DrishtiSharma
commited on
Commit
·
403c1b4
1
Parent(s):
7326e2b
End of training
Browse files- README.md +11 -11
- all_results.json +17 -0
- eval_results.json +17 -0
- trainer_state.json +154 -0
README.md
CHANGED
@@ -17,17 +17,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
|
18 |
This model is a fine-tuned version of [bert-large-uncased](https://huggingface.co/bert-large-uncased) on the None dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss: 0.
|
21 |
-
- Accuracy: 0.
|
22 |
-
- Weighted f1: 0.
|
23 |
-
- Weighted recall: 0.
|
24 |
-
- Weighted precision: 0.
|
25 |
-
- Micro f1: 0.
|
26 |
-
- Micro recall: 0.
|
27 |
-
- Micro precision: 0.
|
28 |
-
- Macro f1: 0.
|
29 |
-
- Macro recall: 0.
|
30 |
-
- Macro precision: 0.
|
31 |
|
32 |
## Model description
|
33 |
|
|
|
17 |
|
18 |
This model is a fine-tuned version of [bert-large-uncased](https://huggingface.co/bert-large-uncased) on the None dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 0.0727
|
21 |
+
- Accuracy: 0.9772
|
22 |
+
- Weighted f1: 0.9772
|
23 |
+
- Weighted recall: 0.9772
|
24 |
+
- Weighted precision: 0.9773
|
25 |
+
- Micro f1: 0.9772
|
26 |
+
- Micro recall: 0.9772
|
27 |
+
- Micro precision: 0.9772
|
28 |
+
- Macro f1: 0.9756
|
29 |
+
- Macro recall: 0.9755
|
30 |
+
- Macro precision: 0.9757
|
31 |
|
32 |
## Model description
|
33 |
|
all_results.json
ADDED
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 5.0,
|
3 |
+
"eval_Macro F1": 0.9756041514089026,
|
4 |
+
"eval_Macro Precision": 0.9756981683292011,
|
5 |
+
"eval_Macro Recall": 0.975548273738675,
|
6 |
+
"eval_Micro F1": 0.9771986970684039,
|
7 |
+
"eval_Micro Precision": 0.9771986970684039,
|
8 |
+
"eval_Micro Recall": 0.9771986970684039,
|
9 |
+
"eval_Weighted F1": 0.977214672387114,
|
10 |
+
"eval_Weighted Precision": 0.9772649217239973,
|
11 |
+
"eval_Weighted Recall": 0.9771986970684039,
|
12 |
+
"eval_accuracy": 0.9771986970684039,
|
13 |
+
"eval_loss": 0.07267534732818604,
|
14 |
+
"eval_runtime": 1.7364,
|
15 |
+
"eval_samples_per_second": 176.805,
|
16 |
+
"eval_steps_per_second": 11.518
|
17 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 5.0,
|
3 |
+
"eval_Macro F1": 0.9756041514089026,
|
4 |
+
"eval_Macro Precision": 0.9756981683292011,
|
5 |
+
"eval_Macro Recall": 0.975548273738675,
|
6 |
+
"eval_Micro F1": 0.9771986970684039,
|
7 |
+
"eval_Micro Precision": 0.9771986970684039,
|
8 |
+
"eval_Micro Recall": 0.9771986970684039,
|
9 |
+
"eval_Weighted F1": 0.977214672387114,
|
10 |
+
"eval_Weighted Precision": 0.9772649217239973,
|
11 |
+
"eval_Weighted Recall": 0.9771986970684039,
|
12 |
+
"eval_accuracy": 0.9771986970684039,
|
13 |
+
"eval_loss": 0.07267534732818604,
|
14 |
+
"eval_runtime": 1.7364,
|
15 |
+
"eval_samples_per_second": 176.805,
|
16 |
+
"eval_steps_per_second": 11.518
|
17 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,154 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.07267534732818604,
|
3 |
+
"best_model_checkpoint": "bert-large-uncased-hate-offensive-normal-speech-lr-1e-05/checkpoint-612",
|
4 |
+
"epoch": 5.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 765,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.01,
|
13 |
+
"learning_rate": 9.986928104575165e-06,
|
14 |
+
"loss": 1.3111,
|
15 |
+
"step": 1
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 1.0,
|
19 |
+
"learning_rate": 8.000000000000001e-06,
|
20 |
+
"loss": 0.9139,
|
21 |
+
"step": 153
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 1.0,
|
25 |
+
"eval_Macro F1": 0.7946891596376133,
|
26 |
+
"eval_Macro Precision": 0.8030025975570529,
|
27 |
+
"eval_Macro Recall": 0.793248276332462,
|
28 |
+
"eval_Micro F1": 0.8110749185667753,
|
29 |
+
"eval_Micro Precision": 0.8110749185667753,
|
30 |
+
"eval_Micro Recall": 0.8110749185667753,
|
31 |
+
"eval_Weighted F1": 0.808597679674613,
|
32 |
+
"eval_Weighted Precision": 0.812490327736885,
|
33 |
+
"eval_Weighted Recall": 0.8110749185667753,
|
34 |
+
"eval_accuracy": 0.8110749185667753,
|
35 |
+
"eval_loss": 0.485382616519928,
|
36 |
+
"eval_runtime": 1.6985,
|
37 |
+
"eval_samples_per_second": 180.753,
|
38 |
+
"eval_steps_per_second": 11.775,
|
39 |
+
"step": 153
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 2.0,
|
43 |
+
"learning_rate": 6e-06,
|
44 |
+
"loss": 0.2278,
|
45 |
+
"step": 306
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 2.0,
|
49 |
+
"eval_Macro F1": 0.965223571946261,
|
50 |
+
"eval_Macro Precision": 0.9654715673031662,
|
51 |
+
"eval_Macro Recall": 0.9657485621293646,
|
52 |
+
"eval_Micro F1": 0.9674267100977199,
|
53 |
+
"eval_Micro Precision": 0.9674267100977199,
|
54 |
+
"eval_Micro Recall": 0.9674267100977199,
|
55 |
+
"eval_Weighted F1": 0.9675123332857436,
|
56 |
+
"eval_Weighted Precision": 0.9682947849236886,
|
57 |
+
"eval_Weighted Recall": 0.9674267100977199,
|
58 |
+
"eval_accuracy": 0.9674267100977199,
|
59 |
+
"eval_loss": 0.0775093287229538,
|
60 |
+
"eval_runtime": 1.6928,
|
61 |
+
"eval_samples_per_second": 181.358,
|
62 |
+
"eval_steps_per_second": 11.815,
|
63 |
+
"step": 306
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 3.0,
|
67 |
+
"learning_rate": 4.000000000000001e-06,
|
68 |
+
"loss": 0.0582,
|
69 |
+
"step": 459
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 3.0,
|
73 |
+
"eval_Macro F1": 0.9756573721562932,
|
74 |
+
"eval_Macro Precision": 0.9768818096344228,
|
75 |
+
"eval_Macro Recall": 0.9753907251940297,
|
76 |
+
"eval_Micro F1": 0.9771986970684039,
|
77 |
+
"eval_Micro Precision": 0.9771986970684039,
|
78 |
+
"eval_Micro Recall": 0.9771986970684039,
|
79 |
+
"eval_Weighted F1": 0.9772671060837873,
|
80 |
+
"eval_Weighted Precision": 0.9781958385960247,
|
81 |
+
"eval_Weighted Recall": 0.9771986970684039,
|
82 |
+
"eval_accuracy": 0.9771986970684039,
|
83 |
+
"eval_loss": 0.10066915303468704,
|
84 |
+
"eval_runtime": 1.6955,
|
85 |
+
"eval_samples_per_second": 181.07,
|
86 |
+
"eval_steps_per_second": 11.796,
|
87 |
+
"step": 459
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 4.0,
|
91 |
+
"learning_rate": 2.0000000000000003e-06,
|
92 |
+
"loss": 0.0228,
|
93 |
+
"step": 612
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 4.0,
|
97 |
+
"eval_Macro F1": 0.9756041514089026,
|
98 |
+
"eval_Macro Precision": 0.9756981683292011,
|
99 |
+
"eval_Macro Recall": 0.975548273738675,
|
100 |
+
"eval_Micro F1": 0.9771986970684039,
|
101 |
+
"eval_Micro Precision": 0.9771986970684039,
|
102 |
+
"eval_Micro Recall": 0.9771986970684039,
|
103 |
+
"eval_Weighted F1": 0.977214672387114,
|
104 |
+
"eval_Weighted Precision": 0.9772649217239973,
|
105 |
+
"eval_Weighted Recall": 0.9771986970684039,
|
106 |
+
"eval_accuracy": 0.9771986970684039,
|
107 |
+
"eval_loss": 0.07267534732818604,
|
108 |
+
"eval_runtime": 1.6931,
|
109 |
+
"eval_samples_per_second": 181.322,
|
110 |
+
"eval_steps_per_second": 11.812,
|
111 |
+
"step": 612
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 5.0,
|
115 |
+
"learning_rate": 0.0,
|
116 |
+
"loss": 0.0159,
|
117 |
+
"step": 765
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 5.0,
|
121 |
+
"eval_Macro F1": 0.9719944646254973,
|
122 |
+
"eval_Macro Precision": 0.9722554765464863,
|
123 |
+
"eval_Macro Recall": 0.9718852700756714,
|
124 |
+
"eval_Micro F1": 0.9739413680781759,
|
125 |
+
"eval_Micro Precision": 0.9739413680781759,
|
126 |
+
"eval_Micro Recall": 0.9739413680781759,
|
127 |
+
"eval_Weighted F1": 0.9739714001894333,
|
128 |
+
"eval_Weighted Precision": 0.9741378480054397,
|
129 |
+
"eval_Weighted Recall": 0.9739413680781759,
|
130 |
+
"eval_accuracy": 0.9739413680781759,
|
131 |
+
"eval_loss": 0.10134366899728775,
|
132 |
+
"eval_runtime": 1.7036,
|
133 |
+
"eval_samples_per_second": 180.21,
|
134 |
+
"eval_steps_per_second": 11.74,
|
135 |
+
"step": 765
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 5.0,
|
139 |
+
"step": 765,
|
140 |
+
"total_flos": 2539812775083840.0,
|
141 |
+
"train_loss": 0.24821887405869228,
|
142 |
+
"train_runtime": 280.5122,
|
143 |
+
"train_samples_per_second": 43.634,
|
144 |
+
"train_steps_per_second": 2.727
|
145 |
+
}
|
146 |
+
],
|
147 |
+
"logging_steps": 500,
|
148 |
+
"max_steps": 765,
|
149 |
+
"num_train_epochs": 5,
|
150 |
+
"save_steps": 500,
|
151 |
+
"total_flos": 2539812775083840.0,
|
152 |
+
"trial_name": null,
|
153 |
+
"trial_params": null
|
154 |
+
}
|