ShengdingHu
commited on
Commit
•
cac7f6a
1
Parent(s):
3ec4a2d
Training in progress, step 200
Browse files- all_results.json +16 -16
- eval_results.json +6 -6
- pytorch_model.bin +2 -2
- runs/Feb01_01-13-39_node1/events.out.tfevents.1643649325.node1 +2 -2
- runs/Feb01_01-13-39_node1/events.out.tfevents.1643650828.node1 +3 -0
- runs/Feb02_15-41-24_node1/1643787763.8152487/events.out.tfevents.1643787763.node1 +3 -0
- runs/Feb02_15-41-24_node1/events.out.tfevents.1643787763.node1 +3 -0
- test_results.json +6 -6
- train_results.json +4 -4
- trainer_state.json +120 -50
- training_args.bin +1 -1
- training_config.json +1 -1
all_results.json
CHANGED
@@ -1,20 +1,20 @@
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
-
"eval_average_metrics":
|
4 |
-
"eval_em":
|
5 |
-
"eval_f1":
|
6 |
-
"eval_loss": 0.
|
7 |
-
"eval_runtime":
|
8 |
-
"eval_samples_per_second":
|
9 |
-
"test_average_metrics":
|
10 |
-
"test_em":
|
11 |
-
"test_f1":
|
12 |
-
"test_loss": 0.
|
13 |
-
"test_runtime":
|
14 |
-
"test_samples_per_second":
|
15 |
-
"train_loss": 0.
|
16 |
-
"train_runtime":
|
17 |
"train_samples": 27243,
|
18 |
-
"train_samples_per_second":
|
19 |
-
"train_steps_per_second":
|
20 |
}
|
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
+
"eval_average_metrics": 62.27396685546466,
|
4 |
+
"eval_em": 52.6431718061674,
|
5 |
+
"eval_f1": 71.90476190476191,
|
6 |
+
"eval_loss": 0.22205480933189392,
|
7 |
+
"eval_runtime": 24.5121,
|
8 |
+
"eval_samples_per_second": 98.89,
|
9 |
+
"test_average_metrics": 63.46612466124662,
|
10 |
+
"test_em": 54.44444444444444,
|
11 |
+
"test_f1": 72.48780487804879,
|
12 |
+
"test_loss": 0.19974233210086823,
|
13 |
+
"test_runtime": 25.265,
|
14 |
+
"test_samples_per_second": 95.943,
|
15 |
+
"train_loss": 0.18877839930180654,
|
16 |
+
"train_runtime": 1478.3404,
|
17 |
"train_samples": 27243,
|
18 |
+
"train_samples_per_second": 55.284,
|
19 |
+
"train_steps_per_second": 1.729
|
20 |
}
|
eval_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
-
"eval_average_metrics":
|
4 |
-
"eval_em":
|
5 |
-
"eval_f1":
|
6 |
-
"eval_loss": 0.
|
7 |
-
"eval_runtime":
|
8 |
-
"eval_samples_per_second":
|
9 |
}
|
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
+
"eval_average_metrics": 62.27396685546466,
|
4 |
+
"eval_em": 52.6431718061674,
|
5 |
+
"eval_f1": 71.90476190476191,
|
6 |
+
"eval_loss": 0.22205480933189392,
|
7 |
+
"eval_runtime": 24.5121,
|
8 |
+
"eval_samples_per_second": 98.89
|
9 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f1ab444282eb590cc07dbabd31e47b50243e70ffccd042252568fb62a630f70c
|
3 |
+
size 2631685
|
runs/Feb01_01-13-39_node1/events.out.tfevents.1643649325.node1
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:32102cbbdcd9c4a795b0aa34ba5c29813b31321cf7915c06cf138ba7acdb7670
|
3 |
+
size 9821
|
runs/Feb01_01-13-39_node1/events.out.tfevents.1643650828.node1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:77e1993e71805ef73e081a4862b663ff96c1aa1575808c78da39cff8d9ac408d
|
3 |
+
size 764
|
runs/Feb02_15-41-24_node1/1643787763.8152487/events.out.tfevents.1643787763.node1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ef1270b5efe2e01fbeb1d60927c997ac38e79a588538ae80052edd4781dd4e59
|
3 |
+
size 5050
|
runs/Feb02_15-41-24_node1/events.out.tfevents.1643787763.node1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ddbeaefe73b2dfc4ea9ad1ad0465fab69d788e5c97f00a965dca9e048367f9af
|
3 |
+
size 4338
|
test_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
-
"test_average_metrics":
|
4 |
-
"test_em":
|
5 |
-
"test_f1":
|
6 |
-
"test_loss": 0.
|
7 |
-
"test_runtime":
|
8 |
-
"test_samples_per_second":
|
9 |
}
|
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
+
"test_average_metrics": 63.46612466124662,
|
4 |
+
"test_em": 54.44444444444444,
|
5 |
+
"test_f1": 72.48780487804879,
|
6 |
+
"test_loss": 0.19974233210086823,
|
7 |
+
"test_runtime": 25.265,
|
8 |
+
"test_samples_per_second": 95.943
|
9 |
}
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
-
"train_loss": 0.
|
4 |
-
"train_runtime":
|
5 |
"train_samples": 27243,
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second":
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 3.0,
|
3 |
+
"train_loss": 0.18877839930180654,
|
4 |
+
"train_runtime": 1478.3404,
|
5 |
"train_samples": 27243,
|
6 |
+
"train_samples_per_second": 55.284,
|
7 |
+
"train_steps_per_second": 1.729
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,105 +1,175 @@
|
|
1 |
{
|
2 |
-
"best_metric":
|
3 |
-
"best_model_checkpoint": "outputs/bitfit/t5-base/superglue-multirc/checkpoint-
|
4 |
"epoch": 3.0,
|
5 |
"global_step": 2556,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
{
|
11 |
"epoch": 0.59,
|
12 |
"learning_rate": 0.0002413145539906103,
|
13 |
-
"loss": 0.
|
14 |
"step": 500
|
15 |
},
|
16 |
{
|
17 |
-
"epoch": 0.
|
18 |
-
"eval_average_metrics":
|
19 |
-
"eval_em": 49.
|
20 |
-
"eval_f1":
|
21 |
-
"eval_loss": 0.
|
22 |
-
"eval_runtime":
|
23 |
-
"eval_samples_per_second":
|
24 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
},
|
26 |
{
|
27 |
"epoch": 1.17,
|
28 |
"learning_rate": 0.00018262910798122063,
|
29 |
-
"loss": 0.
|
30 |
"step": 1000
|
31 |
},
|
32 |
{
|
33 |
"epoch": 1.17,
|
34 |
-
"eval_average_metrics":
|
35 |
-
"eval_em":
|
36 |
-
"eval_f1":
|
37 |
-
"eval_loss": 0.
|
38 |
-
"eval_runtime":
|
39 |
-
"eval_samples_per_second":
|
40 |
"step": 1000
|
41 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
{
|
43 |
"epoch": 1.76,
|
44 |
"learning_rate": 0.00012394366197183098,
|
45 |
-
"loss": 0.
|
46 |
"step": 1500
|
47 |
},
|
48 |
{
|
49 |
-
"epoch": 1.
|
50 |
-
"eval_average_metrics":
|
51 |
-
"eval_em":
|
52 |
-
"eval_f1":
|
53 |
-
"eval_loss": 0.
|
54 |
-
"eval_runtime":
|
55 |
-
"eval_samples_per_second":
|
56 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
},
|
58 |
{
|
59 |
"epoch": 2.35,
|
60 |
"learning_rate": 6.525821596244131e-05,
|
61 |
-
"loss": 0.
|
62 |
"step": 2000
|
63 |
},
|
64 |
{
|
65 |
"epoch": 2.35,
|
66 |
-
"eval_average_metrics":
|
67 |
-
"eval_em":
|
68 |
-
"eval_f1":
|
69 |
-
"eval_loss": 0.
|
70 |
-
"eval_runtime":
|
71 |
-
"eval_samples_per_second":
|
72 |
"step": 2000
|
73 |
},
|
74 |
{
|
75 |
-
"epoch": 2.
|
76 |
-
"
|
77 |
-
"
|
78 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
79 |
},
|
80 |
{
|
81 |
"epoch": 2.93,
|
82 |
-
"
|
83 |
-
"
|
84 |
-
"eval_f1": 67.38351254480287,
|
85 |
-
"eval_loss": 0.21952344477176666,
|
86 |
-
"eval_runtime": 10.193,
|
87 |
-
"eval_samples_per_second": 237.809,
|
88 |
"step": 2500
|
89 |
},
|
90 |
{
|
91 |
"epoch": 3.0,
|
92 |
"step": 2556,
|
93 |
-
"total_flos": 2.
|
94 |
-
"train_loss": 0.
|
95 |
-
"train_runtime":
|
96 |
-
"train_samples_per_second":
|
97 |
-
"train_steps_per_second":
|
98 |
}
|
99 |
],
|
100 |
"max_steps": 2556,
|
101 |
"num_train_epochs": 3,
|
102 |
-
"total_flos": 2.
|
103 |
"trial_name": null,
|
104 |
"trial_params": null
|
105 |
}
|
|
|
1 |
{
|
2 |
+
"best_metric": 62.27396685546466,
|
3 |
+
"best_model_checkpoint": "outputs/bitfit/t5-base/superglue-multirc/checkpoint-2400",
|
4 |
"epoch": 3.0,
|
5 |
"global_step": 2556,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.23,
|
12 |
+
"eval_average_metrics": 58.45389071733288,
|
13 |
+
"eval_em": 49.88986784140969,
|
14 |
+
"eval_f1": 67.01791359325607,
|
15 |
+
"eval_loss": 0.21679049730300903,
|
16 |
+
"eval_runtime": 23.0565,
|
17 |
+
"eval_samples_per_second": 105.133,
|
18 |
+
"step": 200
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"epoch": 0.47,
|
22 |
+
"eval_average_metrics": 61.27063054507724,
|
23 |
+
"eval_em": 51.98237885462555,
|
24 |
+
"eval_f1": 70.55888223552893,
|
25 |
+
"eval_loss": 0.2198609709739685,
|
26 |
+
"eval_runtime": 23.4946,
|
27 |
+
"eval_samples_per_second": 103.173,
|
28 |
+
"step": 400
|
29 |
+
},
|
30 |
{
|
31 |
"epoch": 0.59,
|
32 |
"learning_rate": 0.0002413145539906103,
|
33 |
+
"loss": 0.2582,
|
34 |
"step": 500
|
35 |
},
|
36 |
{
|
37 |
+
"epoch": 0.7,
|
38 |
+
"eval_average_metrics": 59.16978670550705,
|
39 |
+
"eval_em": 49.779735682819386,
|
40 |
+
"eval_f1": 68.55983772819472,
|
41 |
+
"eval_loss": 0.20913197100162506,
|
42 |
+
"eval_runtime": 23.7161,
|
43 |
+
"eval_samples_per_second": 102.209,
|
44 |
+
"step": 600
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.94,
|
48 |
+
"eval_average_metrics": 61.270356888999785,
|
49 |
+
"eval_em": 51.43171806167401,
|
50 |
+
"eval_f1": 71.10899571632555,
|
51 |
+
"eval_loss": 0.20762182772159576,
|
52 |
+
"eval_runtime": 26.3745,
|
53 |
+
"eval_samples_per_second": 91.907,
|
54 |
+
"step": 800
|
55 |
},
|
56 |
{
|
57 |
"epoch": 1.17,
|
58 |
"learning_rate": 0.00018262910798122063,
|
59 |
+
"loss": 0.1865,
|
60 |
"step": 1000
|
61 |
},
|
62 |
{
|
63 |
"epoch": 1.17,
|
64 |
+
"eval_average_metrics": 61.45286279488806,
|
65 |
+
"eval_em": 51.87224669603524,
|
66 |
+
"eval_f1": 71.03347889374089,
|
67 |
+
"eval_loss": 0.2089201658964157,
|
68 |
+
"eval_runtime": 24.8833,
|
69 |
+
"eval_samples_per_second": 97.415,
|
70 |
"step": 1000
|
71 |
},
|
72 |
+
{
|
73 |
+
"epoch": 1.41,
|
74 |
+
"eval_average_metrics": 61.79826184816632,
|
75 |
+
"eval_em": 52.202643171806166,
|
76 |
+
"eval_f1": 71.39388052452648,
|
77 |
+
"eval_loss": 0.20319344103336334,
|
78 |
+
"eval_runtime": 25.4876,
|
79 |
+
"eval_samples_per_second": 95.105,
|
80 |
+
"step": 1200
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 1.64,
|
84 |
+
"eval_average_metrics": 62.13396621755072,
|
85 |
+
"eval_em": 52.53303964757709,
|
86 |
+
"eval_f1": 71.73489278752436,
|
87 |
+
"eval_loss": 0.21759574115276337,
|
88 |
+
"eval_runtime": 26.3526,
|
89 |
+
"eval_samples_per_second": 91.983,
|
90 |
+
"step": 1400
|
91 |
+
},
|
92 |
{
|
93 |
"epoch": 1.76,
|
94 |
"learning_rate": 0.00012394366197183098,
|
95 |
+
"loss": 0.1745,
|
96 |
"step": 1500
|
97 |
},
|
98 |
{
|
99 |
+
"epoch": 1.88,
|
100 |
+
"eval_average_metrics": 62.159955413411296,
|
101 |
+
"eval_em": 52.09251101321586,
|
102 |
+
"eval_f1": 72.22739981360672,
|
103 |
+
"eval_loss": 0.21236789226531982,
|
104 |
+
"eval_runtime": 22.8791,
|
105 |
+
"eval_samples_per_second": 105.948,
|
106 |
+
"step": 1600
|
107 |
+
},
|
108 |
+
{
|
109 |
+
"epoch": 2.11,
|
110 |
+
"eval_average_metrics": 61.48864179183695,
|
111 |
+
"eval_em": 51.651982378854626,
|
112 |
+
"eval_f1": 71.32530120481927,
|
113 |
+
"eval_loss": 0.2192794233560562,
|
114 |
+
"eval_runtime": 23.8863,
|
115 |
+
"eval_samples_per_second": 101.481,
|
116 |
+
"step": 1800
|
117 |
},
|
118 |
{
|
119 |
"epoch": 2.35,
|
120 |
"learning_rate": 6.525821596244131e-05,
|
121 |
+
"loss": 0.1692,
|
122 |
"step": 2000
|
123 |
},
|
124 |
{
|
125 |
"epoch": 2.35,
|
126 |
+
"eval_average_metrics": 62.06655934464116,
|
127 |
+
"eval_em": 52.75330396475771,
|
128 |
+
"eval_f1": 71.37981472452462,
|
129 |
+
"eval_loss": 0.2168177217245102,
|
130 |
+
"eval_runtime": 23.5217,
|
131 |
+
"eval_samples_per_second": 103.054,
|
132 |
"step": 2000
|
133 |
},
|
134 |
{
|
135 |
+
"epoch": 2.58,
|
136 |
+
"eval_average_metrics": 62.263646102553764,
|
137 |
+
"eval_em": 52.31277533039648,
|
138 |
+
"eval_f1": 72.21451687471105,
|
139 |
+
"eval_loss": 0.22297202050685883,
|
140 |
+
"eval_runtime": 22.3978,
|
141 |
+
"eval_samples_per_second": 108.225,
|
142 |
+
"step": 2200
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 2.82,
|
146 |
+
"eval_average_metrics": 62.27396685546466,
|
147 |
+
"eval_em": 52.6431718061674,
|
148 |
+
"eval_f1": 71.90476190476191,
|
149 |
+
"eval_loss": 0.22205480933189392,
|
150 |
+
"eval_runtime": 23.6577,
|
151 |
+
"eval_samples_per_second": 102.461,
|
152 |
+
"step": 2400
|
153 |
},
|
154 |
{
|
155 |
"epoch": 2.93,
|
156 |
+
"learning_rate": 6.5727699530516425e-06,
|
157 |
+
"loss": 0.1589,
|
|
|
|
|
|
|
|
|
158 |
"step": 2500
|
159 |
},
|
160 |
{
|
161 |
"epoch": 3.0,
|
162 |
"step": 2556,
|
163 |
+
"total_flos": 2.511166406964019e+16,
|
164 |
+
"train_loss": 0.18877839930180654,
|
165 |
+
"train_runtime": 1478.3404,
|
166 |
+
"train_samples_per_second": 55.284,
|
167 |
+
"train_steps_per_second": 1.729
|
168 |
}
|
169 |
],
|
170 |
"max_steps": 2556,
|
171 |
"num_train_epochs": 3,
|
172 |
+
"total_flos": 2.511166406964019e+16,
|
173 |
"trial_name": null,
|
174 |
"trial_params": null
|
175 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3183
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:83e35f5470e9729f322b58c3e0e205808f864a2886e2f0aa7c9a151f826d1f0f
|
3 |
size 3183
|
training_config.json
CHANGED
@@ -1 +1 @@
|
|
1 |
-
{"
|
|
|
1 |
+
{"dataset_config_name": ["en"], "delta_type": "lora", "do_eval": true, "do_test": true, "do_train": true, "eval_dataset_config_name": ["en"], "eval_dataset_name": "superglue-multirc", "eval_steps": 200, "evaluation_strategy": "steps", "greater_is_better": true, "learning_rate": 0.0003, "load_best_model_at_end": true, "lora_r": 8, "max_source_length": 256, "metric_for_best_model": "average_metrics", "model_name_or_path": "../../../../plm_cache/t5-base", "num_train_epochs": 3, "output_dir": "outputs/bitfit/t5-base/superglue-multirc", "overwrite_output_dir": true, "per_device_eval_batch_size": 32, "per_device_train_batch_size": 32, "predict_with_generate": true, "push_to_hub": true, "save_steps": 200, "save_strategy": "steps", "save_total_limit": 1, "seed": 42, "split_validation_test": true, "task_name": "superglue-multirc", "test_dataset_config_name": ["en"], "test_dataset_name": "superglue-multirc", "tokenizer_name": "../../../../plm_cache/t5-base", "unfrozen_modules": ["deltas", "layer_norm", "final_layer_norm"], "warmup_steps": 0}
|