Hugging Face's logo

kevinoli
/
clip-finetuned-csu-p14-336-e4l59-l

Zero-Shot Image Classification
Transformers
TensorBoard
Safetensors
clip
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
kevinoli commited on
Commit
a67aec6
•
1 Parent(s): 9b95db0

Training in progress, step 8000, checkpoint

Browse files
Files changed (7) hide show
  1. checkpoint-8000/config.json +31 -0
  2. checkpoint-8000/model.safetensors +3 -0
  3. checkpoint-8000/optimizer.pt +3 -0
  4. checkpoint-8000/rng_state.pth +3 -0
  5. checkpoint-8000/scheduler.pt +3 -0
  6. checkpoint-8000/trainer_state.json +273 -0
  7. checkpoint-8000/training_args.bin +3 -0
checkpoint-8000/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/clip-vit-large-patch14-336",
3
+ "architectures": [
4
+ "CLIPModel"
5
+ ],
6
+ "initializer_factor": 1.0,
7
+ "logit_scale_init_value": 2.6592,
8
+ "model_type": "clip",
9
+ "projection_dim": 768,
10
+ "text_config": {
11
+ "dropout": 0.0,
12
+ "hidden_size": 768,
13
+ "intermediate_size": 3072,
14
+ "model_type": "clip_text_model",
15
+ "num_attention_heads": 12,
16
+ "projection_dim": 768
17
+ },
18
+ "torch_dtype": "float32",
19
+ "transformers_version": "4.45.0.dev0",
20
+ "vision_config": {
21
+ "dropout": 0.0,
22
+ "hidden_size": 1024,
23
+ "image_size": 336,
24
+ "intermediate_size": 4096,
25
+ "model_type": "clip_vision_model",
26
+ "num_attention_heads": 16,
27
+ "num_hidden_layers": 24,
28
+ "patch_size": 14,
29
+ "projection_dim": 768
30
+ }
31
+ }
checkpoint-8000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5086c06895b2790b6349210b0ae23592ebbef81f64e12b56dc522f73e1463b69
3
+ size 1711848436
checkpoint-8000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:813d209a37464141e49b8dcc92328f8da4a476c035395bed2d086b74ba4d4cd5
3
+ size 3424043887
checkpoint-8000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41054f6a7f29c2148ed6fea73ad66d1725aa683821303aeecc507f856769e9c4
3
+ size 14503
checkpoint-8000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8952714691cd32cebfa4a404176461e31aacb95f8264d47601c760321299c7d3
3
+ size 623
checkpoint-8000/trainer_state.json ADDED
@@ -0,0 +1,273 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.3937218189239502,
3
+ "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l59-l/checkpoint-8000",
4
+ "epoch": 1.4738393515106853,
5
+ "eval_steps": 500,
6
+ "global_step": 8000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.09211495946941783,
13
+ "grad_norm": 20.952665328979492,
14
+ "learning_rate": 4.884856300663228e-09,
15
+ "loss": 0.3952,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.09211495946941783,
20
+ "eval_loss": 1.4940427541732788,
21
+ "eval_runtime": 73.9282,
22
+ "eval_samples_per_second": 16.327,
23
+ "eval_steps_per_second": 2.043,
24
+ "step": 500
25
+ },
26
+ {
27
+ "epoch": 0.18422991893883567,
28
+ "grad_norm": 59.28666687011719,
29
+ "learning_rate": 4.769712601326456e-09,
30
+ "loss": 0.4562,
31
+ "step": 1000
32
+ },
33
+ {
34
+ "epoch": 0.18422991893883567,
35
+ "eval_loss": 1.4853259325027466,
36
+ "eval_runtime": 74.6533,
37
+ "eval_samples_per_second": 16.168,
38
+ "eval_steps_per_second": 2.023,
39
+ "step": 1000
40
+ },
41
+ {
42
+ "epoch": 0.2763448784082535,
43
+ "grad_norm": 1.179814338684082,
44
+ "learning_rate": 4.654568901989683e-09,
45
+ "loss": 0.5131,
46
+ "step": 1500
47
+ },
48
+ {
49
+ "epoch": 0.2763448784082535,
50
+ "eval_loss": 1.4757832288742065,
51
+ "eval_runtime": 76.2866,
52
+ "eval_samples_per_second": 15.822,
53
+ "eval_steps_per_second": 1.979,
54
+ "step": 1500
55
+ },
56
+ {
57
+ "epoch": 0.36845983787767134,
58
+ "grad_norm": 0.900991678237915,
59
+ "learning_rate": 4.5394252026529105e-09,
60
+ "loss": 0.4481,
61
+ "step": 2000
62
+ },
63
+ {
64
+ "epoch": 0.36845983787767134,
65
+ "eval_loss": 1.4675536155700684,
66
+ "eval_runtime": 75.9348,
67
+ "eval_samples_per_second": 15.895,
68
+ "eval_steps_per_second": 1.989,
69
+ "step": 2000
70
+ },
71
+ {
72
+ "epoch": 0.46057479734708917,
73
+ "grad_norm": 346.93853759765625,
74
+ "learning_rate": 4.424281503316139e-09,
75
+ "loss": 0.4839,
76
+ "step": 2500
77
+ },
78
+ {
79
+ "epoch": 0.46057479734708917,
80
+ "eval_loss": 1.4585391283035278,
81
+ "eval_runtime": 76.7708,
82
+ "eval_samples_per_second": 15.722,
83
+ "eval_steps_per_second": 1.967,
84
+ "step": 2500
85
+ },
86
+ {
87
+ "epoch": 0.552689756816507,
88
+ "grad_norm": 318.47515869140625,
89
+ "learning_rate": 4.3091378039793666e-09,
90
+ "loss": 0.4377,
91
+ "step": 3000
92
+ },
93
+ {
94
+ "epoch": 0.552689756816507,
95
+ "eval_loss": 1.4507631063461304,
96
+ "eval_runtime": 76.5424,
97
+ "eval_samples_per_second": 15.769,
98
+ "eval_steps_per_second": 1.973,
99
+ "step": 3000
100
+ },
101
+ {
102
+ "epoch": 0.6448047162859248,
103
+ "grad_norm": 112.41814422607422,
104
+ "learning_rate": 4.193994104642594e-09,
105
+ "loss": 0.4231,
106
+ "step": 3500
107
+ },
108
+ {
109
+ "epoch": 0.6448047162859248,
110
+ "eval_loss": 1.4432125091552734,
111
+ "eval_runtime": 76.4812,
112
+ "eval_samples_per_second": 15.782,
113
+ "eval_steps_per_second": 1.974,
114
+ "step": 3500
115
+ },
116
+ {
117
+ "epoch": 0.7369196757553427,
118
+ "grad_norm": 83.51781463623047,
119
+ "learning_rate": 4.078850405305822e-09,
120
+ "loss": 0.4369,
121
+ "step": 4000
122
+ },
123
+ {
124
+ "epoch": 0.7369196757553427,
125
+ "eval_loss": 1.4365661144256592,
126
+ "eval_runtime": 76.4405,
127
+ "eval_samples_per_second": 15.79,
128
+ "eval_steps_per_second": 1.975,
129
+ "step": 4000
130
+ },
131
+ {
132
+ "epoch": 0.8290346352247605,
133
+ "grad_norm": 460.5648498535156,
134
+ "learning_rate": 3.963706705969049e-09,
135
+ "loss": 0.4082,
136
+ "step": 4500
137
+ },
138
+ {
139
+ "epoch": 0.8290346352247605,
140
+ "eval_loss": 1.4302303791046143,
141
+ "eval_runtime": 75.042,
142
+ "eval_samples_per_second": 16.084,
143
+ "eval_steps_per_second": 2.012,
144
+ "step": 4500
145
+ },
146
+ {
147
+ "epoch": 0.9211495946941783,
148
+ "grad_norm": 5.289137363433838,
149
+ "learning_rate": 3.848563006632278e-09,
150
+ "loss": 0.4234,
151
+ "step": 5000
152
+ },
153
+ {
154
+ "epoch": 0.9211495946941783,
155
+ "eval_loss": 1.4243344068527222,
156
+ "eval_runtime": 76.6198,
157
+ "eval_samples_per_second": 15.753,
158
+ "eval_steps_per_second": 1.971,
159
+ "step": 5000
160
+ },
161
+ {
162
+ "epoch": 1.013264554163596,
163
+ "grad_norm": 0.6147650480270386,
164
+ "learning_rate": 3.733419307295505e-09,
165
+ "loss": 0.4266,
166
+ "step": 5500
167
+ },
168
+ {
169
+ "epoch": 1.013264554163596,
170
+ "eval_loss": 1.4190788269042969,
171
+ "eval_runtime": 75.4536,
172
+ "eval_samples_per_second": 15.997,
173
+ "eval_steps_per_second": 2.001,
174
+ "step": 5500
175
+ },
176
+ {
177
+ "epoch": 1.105379513633014,
178
+ "grad_norm": 231.71815490722656,
179
+ "learning_rate": 3.6182756079587326e-09,
180
+ "loss": 0.4438,
181
+ "step": 6000
182
+ },
183
+ {
184
+ "epoch": 1.105379513633014,
185
+ "eval_loss": 1.4136912822723389,
186
+ "eval_runtime": 76.5978,
187
+ "eval_samples_per_second": 15.758,
188
+ "eval_steps_per_second": 1.971,
189
+ "step": 6000
190
+ },
191
+ {
192
+ "epoch": 1.1974944731024317,
193
+ "grad_norm": 464.0763854980469,
194
+ "learning_rate": 3.5031319086219602e-09,
195
+ "loss": 0.3814,
196
+ "step": 6500
197
+ },
198
+ {
199
+ "epoch": 1.1974944731024317,
200
+ "eval_loss": 1.4085203409194946,
201
+ "eval_runtime": 75.4916,
202
+ "eval_samples_per_second": 15.989,
203
+ "eval_steps_per_second": 2.0,
204
+ "step": 6500
205
+ },
206
+ {
207
+ "epoch": 1.2896094325718497,
208
+ "grad_norm": 2.5053436756134033,
209
+ "learning_rate": 3.3879882092851882e-09,
210
+ "loss": 0.3327,
211
+ "step": 7000
212
+ },
213
+ {
214
+ "epoch": 1.2896094325718497,
215
+ "eval_loss": 1.4041872024536133,
216
+ "eval_runtime": 75.2076,
217
+ "eval_samples_per_second": 16.049,
218
+ "eval_steps_per_second": 2.008,
219
+ "step": 7000
220
+ },
221
+ {
222
+ "epoch": 1.3817243920412676,
223
+ "grad_norm": 126.44642639160156,
224
+ "learning_rate": 3.2728445099484154e-09,
225
+ "loss": 0.4045,
226
+ "step": 7500
227
+ },
228
+ {
229
+ "epoch": 1.3817243920412676,
230
+ "eval_loss": 1.398864984512329,
231
+ "eval_runtime": 75.3652,
232
+ "eval_samples_per_second": 16.015,
233
+ "eval_steps_per_second": 2.004,
234
+ "step": 7500
235
+ },
236
+ {
237
+ "epoch": 1.4738393515106853,
238
+ "grad_norm": 1.046550989151001,
239
+ "learning_rate": 3.1577008106116435e-09,
240
+ "loss": 0.4038,
241
+ "step": 8000
242
+ },
243
+ {
244
+ "epoch": 1.4738393515106853,
245
+ "eval_loss": 1.3937218189239502,
246
+ "eval_runtime": 75.1378,
247
+ "eval_samples_per_second": 16.064,
248
+ "eval_steps_per_second": 2.01,
249
+ "step": 8000
250
+ }
251
+ ],
252
+ "logging_steps": 500,
253
+ "max_steps": 21712,
254
+ "num_input_tokens_seen": 0,
255
+ "num_train_epochs": 4,
256
+ "save_steps": 500,
257
+ "stateful_callbacks": {
258
+ "TrainerControl": {
259
+ "args": {
260
+ "should_epoch_stop": false,
261
+ "should_evaluate": false,
262
+ "should_log": false,
263
+ "should_save": true,
264
+ "should_training_stop": false
265
+ },
266
+ "attributes": {}
267
+ }
268
+ },
269
+ "total_flos": 2878786326782520.0,
270
+ "train_batch_size": 2,
271
+ "trial_name": null,
272
+ "trial_params": null
273
+ }
checkpoint-8000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17dd82d53ab261b7081e3e93503b374d907fea92ca3e505d0ff1319c5765d331
3
+ size 4847