Model save
Browse files- README.md +1 -1
- all_results.json +8 -0
- runs/Jul19_11-52-50_c85a676f2b00/events.out.tfevents.1721390865.c85a676f2b00.34.6 +2 -2
- test_results.json +8 -0
- trainer_state.json +230 -0
README.md
CHANGED
@@ -21,7 +21,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
21 |
|
22 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
23 |
It achieves the following results on the evaluation set:
|
24 |
-
- Loss: 0.
|
25 |
- Accuracy: 0.9097
|
26 |
|
27 |
## Model description
|
|
|
21 |
|
22 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
23 |
It achieves the following results on the evaluation set:
|
24 |
+
- Loss: 0.2617
|
25 |
- Accuracy: 0.9097
|
26 |
|
27 |
## Model description
|
all_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 7.101351351351352,
|
3 |
+
"eval_accuracy": 0.9096774193548387,
|
4 |
+
"eval_loss": 0.26174646615982056,
|
5 |
+
"eval_runtime": 27.2537,
|
6 |
+
"eval_samples_per_second": 5.687,
|
7 |
+
"eval_steps_per_second": 0.367
|
8 |
+
}
|
runs/Jul19_11-52-50_c85a676f2b00/events.out.tfevents.1721390865.c85a676f2b00.34.6
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 sha256:5115b6c3271b0d797e06b0c6df40180d095b91da0fa5e9c877973b2fbded9b87
|
3 |
+
size 734
|
test_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 7.101351351351352,
|
3 |
+
"eval_accuracy": 0.9096774193548387,
|
4 |
+
"eval_loss": 0.26174646615982056,
|
5 |
+
"eval_runtime": 27.2537,
|
6 |
+
"eval_samples_per_second": 5.687,
|
7 |
+
"eval_steps_per_second": 0.367
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,230 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.9571428571428572,
|
3 |
+
"best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-133",
|
4 |
+
"epoch": 7.101351351351352,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 148,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.06756756756756757,
|
13 |
+
"grad_norm": 6.901234149932861,
|
14 |
+
"learning_rate": 3.3333333333333335e-05,
|
15 |
+
"loss": 2.0115,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.12837837837837837,
|
20 |
+
"eval_accuracy": 0.5428571428571428,
|
21 |
+
"eval_loss": 1.546947956085205,
|
22 |
+
"eval_runtime": 10.0454,
|
23 |
+
"eval_samples_per_second": 6.968,
|
24 |
+
"eval_steps_per_second": 0.498,
|
25 |
+
"step": 19
|
26 |
+
},
|
27 |
+
{
|
28 |
+
"epoch": 1.0067567567567568,
|
29 |
+
"grad_norm": 7.984360218048096,
|
30 |
+
"learning_rate": 4.81203007518797e-05,
|
31 |
+
"loss": 1.6913,
|
32 |
+
"step": 20
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 1.0743243243243243,
|
36 |
+
"grad_norm": 6.522056579589844,
|
37 |
+
"learning_rate": 4.43609022556391e-05,
|
38 |
+
"loss": 1.2145,
|
39 |
+
"step": 30
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 1.1283783783783783,
|
43 |
+
"eval_accuracy": 0.7,
|
44 |
+
"eval_loss": 0.9201391935348511,
|
45 |
+
"eval_runtime": 10.5435,
|
46 |
+
"eval_samples_per_second": 6.639,
|
47 |
+
"eval_steps_per_second": 0.474,
|
48 |
+
"step": 38
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 2.0135135135135136,
|
52 |
+
"grad_norm": 6.391108989715576,
|
53 |
+
"learning_rate": 4.0601503759398494e-05,
|
54 |
+
"loss": 0.8009,
|
55 |
+
"step": 40
|
56 |
+
},
|
57 |
+
{
|
58 |
+
"epoch": 2.081081081081081,
|
59 |
+
"grad_norm": 8.276566505432129,
|
60 |
+
"learning_rate": 3.6842105263157895e-05,
|
61 |
+
"loss": 0.6166,
|
62 |
+
"step": 50
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 2.1283783783783785,
|
66 |
+
"eval_accuracy": 0.8285714285714286,
|
67 |
+
"eval_loss": 0.5547969341278076,
|
68 |
+
"eval_runtime": 11.9084,
|
69 |
+
"eval_samples_per_second": 5.878,
|
70 |
+
"eval_steps_per_second": 0.42,
|
71 |
+
"step": 57
|
72 |
+
},
|
73 |
+
{
|
74 |
+
"epoch": 3.02027027027027,
|
75 |
+
"grad_norm": 21.176633834838867,
|
76 |
+
"learning_rate": 3.3082706766917295e-05,
|
77 |
+
"loss": 0.4484,
|
78 |
+
"step": 60
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"epoch": 3.0878378378378377,
|
82 |
+
"grad_norm": 3.571500062942505,
|
83 |
+
"learning_rate": 2.9323308270676693e-05,
|
84 |
+
"loss": 0.3255,
|
85 |
+
"step": 70
|
86 |
+
},
|
87 |
+
{
|
88 |
+
"epoch": 3.1283783783783785,
|
89 |
+
"eval_accuracy": 0.9,
|
90 |
+
"eval_loss": 0.35559970140457153,
|
91 |
+
"eval_runtime": 11.9471,
|
92 |
+
"eval_samples_per_second": 5.859,
|
93 |
+
"eval_steps_per_second": 0.419,
|
94 |
+
"step": 76
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 4.027027027027027,
|
98 |
+
"grad_norm": 3.0831594467163086,
|
99 |
+
"learning_rate": 2.556390977443609e-05,
|
100 |
+
"loss": 0.2372,
|
101 |
+
"step": 80
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 4.094594594594595,
|
105 |
+
"grad_norm": 1.0621541738510132,
|
106 |
+
"learning_rate": 2.1804511278195487e-05,
|
107 |
+
"loss": 0.1945,
|
108 |
+
"step": 90
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 4.128378378378378,
|
112 |
+
"eval_accuracy": 0.8857142857142857,
|
113 |
+
"eval_loss": 0.2917786240577698,
|
114 |
+
"eval_runtime": 11.1985,
|
115 |
+
"eval_samples_per_second": 6.251,
|
116 |
+
"eval_steps_per_second": 0.446,
|
117 |
+
"step": 95
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 5.033783783783784,
|
121 |
+
"grad_norm": 7.335643291473389,
|
122 |
+
"learning_rate": 1.8045112781954888e-05,
|
123 |
+
"loss": 0.1863,
|
124 |
+
"step": 100
|
125 |
+
},
|
126 |
+
{
|
127 |
+
"epoch": 5.101351351351352,
|
128 |
+
"grad_norm": 3.736997604370117,
|
129 |
+
"learning_rate": 1.4285714285714285e-05,
|
130 |
+
"loss": 0.098,
|
131 |
+
"step": 110
|
132 |
+
},
|
133 |
+
{
|
134 |
+
"epoch": 5.128378378378378,
|
135 |
+
"eval_accuracy": 0.8714285714285714,
|
136 |
+
"eval_loss": 0.38738560676574707,
|
137 |
+
"eval_runtime": 11.6146,
|
138 |
+
"eval_samples_per_second": 6.027,
|
139 |
+
"eval_steps_per_second": 0.43,
|
140 |
+
"step": 114
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 6.04054054054054,
|
144 |
+
"grad_norm": 4.334865570068359,
|
145 |
+
"learning_rate": 1.0526315789473684e-05,
|
146 |
+
"loss": 0.0613,
|
147 |
+
"step": 120
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"epoch": 6.108108108108108,
|
151 |
+
"grad_norm": 6.393611907958984,
|
152 |
+
"learning_rate": 6.766917293233083e-06,
|
153 |
+
"loss": 0.0571,
|
154 |
+
"step": 130
|
155 |
+
},
|
156 |
+
{
|
157 |
+
"epoch": 6.128378378378378,
|
158 |
+
"eval_accuracy": 0.9571428571428572,
|
159 |
+
"eval_loss": 0.1539522409439087,
|
160 |
+
"eval_runtime": 11.8139,
|
161 |
+
"eval_samples_per_second": 5.925,
|
162 |
+
"eval_steps_per_second": 0.423,
|
163 |
+
"step": 133
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 7.047297297297297,
|
167 |
+
"grad_norm": 0.5439800024032593,
|
168 |
+
"learning_rate": 3.007518796992481e-06,
|
169 |
+
"loss": 0.0387,
|
170 |
+
"step": 140
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 7.101351351351352,
|
174 |
+
"eval_accuracy": 0.8571428571428571,
|
175 |
+
"eval_loss": 0.25466597080230713,
|
176 |
+
"eval_runtime": 12.6936,
|
177 |
+
"eval_samples_per_second": 5.515,
|
178 |
+
"eval_steps_per_second": 0.394,
|
179 |
+
"step": 148
|
180 |
+
},
|
181 |
+
{
|
182 |
+
"epoch": 7.101351351351352,
|
183 |
+
"step": 148,
|
184 |
+
"total_flos": 2.915999166844109e+18,
|
185 |
+
"train_loss": 0.5411020509294562,
|
186 |
+
"train_runtime": 849.4097,
|
187 |
+
"train_samples_per_second": 2.788,
|
188 |
+
"train_steps_per_second": 0.174
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 7.101351351351352,
|
192 |
+
"eval_accuracy": 0.9096774193548387,
|
193 |
+
"eval_loss": 0.25891318917274475,
|
194 |
+
"eval_runtime": 26.949,
|
195 |
+
"eval_samples_per_second": 5.752,
|
196 |
+
"eval_steps_per_second": 0.371,
|
197 |
+
"step": 148
|
198 |
+
},
|
199 |
+
{
|
200 |
+
"epoch": 7.101351351351352,
|
201 |
+
"eval_accuracy": 0.9096774193548387,
|
202 |
+
"eval_loss": 0.26174646615982056,
|
203 |
+
"eval_runtime": 27.2537,
|
204 |
+
"eval_samples_per_second": 5.687,
|
205 |
+
"eval_steps_per_second": 0.367,
|
206 |
+
"step": 148
|
207 |
+
}
|
208 |
+
],
|
209 |
+
"logging_steps": 10,
|
210 |
+
"max_steps": 148,
|
211 |
+
"num_input_tokens_seen": 0,
|
212 |
+
"num_train_epochs": 9223372036854775807,
|
213 |
+
"save_steps": 500,
|
214 |
+
"stateful_callbacks": {
|
215 |
+
"TrainerControl": {
|
216 |
+
"args": {
|
217 |
+
"should_epoch_stop": false,
|
218 |
+
"should_evaluate": false,
|
219 |
+
"should_log": false,
|
220 |
+
"should_save": true,
|
221 |
+
"should_training_stop": true
|
222 |
+
},
|
223 |
+
"attributes": {}
|
224 |
+
}
|
225 |
+
},
|
226 |
+
"total_flos": 2.915999166844109e+18,
|
227 |
+
"train_batch_size": 16,
|
228 |
+
"trial_name": null,
|
229 |
+
"trial_params": null
|
230 |
+
}
|