alex2awesome
commited on
Commit
•
b17c8b8
1
Parent(s):
fec8243
- all_results.json +21 -0
- config.json +38 -0
- post-training eval_results.json +9 -0
- pre-training eval_results.json +8 -0
- prediction_output.jsonl +0 -0
- pytorch_model.bin +3 -0
- runs/Jan12_19-16-57_instance-6/1673551070.453638/events.out.tfevents.1673551070.instance-6.21467.1 +3 -0
- runs/Jan12_19-16-57_instance-6/events.out.tfevents.1673551070.instance-6.21467.0 +3 -0
- runs/Jan12_19-20-53_instance-6/1673551304.5341797/events.out.tfevents.1673551304.instance-6.24298.1 +3 -0
- runs/Jan12_19-20-53_instance-6/events.out.tfevents.1673551304.instance-6.24298.0 +3 -0
- runs/Jan12_19-23-48_instance-6/1673551572.4730556/events.out.tfevents.1673551572.instance-6.26426.1 +3 -0
- runs/Jan12_19-23-48_instance-6/events.out.tfevents.1673551572.instance-6.26426.0 +3 -0
- train_results.json +8 -0
- trainer_state.json +501 -0
- training_args.bin +3 -0
all_results.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 3.0,
|
3 |
+
"eval_e": 0.0,
|
4 |
+
"eval_f1": 0.0024405878282630906,
|
5 |
+
"eval_loss": 7.580255031585693,
|
6 |
+
"eval_runtime": 121.173,
|
7 |
+
"eval_samples": 3336,
|
8 |
+
"eval_samples_per_second": 27.531,
|
9 |
+
"eval_steps_per_second": 27.531,
|
10 |
+
"test_e": 0.48231414868105515,
|
11 |
+
"test_f1": 0.4307979810407354,
|
12 |
+
"test_loss": 4.15036678314209,
|
13 |
+
"test_runtime": 121.1908,
|
14 |
+
"test_samples_per_second": 27.527,
|
15 |
+
"test_steps_per_second": 27.527,
|
16 |
+
"train_loss": 1.6027517270257226,
|
17 |
+
"train_runtime": 3981.5918,
|
18 |
+
"train_samples": 9379,
|
19 |
+
"train_samples_per_second": 7.067,
|
20 |
+
"train_steps_per_second": 7.067
|
21 |
+
}
|
config.json
ADDED
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "google/bigbird-roberta-base",
|
3 |
+
"architectures": [
|
4 |
+
"QAModel"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"attention_type": "original_full",
|
8 |
+
"block_size": 64,
|
9 |
+
"bos_token_id": 1,
|
10 |
+
"classifier_dropout": null,
|
11 |
+
"eos_token_id": 2,
|
12 |
+
"freeze_layers": null,
|
13 |
+
"gradient_checkpointing": false,
|
14 |
+
"hidden_act": "gelu_new",
|
15 |
+
"hidden_dropout_prob": 0.1,
|
16 |
+
"hidden_size": 768,
|
17 |
+
"include_nones_as_positives": false,
|
18 |
+
"initializer_range": 0.02,
|
19 |
+
"intermediate_size": 3072,
|
20 |
+
"layer_norm_eps": 1e-12,
|
21 |
+
"loss_window": null,
|
22 |
+
"max_position_embeddings": 4096,
|
23 |
+
"model_type": "big_bird",
|
24 |
+
"num_attention_heads": 12,
|
25 |
+
"num_hidden_layers": 12,
|
26 |
+
"num_random_blocks": 3,
|
27 |
+
"pad_token_id": 0,
|
28 |
+
"position_embedding_type": "absolute",
|
29 |
+
"qa_head": {},
|
30 |
+
"rescale_embeddings": false,
|
31 |
+
"sep_token_id": 66,
|
32 |
+
"torch_dtype": "float32",
|
33 |
+
"transformers_version": "4.12.2",
|
34 |
+
"type_vocab_size": 2,
|
35 |
+
"use_bias": true,
|
36 |
+
"use_cache": true,
|
37 |
+
"vocab_size": 50358
|
38 |
+
}
|
post-training eval_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"eval_samples": 3336,
|
3 |
+
"test_e": 0.48231414868105515,
|
4 |
+
"test_f1": 0.4307979810407354,
|
5 |
+
"test_loss": 4.15036678314209,
|
6 |
+
"test_runtime": 121.1908,
|
7 |
+
"test_samples_per_second": 27.527,
|
8 |
+
"test_steps_per_second": 27.527
|
9 |
+
}
|
pre-training eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"eval_e": 0.0,
|
3 |
+
"eval_f1": 0.0024405878282630906,
|
4 |
+
"eval_loss": 7.580255031585693,
|
5 |
+
"eval_runtime": 121.173,
|
6 |
+
"eval_samples_per_second": 27.531,
|
7 |
+
"eval_steps_per_second": 27.531
|
8 |
+
}
|
prediction_output.jsonl
ADDED
The diff for this file is too large to render.
See raw diff
|
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:514059838c0cfa3ef94c4fe736619c873ff88f37ada6808e3e03757134c04a0a
|
3 |
+
size 509994093
|
runs/Jan12_19-16-57_instance-6/1673551070.453638/events.out.tfevents.1673551070.instance-6.21467.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2a62707baf54b67d351abb6d2cc801baf39123647161167a126622c4b9c58ed1
|
3 |
+
size 4630
|
runs/Jan12_19-16-57_instance-6/events.out.tfevents.1673551070.instance-6.21467.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:236ce707a140c2f935f504ae595c29e2abf0000e8db40907e8c06c80df7a0b14
|
3 |
+
size 4369
|
runs/Jan12_19-20-53_instance-6/1673551304.5341797/events.out.tfevents.1673551304.instance-6.24298.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8225dcc8d727d46e800b6f5584cf77d52d1929cc3fb96555a8c98c31333f3776
|
3 |
+
size 4630
|
runs/Jan12_19-20-53_instance-6/events.out.tfevents.1673551304.instance-6.24298.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:eaf69355bbf2d8fc0b4006c582519fba893a2c245ea5894c56e8b6a1865a3166
|
3 |
+
size 4212
|
runs/Jan12_19-23-48_instance-6/1673551572.4730556/events.out.tfevents.1673551572.instance-6.26426.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c024db8c451ec305f39ccce96ddf4ee8bd0bcb9fa2719f7bafcdc24e8828f8e7
|
3 |
+
size 4630
|
runs/Jan12_19-23-48_instance-6/events.out.tfevents.1673551572.instance-6.26426.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:233387f4affbe4f55bb7081d196f7182defa6f7069ab7af81e1097fc446e32f6
|
3 |
+
size 18232
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 3.0,
|
3 |
+
"train_loss": 1.6027517270257226,
|
4 |
+
"train_runtime": 3981.5918,
|
5 |
+
"train_samples": 9379,
|
6 |
+
"train_samples_per_second": 7.067,
|
7 |
+
"train_steps_per_second": 7.067
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,501 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.0,
|
5 |
+
"global_step": 28137,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.05,
|
12 |
+
"learning_rate": 4.91114902086221e-05,
|
13 |
+
"loss": 4.1307,
|
14 |
+
"step": 500
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.11,
|
18 |
+
"learning_rate": 4.82229804172442e-05,
|
19 |
+
"loss": 3.2328,
|
20 |
+
"step": 1000
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.16,
|
24 |
+
"learning_rate": 4.73344706258663e-05,
|
25 |
+
"loss": 2.9206,
|
26 |
+
"step": 1500
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.21,
|
30 |
+
"learning_rate": 4.64459608344884e-05,
|
31 |
+
"loss": 2.8462,
|
32 |
+
"step": 2000
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.21,
|
36 |
+
"eval_e": 0.23890887290167867,
|
37 |
+
"eval_f1": 0.23467707071366814,
|
38 |
+
"eval_loss": 5.362636089324951,
|
39 |
+
"eval_runtime": 121.7016,
|
40 |
+
"eval_samples_per_second": 27.411,
|
41 |
+
"eval_steps_per_second": 27.411,
|
42 |
+
"step": 2000
|
43 |
+
},
|
44 |
+
{
|
45 |
+
"epoch": 0.27,
|
46 |
+
"learning_rate": 4.55574510431105e-05,
|
47 |
+
"loss": 2.9308,
|
48 |
+
"step": 2500
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 0.32,
|
52 |
+
"learning_rate": 4.46689412517326e-05,
|
53 |
+
"loss": 2.8088,
|
54 |
+
"step": 3000
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 0.37,
|
58 |
+
"learning_rate": 4.378043146035469e-05,
|
59 |
+
"loss": 2.5237,
|
60 |
+
"step": 3500
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"epoch": 0.43,
|
64 |
+
"learning_rate": 4.2891921668976795e-05,
|
65 |
+
"loss": 2.5704,
|
66 |
+
"step": 4000
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.43,
|
70 |
+
"eval_e": 0.3579136690647482,
|
71 |
+
"eval_f1": 0.3288601709033254,
|
72 |
+
"eval_loss": 3.1302592754364014,
|
73 |
+
"eval_runtime": 121.4271,
|
74 |
+
"eval_samples_per_second": 27.473,
|
75 |
+
"eval_steps_per_second": 27.473,
|
76 |
+
"step": 4000
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 0.48,
|
80 |
+
"learning_rate": 4.200341187759889e-05,
|
81 |
+
"loss": 2.2721,
|
82 |
+
"step": 4500
|
83 |
+
},
|
84 |
+
{
|
85 |
+
"epoch": 0.53,
|
86 |
+
"learning_rate": 4.111490208622099e-05,
|
87 |
+
"loss": 2.5344,
|
88 |
+
"step": 5000
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 0.59,
|
92 |
+
"learning_rate": 4.022639229484309e-05,
|
93 |
+
"loss": 2.3429,
|
94 |
+
"step": 5500
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.64,
|
98 |
+
"learning_rate": 3.933788250346519e-05,
|
99 |
+
"loss": 2.3649,
|
100 |
+
"step": 6000
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.64,
|
104 |
+
"eval_e": 0.3761990407673861,
|
105 |
+
"eval_f1": 0.3487221168810336,
|
106 |
+
"eval_loss": 4.601644039154053,
|
107 |
+
"eval_runtime": 121.4022,
|
108 |
+
"eval_samples_per_second": 27.479,
|
109 |
+
"eval_steps_per_second": 27.479,
|
110 |
+
"step": 6000
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 0.69,
|
114 |
+
"learning_rate": 3.844937271208729e-05,
|
115 |
+
"loss": 2.125,
|
116 |
+
"step": 6500
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 0.75,
|
120 |
+
"learning_rate": 3.756086292070939e-05,
|
121 |
+
"loss": 2.241,
|
122 |
+
"step": 7000
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.8,
|
126 |
+
"learning_rate": 3.6672353129331485e-05,
|
127 |
+
"loss": 2.0385,
|
128 |
+
"step": 7500
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.85,
|
132 |
+
"learning_rate": 3.578384333795359e-05,
|
133 |
+
"loss": 2.3278,
|
134 |
+
"step": 8000
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.85,
|
138 |
+
"eval_e": 0.4094724220623501,
|
139 |
+
"eval_f1": 0.3706672296662482,
|
140 |
+
"eval_loss": 4.3857340812683105,
|
141 |
+
"eval_runtime": 121.5065,
|
142 |
+
"eval_samples_per_second": 27.455,
|
143 |
+
"eval_steps_per_second": 27.455,
|
144 |
+
"step": 8000
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 0.91,
|
148 |
+
"learning_rate": 3.4895333546575685e-05,
|
149 |
+
"loss": 2.2073,
|
150 |
+
"step": 8500
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.96,
|
154 |
+
"learning_rate": 3.400682375519779e-05,
|
155 |
+
"loss": 2.089,
|
156 |
+
"step": 9000
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.01,
|
160 |
+
"learning_rate": 3.311831396381988e-05,
|
161 |
+
"loss": 1.8698,
|
162 |
+
"step": 9500
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 1.07,
|
166 |
+
"learning_rate": 3.222980417244198e-05,
|
167 |
+
"loss": 1.4806,
|
168 |
+
"step": 10000
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"epoch": 1.07,
|
172 |
+
"eval_e": 0.39718225419664266,
|
173 |
+
"eval_f1": 0.3538201477224622,
|
174 |
+
"eval_loss": 4.4642109870910645,
|
175 |
+
"eval_runtime": 121.3475,
|
176 |
+
"eval_samples_per_second": 27.491,
|
177 |
+
"eval_steps_per_second": 27.491,
|
178 |
+
"step": 10000
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 1.12,
|
182 |
+
"learning_rate": 3.134129438106408e-05,
|
183 |
+
"loss": 1.381,
|
184 |
+
"step": 10500
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.17,
|
188 |
+
"learning_rate": 3.0452784589686178e-05,
|
189 |
+
"loss": 1.4917,
|
190 |
+
"step": 11000
|
191 |
+
},
|
192 |
+
{
|
193 |
+
"epoch": 1.23,
|
194 |
+
"learning_rate": 2.9564274798308278e-05,
|
195 |
+
"loss": 1.5309,
|
196 |
+
"step": 11500
|
197 |
+
},
|
198 |
+
{
|
199 |
+
"epoch": 1.28,
|
200 |
+
"learning_rate": 2.8675765006930378e-05,
|
201 |
+
"loss": 1.534,
|
202 |
+
"step": 12000
|
203 |
+
},
|
204 |
+
{
|
205 |
+
"epoch": 1.28,
|
206 |
+
"eval_e": 0.3983812949640288,
|
207 |
+
"eval_f1": 0.361609584220443,
|
208 |
+
"eval_loss": 4.218064785003662,
|
209 |
+
"eval_runtime": 121.4358,
|
210 |
+
"eval_samples_per_second": 27.471,
|
211 |
+
"eval_steps_per_second": 27.471,
|
212 |
+
"step": 12000
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.33,
|
216 |
+
"learning_rate": 2.7787255215552478e-05,
|
217 |
+
"loss": 1.4109,
|
218 |
+
"step": 12500
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 1.39,
|
222 |
+
"learning_rate": 2.6898745424174578e-05,
|
223 |
+
"loss": 1.562,
|
224 |
+
"step": 13000
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 1.44,
|
228 |
+
"learning_rate": 2.6010235632796674e-05,
|
229 |
+
"loss": 1.6473,
|
230 |
+
"step": 13500
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 1.49,
|
234 |
+
"learning_rate": 2.5121725841418774e-05,
|
235 |
+
"loss": 1.2592,
|
236 |
+
"step": 14000
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 1.49,
|
240 |
+
"eval_e": 0.4520383693045564,
|
241 |
+
"eval_f1": 0.40414986746385745,
|
242 |
+
"eval_loss": 4.98923397064209,
|
243 |
+
"eval_runtime": 121.4673,
|
244 |
+
"eval_samples_per_second": 27.464,
|
245 |
+
"eval_steps_per_second": 27.464,
|
246 |
+
"step": 14000
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 1.55,
|
250 |
+
"learning_rate": 2.423321605004087e-05,
|
251 |
+
"loss": 1.3429,
|
252 |
+
"step": 14500
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 1.6,
|
256 |
+
"learning_rate": 2.334470625866297e-05,
|
257 |
+
"loss": 1.4802,
|
258 |
+
"step": 15000
|
259 |
+
},
|
260 |
+
{
|
261 |
+
"epoch": 1.65,
|
262 |
+
"learning_rate": 2.245619646728507e-05,
|
263 |
+
"loss": 1.4203,
|
264 |
+
"step": 15500
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 1.71,
|
268 |
+
"learning_rate": 2.156768667590717e-05,
|
269 |
+
"loss": 1.2868,
|
270 |
+
"step": 16000
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"epoch": 1.71,
|
274 |
+
"eval_e": 0.4475419664268585,
|
275 |
+
"eval_f1": 0.40389977547135164,
|
276 |
+
"eval_loss": 4.433777809143066,
|
277 |
+
"eval_runtime": 121.4975,
|
278 |
+
"eval_samples_per_second": 27.457,
|
279 |
+
"eval_steps_per_second": 27.457,
|
280 |
+
"step": 16000
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"epoch": 1.76,
|
284 |
+
"learning_rate": 2.067917688452927e-05,
|
285 |
+
"loss": 1.3837,
|
286 |
+
"step": 16500
|
287 |
+
},
|
288 |
+
{
|
289 |
+
"epoch": 1.81,
|
290 |
+
"learning_rate": 1.9790667093151367e-05,
|
291 |
+
"loss": 1.3352,
|
292 |
+
"step": 17000
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"epoch": 1.87,
|
296 |
+
"learning_rate": 1.8902157301773464e-05,
|
297 |
+
"loss": 1.3861,
|
298 |
+
"step": 17500
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 1.92,
|
302 |
+
"learning_rate": 1.8013647510395564e-05,
|
303 |
+
"loss": 1.4105,
|
304 |
+
"step": 18000
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 1.92,
|
308 |
+
"eval_e": 0.4577338129496403,
|
309 |
+
"eval_f1": 0.408643956398686,
|
310 |
+
"eval_loss": 4.376325607299805,
|
311 |
+
"eval_runtime": 121.4273,
|
312 |
+
"eval_samples_per_second": 27.473,
|
313 |
+
"eval_steps_per_second": 27.473,
|
314 |
+
"step": 18000
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 1.97,
|
318 |
+
"learning_rate": 1.7125137719017664e-05,
|
319 |
+
"loss": 1.2234,
|
320 |
+
"step": 18500
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 2.03,
|
324 |
+
"learning_rate": 1.6236627927639764e-05,
|
325 |
+
"loss": 1.0492,
|
326 |
+
"step": 19000
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 2.08,
|
330 |
+
"learning_rate": 1.5348118136261864e-05,
|
331 |
+
"loss": 0.8572,
|
332 |
+
"step": 19500
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 2.13,
|
336 |
+
"learning_rate": 1.4459608344883962e-05,
|
337 |
+
"loss": 0.7944,
|
338 |
+
"step": 20000
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 2.13,
|
342 |
+
"eval_e": 0.473621103117506,
|
343 |
+
"eval_f1": 0.42407242585653226,
|
344 |
+
"eval_loss": 4.196831703186035,
|
345 |
+
"eval_runtime": 121.4007,
|
346 |
+
"eval_samples_per_second": 27.479,
|
347 |
+
"eval_steps_per_second": 27.479,
|
348 |
+
"step": 20000
|
349 |
+
},
|
350 |
+
{
|
351 |
+
"epoch": 2.19,
|
352 |
+
"learning_rate": 1.3571098553506059e-05,
|
353 |
+
"loss": 0.9145,
|
354 |
+
"step": 20500
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 2.24,
|
358 |
+
"learning_rate": 1.2682588762128159e-05,
|
359 |
+
"loss": 0.8874,
|
360 |
+
"step": 21000
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 2.29,
|
364 |
+
"learning_rate": 1.1794078970750259e-05,
|
365 |
+
"loss": 0.9073,
|
366 |
+
"step": 21500
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 2.35,
|
370 |
+
"learning_rate": 1.0905569179372357e-05,
|
371 |
+
"loss": 0.791,
|
372 |
+
"step": 22000
|
373 |
+
},
|
374 |
+
{
|
375 |
+
"epoch": 2.35,
|
376 |
+
"eval_e": 0.4697242206235012,
|
377 |
+
"eval_f1": 0.4240145974837222,
|
378 |
+
"eval_loss": 4.208236217498779,
|
379 |
+
"eval_runtime": 121.3326,
|
380 |
+
"eval_samples_per_second": 27.495,
|
381 |
+
"eval_steps_per_second": 27.495,
|
382 |
+
"step": 22000
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 2.4,
|
386 |
+
"learning_rate": 1.0017059387994457e-05,
|
387 |
+
"loss": 0.753,
|
388 |
+
"step": 22500
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 2.45,
|
392 |
+
"learning_rate": 9.128549596616555e-06,
|
393 |
+
"loss": 0.9797,
|
394 |
+
"step": 23000
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 2.51,
|
398 |
+
"learning_rate": 8.240039805238655e-06,
|
399 |
+
"loss": 0.8473,
|
400 |
+
"step": 23500
|
401 |
+
},
|
402 |
+
{
|
403 |
+
"epoch": 2.56,
|
404 |
+
"learning_rate": 7.351530013860753e-06,
|
405 |
+
"loss": 0.8997,
|
406 |
+
"step": 24000
|
407 |
+
},
|
408 |
+
{
|
409 |
+
"epoch": 2.56,
|
410 |
+
"eval_e": 0.4766187050359712,
|
411 |
+
"eval_f1": 0.42464439616671873,
|
412 |
+
"eval_loss": 4.4838666915893555,
|
413 |
+
"eval_runtime": 121.474,
|
414 |
+
"eval_samples_per_second": 27.463,
|
415 |
+
"eval_steps_per_second": 27.463,
|
416 |
+
"step": 24000
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 2.61,
|
420 |
+
"learning_rate": 6.463020222482852e-06,
|
421 |
+
"loss": 0.8912,
|
422 |
+
"step": 24500
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 2.67,
|
426 |
+
"learning_rate": 5.574510431104952e-06,
|
427 |
+
"loss": 0.7544,
|
428 |
+
"step": 25000
|
429 |
+
},
|
430 |
+
{
|
431 |
+
"epoch": 2.72,
|
432 |
+
"learning_rate": 4.68600063972705e-06,
|
433 |
+
"loss": 0.823,
|
434 |
+
"step": 25500
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 2.77,
|
438 |
+
"learning_rate": 3.7974908483491486e-06,
|
439 |
+
"loss": 0.835,
|
440 |
+
"step": 26000
|
441 |
+
},
|
442 |
+
{
|
443 |
+
"epoch": 2.77,
|
444 |
+
"eval_e": 0.48081534772182255,
|
445 |
+
"eval_f1": 0.4289052534622072,
|
446 |
+
"eval_loss": 3.9213685989379883,
|
447 |
+
"eval_runtime": 121.3974,
|
448 |
+
"eval_samples_per_second": 27.48,
|
449 |
+
"eval_steps_per_second": 27.48,
|
450 |
+
"step": 26000
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 2.83,
|
454 |
+
"learning_rate": 2.908981056971248e-06,
|
455 |
+
"loss": 0.8194,
|
456 |
+
"step": 26500
|
457 |
+
},
|
458 |
+
{
|
459 |
+
"epoch": 2.88,
|
460 |
+
"learning_rate": 2.020471265593347e-06,
|
461 |
+
"loss": 0.7346,
|
462 |
+
"step": 27000
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"epoch": 2.93,
|
466 |
+
"learning_rate": 1.131961474215446e-06,
|
467 |
+
"loss": 0.642,
|
468 |
+
"step": 27500
|
469 |
+
},
|
470 |
+
{
|
471 |
+
"epoch": 2.99,
|
472 |
+
"learning_rate": 2.434516828375449e-07,
|
473 |
+
"loss": 0.8905,
|
474 |
+
"step": 28000
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 2.99,
|
478 |
+
"eval_e": 0.48231414868105515,
|
479 |
+
"eval_f1": 0.4308855987601117,
|
480 |
+
"eval_loss": 4.153212070465088,
|
481 |
+
"eval_runtime": 121.3729,
|
482 |
+
"eval_samples_per_second": 27.486,
|
483 |
+
"eval_steps_per_second": 27.486,
|
484 |
+
"step": 28000
|
485 |
+
},
|
486 |
+
{
|
487 |
+
"epoch": 3.0,
|
488 |
+
"step": 28137,
|
489 |
+
"total_flos": 1.867929495867468e+16,
|
490 |
+
"train_loss": 1.6027517270257226,
|
491 |
+
"train_runtime": 3981.5918,
|
492 |
+
"train_samples_per_second": 7.067,
|
493 |
+
"train_steps_per_second": 7.067
|
494 |
+
}
|
495 |
+
],
|
496 |
+
"max_steps": 28137,
|
497 |
+
"num_train_epochs": 3,
|
498 |
+
"total_flos": 1.867929495867468e+16,
|
499 |
+
"trial_name": null,
|
500 |
+
"trial_params": null
|
501 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:de20ea039fc61092afb26d778cb37a099fccbfb961575e551b5cb0476fcf363c
|
3 |
+
size 2927
|