
|
1189 |
+
"grad_norm": 14.061014175415039,
|
1190 |
+
"learning_rate": 1.9163035588342315e-06,
|
1191 |
+
"loss": 0.1107,
|
1192 |
+
"step": 14700
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 0.9183420203524447,
|
1196 |
+
"grad_norm": 13.071681022644043,
|
1197 |
+
"learning_rate": 1.6566488568429294e-06,
|
1198 |
+
"loss": 0.1163,
|
1199 |
+
"step": 14800
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 0.9245470340034748,
|
1203 |
+
"grad_norm": 12.697420120239258,
|
1204 |
+
"learning_rate": 1.415594219441585e-06,
|
1205 |
+
"loss": 0.0843,
|
1206 |
+
"step": 14900
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 0.9307520476545048,
|
1210 |
+
"grad_norm": 10.601663589477539,
|
1211 |
+
"learning_rate": 1.1932323922036803e-06,
|
1212 |
+
"loss": 0.1068,
|
1213 |
+
"step": 15000
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 0.9307520476545048,
|
1217 |
+
"eval_avg_non_pair_similarity": 0.0001585147638006401,
|
1218 |
+
"eval_avg_pair_similarity": 0.007807339721824974,
|
1219 |
+
"eval_loss": 0.11420850455760956,
|
1220 |
+
"eval_runtime": 52.8258,
|
1221 |
+
"eval_samples_per_second": 9.465,
|
1222 |
+
"eval_similarity_ratio": 49.25307608346225,
|
1223 |
+
"eval_steps_per_second": 0.303,
|
1224 |
+
"step": 15000
|
1225 |
+
}
|
1226 |
+
],
|
1227 |
+
"logging_steps": 100,
|
1228 |
+
"max_steps": 16116,
|
1229 |
+
"num_input_tokens_seen": 0,
|
1230 |
+
"num_train_epochs": 1,
|
1231 |
+
"save_steps": 1000,
|
1232 |
+
"stateful_callbacks": {
|
1233 |
+
"TrainerControl": {
|
1234 |
+
"args": {
|
1235 |
+
"should_epoch_stop": false,
|
1236 |
+
"should_evaluate": false,
|
1237 |
+
"should_log": false,
|
1238 |
+
"should_save": true,
|
1239 |
+
"should_training_stop": false
|
1240 |
+
},
|
1241 |
+
"attributes": {}
|
1242 |
+
}
|
1243 |
+
},
|
1244 |
+
"total_flos": 0.0,
|
1245 |
+
"train_batch_size": 32,
|
1246 |
+
"trial_name": null,
|
1247 |
+
"trial_params": null
|
1248 |
+
}
|
checkpoint-15000/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:45f6f301aa93fd04cb88245f1be6b5d49fae9be8dd0abe1d8a346a6aadd21663
|
3 |
+
size 5112
|