gokulsrinivasagan
commited on
Commit
•
1af6413
1
Parent(s):
eb9619a
End of training
Browse files- README.md +19 -4
- all_results.json +11 -11
- eval_results.json +6 -6
- logs/events.out.tfevents.1733327293.ki-g0008.1208741.29 +3 -0
- train_results.json +6 -6
- trainer_state.json +60 -108
README.md
CHANGED
@@ -1,13 +1,28 @@
|
|
1 |
---
|
2 |
library_name: transformers
|
|
|
|
|
3 |
base_model: gokulsrinivasagan/bert_tiny_lda_100_v1
|
4 |
tags:
|
5 |
- generated_from_trainer
|
|
|
|
|
6 |
metrics:
|
7 |
- accuracy
|
8 |
model-index:
|
9 |
- name: bert_tiny_lda_100_v1_sst2
|
10 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
---
|
12 |
|
13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -15,10 +30,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# bert_tiny_lda_100_v1_sst2
|
17 |
|
18 |
-
This model is a fine-tuned version of [gokulsrinivasagan/bert_tiny_lda_100_v1](https://huggingface.co/gokulsrinivasagan/bert_tiny_lda_100_v1) on
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss: 0.
|
21 |
-
- Accuracy: 0.
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
1 |
---
|
2 |
library_name: transformers
|
3 |
+
language:
|
4 |
+
- en
|
5 |
base_model: gokulsrinivasagan/bert_tiny_lda_100_v1
|
6 |
tags:
|
7 |
- generated_from_trainer
|
8 |
+
datasets:
|
9 |
+
- glue
|
10 |
metrics:
|
11 |
- accuracy
|
12 |
model-index:
|
13 |
- name: bert_tiny_lda_100_v1_sst2
|
14 |
+
results:
|
15 |
+
- task:
|
16 |
+
name: Text Classification
|
17 |
+
type: text-classification
|
18 |
+
dataset:
|
19 |
+
name: GLUE SST2
|
20 |
+
type: glue
|
21 |
+
args: sst2
|
22 |
+
metrics:
|
23 |
+
- name: Accuracy
|
24 |
+
type: accuracy
|
25 |
+
value: 0.805045871559633
|
26 |
---
|
27 |
|
28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
30 |
|
31 |
# bert_tiny_lda_100_v1_sst2
|
32 |
|
33 |
+
This model is a fine-tuned version of [gokulsrinivasagan/bert_tiny_lda_100_v1](https://huggingface.co/gokulsrinivasagan/bert_tiny_lda_100_v1) on the GLUE SST2 dataset.
|
34 |
It achieves the following results on the evaluation set:
|
35 |
+
- Loss: 0.4466
|
36 |
+
- Accuracy: 0.8050
|
37 |
|
38 |
## Model description
|
39 |
|
all_results.json
CHANGED
@@ -1,15 +1,15 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime": 0.
|
6 |
"eval_samples": 872,
|
7 |
-
"eval_samples_per_second":
|
8 |
-
"eval_steps_per_second":
|
9 |
-
"total_flos": 1.
|
10 |
-
"train_loss": 0.
|
11 |
-
"train_runtime":
|
12 |
"train_samples": 67349,
|
13 |
-
"train_samples_per_second":
|
14 |
-
"train_steps_per_second":
|
15 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 6.0,
|
3 |
+
"eval_accuracy": 0.805045871559633,
|
4 |
+
"eval_loss": 0.4466005265712738,
|
5 |
+
"eval_runtime": 0.2786,
|
6 |
"eval_samples": 872,
|
7 |
+
"eval_samples_per_second": 3130.287,
|
8 |
+
"eval_steps_per_second": 14.359,
|
9 |
+
"total_flos": 1.0596790573590528e+16,
|
10 |
+
"train_loss": 0.21827046196870128,
|
11 |
+
"train_runtime": 225.56,
|
12 |
"train_samples": 67349,
|
13 |
+
"train_samples_per_second": 14929.289,
|
14 |
+
"train_steps_per_second": 58.521
|
15 |
}
|
eval_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime": 0.
|
6 |
"eval_samples": 872,
|
7 |
-
"eval_samples_per_second":
|
8 |
-
"eval_steps_per_second":
|
9 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 6.0,
|
3 |
+
"eval_accuracy": 0.805045871559633,
|
4 |
+
"eval_loss": 0.4466005265712738,
|
5 |
+
"eval_runtime": 0.2786,
|
6 |
"eval_samples": 872,
|
7 |
+
"eval_samples_per_second": 3130.287,
|
8 |
+
"eval_steps_per_second": 14.359
|
9 |
}
|
logs/events.out.tfevents.1733327293.ki-g0008.1208741.29
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:507d136c425b1c061096c3a4128514b97cc5871118f07e1af01b1380a2a5fb91
|
3 |
+
size 411
|
train_results.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"total_flos": 1.
|
4 |
-
"train_loss": 0.
|
5 |
-
"train_runtime":
|
6 |
"train_samples": 67349,
|
7 |
-
"train_samples_per_second":
|
8 |
-
"train_steps_per_second":
|
9 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 6.0,
|
3 |
+
"total_flos": 1.0596790573590528e+16,
|
4 |
+
"train_loss": 0.21827046196870128,
|
5 |
+
"train_runtime": 225.56,
|
6 |
"train_samples": 67349,
|
7 |
+
"train_samples_per_second": 14929.289,
|
8 |
+
"train_steps_per_second": 58.521
|
9 |
}
|
trainer_state.json
CHANGED
@@ -1,165 +1,117 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "bert_tiny_lda_100_v1_sst2/checkpoint-
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 1.0,
|
13 |
-
"grad_norm":
|
14 |
-
"learning_rate":
|
15 |
-
"loss": 0.
|
16 |
"step": 264
|
17 |
},
|
18 |
{
|
19 |
"epoch": 1.0,
|
20 |
-
"eval_accuracy": 0.
|
21 |
-
"eval_loss": 0.
|
22 |
-
"eval_runtime": 0.
|
23 |
-
"eval_samples_per_second":
|
24 |
-
"eval_steps_per_second":
|
25 |
"step": 264
|
26 |
},
|
27 |
{
|
28 |
"epoch": 2.0,
|
29 |
-
"grad_norm":
|
30 |
-
"learning_rate":
|
31 |
-
"loss": 0.
|
32 |
"step": 528
|
33 |
},
|
34 |
{
|
35 |
"epoch": 2.0,
|
36 |
-
"eval_accuracy": 0.
|
37 |
-
"eval_loss": 0.
|
38 |
-
"eval_runtime": 0.
|
39 |
-
"eval_samples_per_second":
|
40 |
-
"eval_steps_per_second": 14.
|
41 |
"step": 528
|
42 |
},
|
43 |
{
|
44 |
"epoch": 3.0,
|
45 |
-
"grad_norm":
|
46 |
-
"learning_rate":
|
47 |
-
"loss": 0.
|
48 |
"step": 792
|
49 |
},
|
50 |
{
|
51 |
"epoch": 3.0,
|
52 |
-
"eval_accuracy": 0.
|
53 |
-
"eval_loss": 0.
|
54 |
-
"eval_runtime": 0.
|
55 |
-
"eval_samples_per_second":
|
56 |
-
"eval_steps_per_second":
|
57 |
"step": 792
|
58 |
},
|
59 |
{
|
60 |
"epoch": 4.0,
|
61 |
-
"grad_norm":
|
62 |
-
"learning_rate":
|
63 |
-
"loss": 0.
|
64 |
"step": 1056
|
65 |
},
|
66 |
{
|
67 |
"epoch": 4.0,
|
68 |
-
"eval_accuracy": 0.
|
69 |
-
"eval_loss": 0.
|
70 |
-
"eval_runtime": 0.
|
71 |
-
"eval_samples_per_second":
|
72 |
-
"eval_steps_per_second": 14.
|
73 |
"step": 1056
|
74 |
},
|
75 |
{
|
76 |
"epoch": 5.0,
|
77 |
-
"grad_norm":
|
78 |
-
"learning_rate":
|
79 |
-
"loss": 0.
|
80 |
"step": 1320
|
81 |
},
|
82 |
{
|
83 |
"epoch": 5.0,
|
84 |
-
"eval_accuracy": 0.
|
85 |
-
"eval_loss": 0.
|
86 |
-
"eval_runtime": 0.
|
87 |
-
"eval_samples_per_second":
|
88 |
-
"eval_steps_per_second":
|
89 |
"step": 1320
|
90 |
},
|
91 |
{
|
92 |
"epoch": 6.0,
|
93 |
-
"grad_norm":
|
94 |
-
"learning_rate":
|
95 |
-
"loss": 0.
|
96 |
"step": 1584
|
97 |
},
|
98 |
{
|
99 |
"epoch": 6.0,
|
100 |
-
"eval_accuracy": 0.
|
101 |
-
"eval_loss": 0.
|
102 |
-
"eval_runtime": 0.
|
103 |
-
"eval_samples_per_second":
|
104 |
-
"eval_steps_per_second":
|
105 |
"step": 1584
|
106 |
},
|
107 |
{
|
108 |
-
"epoch":
|
109 |
-
"
|
110 |
-
"
|
111 |
-
"
|
112 |
-
"
|
113 |
-
|
114 |
-
|
115 |
-
"epoch": 7.0,
|
116 |
-
"eval_accuracy": 0.5091743119266054,
|
117 |
-
"eval_loss": 0.698842465877533,
|
118 |
-
"eval_runtime": 0.2611,
|
119 |
-
"eval_samples_per_second": 3339.088,
|
120 |
-
"eval_steps_per_second": 15.317,
|
121 |
-
"step": 1848
|
122 |
-
},
|
123 |
-
{
|
124 |
-
"epoch": 8.0,
|
125 |
-
"grad_norm": 0.049795445054769516,
|
126 |
-
"learning_rate": 0.00084,
|
127 |
-
"loss": 0.6865,
|
128 |
-
"step": 2112
|
129 |
-
},
|
130 |
-
{
|
131 |
-
"epoch": 8.0,
|
132 |
-
"eval_accuracy": 0.5091743119266054,
|
133 |
-
"eval_loss": 0.6989858150482178,
|
134 |
-
"eval_runtime": 0.2626,
|
135 |
-
"eval_samples_per_second": 3320.526,
|
136 |
-
"eval_steps_per_second": 15.232,
|
137 |
-
"step": 2112
|
138 |
-
},
|
139 |
-
{
|
140 |
-
"epoch": 9.0,
|
141 |
-
"grad_norm": 0.18644928932189941,
|
142 |
-
"learning_rate": 0.00082,
|
143 |
-
"loss": 0.6866,
|
144 |
-
"step": 2376
|
145 |
-
},
|
146 |
-
{
|
147 |
-
"epoch": 9.0,
|
148 |
-
"eval_accuracy": 0.5091743119266054,
|
149 |
-
"eval_loss": 0.6969968676567078,
|
150 |
-
"eval_runtime": 0.2637,
|
151 |
-
"eval_samples_per_second": 3306.531,
|
152 |
-
"eval_steps_per_second": 15.168,
|
153 |
-
"step": 2376
|
154 |
-
},
|
155 |
-
{
|
156 |
-
"epoch": 9.0,
|
157 |
-
"step": 2376,
|
158 |
-
"total_flos": 1.5895185860385792e+16,
|
159 |
-
"train_loss": 0.6873250633779199,
|
160 |
-
"train_runtime": 324.1575,
|
161 |
-
"train_samples_per_second": 10388.315,
|
162 |
-
"train_steps_per_second": 40.721
|
163 |
}
|
164 |
],
|
165 |
"logging_steps": 1,
|
@@ -188,7 +140,7 @@
|
|
188 |
"attributes": {}
|
189 |
}
|
190 |
},
|
191 |
-
"total_flos": 1.
|
192 |
"train_batch_size": 256,
|
193 |
"trial_name": null,
|
194 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.4466005265712738,
|
3 |
+
"best_model_checkpoint": "bert_tiny_lda_100_v1_sst2/checkpoint-264",
|
4 |
+
"epoch": 6.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 1584,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 1.0,
|
13 |
+
"grad_norm": 10.585709571838379,
|
14 |
+
"learning_rate": 4.9e-05,
|
15 |
+
"loss": 0.4292,
|
16 |
"step": 264
|
17 |
},
|
18 |
{
|
19 |
"epoch": 1.0,
|
20 |
+
"eval_accuracy": 0.805045871559633,
|
21 |
+
"eval_loss": 0.4466005265712738,
|
22 |
+
"eval_runtime": 0.2781,
|
23 |
+
"eval_samples_per_second": 3135.372,
|
24 |
+
"eval_steps_per_second": 14.382,
|
25 |
"step": 264
|
26 |
},
|
27 |
{
|
28 |
"epoch": 2.0,
|
29 |
+
"grad_norm": 6.0157647132873535,
|
30 |
+
"learning_rate": 4.8e-05,
|
31 |
+
"loss": 0.2589,
|
32 |
"step": 528
|
33 |
},
|
34 |
{
|
35 |
"epoch": 2.0,
|
36 |
+
"eval_accuracy": 0.8153669724770642,
|
37 |
+
"eval_loss": 0.4592009484767914,
|
38 |
+
"eval_runtime": 0.2832,
|
39 |
+
"eval_samples_per_second": 3079.044,
|
40 |
+
"eval_steps_per_second": 14.124,
|
41 |
"step": 528
|
42 |
},
|
43 |
{
|
44 |
"epoch": 3.0,
|
45 |
+
"grad_norm": 4.999380111694336,
|
46 |
+
"learning_rate": 4.7e-05,
|
47 |
+
"loss": 0.2007,
|
48 |
"step": 792
|
49 |
},
|
50 |
{
|
51 |
"epoch": 3.0,
|
52 |
+
"eval_accuracy": 0.8130733944954128,
|
53 |
+
"eval_loss": 0.4953426420688629,
|
54 |
+
"eval_runtime": 0.2791,
|
55 |
+
"eval_samples_per_second": 3124.457,
|
56 |
+
"eval_steps_per_second": 14.332,
|
57 |
"step": 792
|
58 |
},
|
59 |
{
|
60 |
"epoch": 4.0,
|
61 |
+
"grad_norm": 6.423249244689941,
|
62 |
+
"learning_rate": 4.600000000000001e-05,
|
63 |
+
"loss": 0.1652,
|
64 |
"step": 1056
|
65 |
},
|
66 |
{
|
67 |
"epoch": 4.0,
|
68 |
+
"eval_accuracy": 0.8302752293577982,
|
69 |
+
"eval_loss": 0.5380478501319885,
|
70 |
+
"eval_runtime": 0.2813,
|
71 |
+
"eval_samples_per_second": 3099.753,
|
72 |
+
"eval_steps_per_second": 14.219,
|
73 |
"step": 1056
|
74 |
},
|
75 |
{
|
76 |
"epoch": 5.0,
|
77 |
+
"grad_norm": 4.089705467224121,
|
78 |
+
"learning_rate": 4.5e-05,
|
79 |
+
"loss": 0.137,
|
80 |
"step": 1320
|
81 |
},
|
82 |
{
|
83 |
"epoch": 5.0,
|
84 |
+
"eval_accuracy": 0.8096330275229358,
|
85 |
+
"eval_loss": 0.5538350939750671,
|
86 |
+
"eval_runtime": 0.2751,
|
87 |
+
"eval_samples_per_second": 3169.498,
|
88 |
+
"eval_steps_per_second": 14.539,
|
89 |
"step": 1320
|
90 |
},
|
91 |
{
|
92 |
"epoch": 6.0,
|
93 |
+
"grad_norm": 6.120477199554443,
|
94 |
+
"learning_rate": 4.4000000000000006e-05,
|
95 |
+
"loss": 0.1185,
|
96 |
"step": 1584
|
97 |
},
|
98 |
{
|
99 |
"epoch": 6.0,
|
100 |
+
"eval_accuracy": 0.8096330275229358,
|
101 |
+
"eval_loss": 0.6460174918174744,
|
102 |
+
"eval_runtime": 0.2755,
|
103 |
+
"eval_samples_per_second": 3164.776,
|
104 |
+
"eval_steps_per_second": 14.517,
|
105 |
"step": 1584
|
106 |
},
|
107 |
{
|
108 |
+
"epoch": 6.0,
|
109 |
+
"step": 1584,
|
110 |
+
"total_flos": 1.0596790573590528e+16,
|
111 |
+
"train_loss": 0.21827046196870128,
|
112 |
+
"train_runtime": 225.56,
|
113 |
+
"train_samples_per_second": 14929.289,
|
114 |
+
"train_steps_per_second": 58.521
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
115 |
}
|
116 |
],
|
117 |
"logging_steps": 1,
|
|
|
140 |
"attributes": {}
|
141 |
}
|
142 |
},
|
143 |
+
"total_flos": 1.0596790573590528e+16,
|
144 |
"train_batch_size": 256,
|
145 |
"trial_name": null,
|
146 |
"trial_params": null
|