kevinoli commited on
Commit
c629db3
1 Parent(s): 6a7a286

Training in progress, step 4000, checkpoint

Browse files
checkpoint-4000/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-4000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1fd6b8f2b0e94dfe260a1dd8a49b317198033c4155d6d15e067bbad16ecbc290
3
+ size 1711848436
checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6194ce851679c69cccd1917b3eba2c3b2c9df0696f497da3cbb1251ad8fb2686
3
+ size 3424043887
checkpoint-4000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0289613f9abb782023ce231f34f2dee59b6b952ce76bf1fa3bd8f2e3b35b3825
3
+ size 14503
checkpoint-4000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cba266763f45c5a92f9dd3e089490f6840349f485cd776d10d2de75265bed452
3
+ size 623
checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,153 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.0176074504852295,
3
+ "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l17-l/checkpoint-4000",
4
+ "epoch": 0.42662116040955633,
5
+ "eval_steps": 500,
6
+ "global_step": 4000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05332764505119454,
13
+ "grad_norm": 247.07174682617188,
14
+ "learning_rate": 9.822241183162683e-08,
15
+ "loss": 0.4515,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.05332764505119454,
20
+ "eval_loss": 1.38526451587677,
21
+ "eval_runtime": 63.6956,
22
+ "eval_samples_per_second": 15.496,
23
+ "eval_steps_per_second": 1.947,
24
+ "step": 500
25
+ },
26
+ {
27
+ "epoch": 0.10665529010238908,
28
+ "grad_norm": 72.11767578125,
29
+ "learning_rate": 9.64448236632537e-08,
30
+ "loss": 0.414,
31
+ "step": 1000
32
+ },
33
+ {
34
+ "epoch": 0.10665529010238908,
35
+ "eval_loss": 1.284197211265564,
36
+ "eval_runtime": 62.7418,
37
+ "eval_samples_per_second": 15.731,
38
+ "eval_steps_per_second": 1.976,
39
+ "step": 1000
40
+ },
41
+ {
42
+ "epoch": 0.1599829351535836,
43
+ "grad_norm": 812.2269287109375,
44
+ "learning_rate": 9.466723549488054e-08,
45
+ "loss": 0.3279,
46
+ "step": 1500
47
+ },
48
+ {
49
+ "epoch": 0.1599829351535836,
50
+ "eval_loss": 1.2100127935409546,
51
+ "eval_runtime": 62.6215,
52
+ "eval_samples_per_second": 15.761,
53
+ "eval_steps_per_second": 1.98,
54
+ "step": 1500
55
+ },
56
+ {
57
+ "epoch": 0.21331058020477817,
58
+ "grad_norm": 4.459943771362305,
59
+ "learning_rate": 9.288964732650739e-08,
60
+ "loss": 0.2912,
61
+ "step": 2000
62
+ },
63
+ {
64
+ "epoch": 0.21331058020477817,
65
+ "eval_loss": 1.1584910154342651,
66
+ "eval_runtime": 64.0453,
67
+ "eval_samples_per_second": 15.411,
68
+ "eval_steps_per_second": 1.936,
69
+ "step": 2000
70
+ },
71
+ {
72
+ "epoch": 0.2666382252559727,
73
+ "grad_norm": 328.06475830078125,
74
+ "learning_rate": 9.111205915813424e-08,
75
+ "loss": 0.3056,
76
+ "step": 2500
77
+ },
78
+ {
79
+ "epoch": 0.2666382252559727,
80
+ "eval_loss": 1.1090214252471924,
81
+ "eval_runtime": 63.2171,
82
+ "eval_samples_per_second": 15.613,
83
+ "eval_steps_per_second": 1.961,
84
+ "step": 2500
85
+ },
86
+ {
87
+ "epoch": 0.3199658703071672,
88
+ "grad_norm": 0.00026008329587057233,
89
+ "learning_rate": 8.933447098976109e-08,
90
+ "loss": 0.2453,
91
+ "step": 3000
92
+ },
93
+ {
94
+ "epoch": 0.3199658703071672,
95
+ "eval_loss": 1.0675184726715088,
96
+ "eval_runtime": 63.9757,
97
+ "eval_samples_per_second": 15.428,
98
+ "eval_steps_per_second": 1.938,
99
+ "step": 3000
100
+ },
101
+ {
102
+ "epoch": 0.37329351535836175,
103
+ "grad_norm": 7.824947834014893,
104
+ "learning_rate": 8.755688282138793e-08,
105
+ "loss": 0.2721,
106
+ "step": 3500
107
+ },
108
+ {
109
+ "epoch": 0.37329351535836175,
110
+ "eval_loss": 1.0416266918182373,
111
+ "eval_runtime": 63.8858,
112
+ "eval_samples_per_second": 15.449,
113
+ "eval_steps_per_second": 1.941,
114
+ "step": 3500
115
+ },
116
+ {
117
+ "epoch": 0.42662116040955633,
118
+ "grad_norm": 4.335076332092285,
119
+ "learning_rate": 8.57792946530148e-08,
120
+ "loss": 0.1925,
121
+ "step": 4000
122
+ },
123
+ {
124
+ "epoch": 0.42662116040955633,
125
+ "eval_loss": 1.0176074504852295,
126
+ "eval_runtime": 64.023,
127
+ "eval_samples_per_second": 15.416,
128
+ "eval_steps_per_second": 1.937,
129
+ "step": 4000
130
+ }
131
+ ],
132
+ "logging_steps": 500,
133
+ "max_steps": 28128,
134
+ "num_input_tokens_seen": 0,
135
+ "num_train_epochs": 3,
136
+ "save_steps": 500,
137
+ "stateful_callbacks": {
138
+ "TrainerControl": {
139
+ "args": {
140
+ "should_epoch_stop": false,
141
+ "should_evaluate": false,
142
+ "should_log": false,
143
+ "should_save": true,
144
+ "should_training_stop": false
145
+ },
146
+ "attributes": {}
147
+ }
148
+ },
149
+ "total_flos": 1439752921742520.0,
150
+ "train_batch_size": 2,
151
+ "trial_name": null,
152
+ "trial_params": null
153
+ }
checkpoint-4000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b52c451dc1fefed1a64b0a23a69d323685c58d3023f0abb7cb18bc82e1be68c7
3
+ size 4847