Training in progress, step 10000
Browse files- .gitignore +1 -0
- config.json +31 -0
- last-checkpoint/config.json +31 -0
- last-checkpoint/optimizer.pt +3 -0
- last-checkpoint/pytorch_model.bin +3 -0
- last-checkpoint/rng_state.pth +3 -0
- last-checkpoint/scaler.pt +3 -0
- last-checkpoint/scheduler.pt +3 -0
- last-checkpoint/trainer_state.json +152 -0
- last-checkpoint/training_args.bin +3 -0
- pytorch_model.bin +3 -0
- training_args.bin +3 -0
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
checkpoint-*/
|
config.json
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "Nadav/PretrainedPHD",
|
3 |
+
"architectures": [
|
4 |
+
"PIXELForPreTraining"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"decoder_hidden_size": 512,
|
8 |
+
"decoder_intermediate_size": 2048,
|
9 |
+
"decoder_num_attention_heads": 16,
|
10 |
+
"decoder_num_hidden_layers": 8,
|
11 |
+
"hidden_act": "gelu",
|
12 |
+
"hidden_dropout_prob": 0.1,
|
13 |
+
"hidden_size": 768,
|
14 |
+
"image_size": [
|
15 |
+
368,
|
16 |
+
368
|
17 |
+
],
|
18 |
+
"initializer_range": 0.02,
|
19 |
+
"intermediate_size": 3072,
|
20 |
+
"layer_norm_eps": 1e-12,
|
21 |
+
"mask_ratio": 0.25,
|
22 |
+
"model_type": "pixel",
|
23 |
+
"norm_pix_loss": true,
|
24 |
+
"num_attention_heads": 12,
|
25 |
+
"num_channels": 3,
|
26 |
+
"num_hidden_layers": 12,
|
27 |
+
"patch_size": 16,
|
28 |
+
"qkv_bias": true,
|
29 |
+
"torch_dtype": "float32",
|
30 |
+
"transformers_version": "4.17.0"
|
31 |
+
}
|
last-checkpoint/config.json
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "Nadav/PretrainedPHD",
|
3 |
+
"architectures": [
|
4 |
+
"PIXELForPreTraining"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"decoder_hidden_size": 512,
|
8 |
+
"decoder_intermediate_size": 2048,
|
9 |
+
"decoder_num_attention_heads": 16,
|
10 |
+
"decoder_num_hidden_layers": 8,
|
11 |
+
"hidden_act": "gelu",
|
12 |
+
"hidden_dropout_prob": 0.1,
|
13 |
+
"hidden_size": 768,
|
14 |
+
"image_size": [
|
15 |
+
368,
|
16 |
+
368
|
17 |
+
],
|
18 |
+
"initializer_range": 0.02,
|
19 |
+
"intermediate_size": 3072,
|
20 |
+
"layer_norm_eps": 1e-12,
|
21 |
+
"mask_ratio": 0.25,
|
22 |
+
"model_type": "pixel",
|
23 |
+
"norm_pix_loss": true,
|
24 |
+
"num_attention_heads": 12,
|
25 |
+
"num_channels": 3,
|
26 |
+
"num_hidden_layers": 12,
|
27 |
+
"patch_size": 16,
|
28 |
+
"qkv_bias": true,
|
29 |
+
"torch_dtype": "float32",
|
30 |
+
"transformers_version": "4.17.0"
|
31 |
+
}
|
last-checkpoint/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b7965023fb90058fbcca08a10a8e9e39d8cf72c5b9f3d0eb2b25bc715dfb7185
|
3 |
+
size 893438545
|
last-checkpoint/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5abd226a9e0cc358d3b07162345d74dc2fab3322999fae6a6de0ab8221263680
|
3 |
+
size 449471589
|
last-checkpoint/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8217c6a790474956dac035c8fd3d42f3d276f02a592aa421e5eed682cc08b9be
|
3 |
+
size 15523
|
last-checkpoint/scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:46171f332d1e5155156d4999fec5406fc9f03857ff686d0152b9c4ca38519697
|
3 |
+
size 559
|
last-checkpoint/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:611657e4e82b38f099b149471e6aaaa5a577926bec8385cc0f6dd743ea4cfd9b
|
3 |
+
size 623
|
last-checkpoint/trainer_state.json
ADDED
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.870852564660803,
|
5 |
+
"global_step": 10000,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.04,
|
12 |
+
"learning_rate": 0.00010077565027123787,
|
13 |
+
"loss": 0.55,
|
14 |
+
"step": 500
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.09,
|
18 |
+
"learning_rate": 0.00010075885246660077,
|
19 |
+
"loss": 0.5027,
|
20 |
+
"step": 1000
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.13,
|
24 |
+
"learning_rate": 0.00010073086073069567,
|
25 |
+
"loss": 0.4909,
|
26 |
+
"step": 1500
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.17,
|
30 |
+
"learning_rate": 0.0001006916819700645,
|
31 |
+
"loss": 0.4828,
|
32 |
+
"step": 2000
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.22,
|
36 |
+
"learning_rate": 0.00010064132585148025,
|
37 |
+
"loss": 0.4791,
|
38 |
+
"step": 2500
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.26,
|
42 |
+
"learning_rate": 0.00010057980479956167,
|
43 |
+
"loss": 0.475,
|
44 |
+
"step": 3000
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.3,
|
48 |
+
"learning_rate": 0.00010050713399370776,
|
49 |
+
"loss": 0.4706,
|
50 |
+
"step": 3500
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.35,
|
54 |
+
"learning_rate": 0.00010042333136435246,
|
55 |
+
"loss": 0.4686,
|
56 |
+
"step": 4000
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 0.39,
|
60 |
+
"learning_rate": 0.0001003284175885406,
|
61 |
+
"loss": 0.4662,
|
62 |
+
"step": 4500
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 0.44,
|
66 |
+
"learning_rate": 0.00010022241608482616,
|
67 |
+
"loss": 0.4639,
|
68 |
+
"step": 5000
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 0.44,
|
72 |
+
"eval_loss": 0.4434024393558502,
|
73 |
+
"eval_runtime": 58.4628,
|
74 |
+
"eval_samples_per_second": 85.524,
|
75 |
+
"eval_steps_per_second": 1.351,
|
76 |
+
"step": 5000
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 0.48,
|
80 |
+
"learning_rate": 0.00010010535300749402,
|
81 |
+
"loss": 0.4617,
|
82 |
+
"step": 5500
|
83 |
+
},
|
84 |
+
{
|
85 |
+
"epoch": 0.52,
|
86 |
+
"learning_rate": 9.997725724010688e-05,
|
87 |
+
"loss": 0.4605,
|
88 |
+
"step": 6000
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 0.57,
|
92 |
+
"learning_rate": 9.983816038837858e-05,
|
93 |
+
"loss": 0.4601,
|
94 |
+
"step": 6500
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.61,
|
98 |
+
"learning_rate": 9.968809677237595e-05,
|
99 |
+
"loss": 0.4567,
|
100 |
+
"step": 7000
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.65,
|
104 |
+
"learning_rate": 9.952710341805077e-05,
|
105 |
+
"loss": 0.4573,
|
106 |
+
"step": 7500
|
107 |
+
},
|
108 |
+
{
|
109 |
+
"epoch": 0.7,
|
110 |
+
"learning_rate": 9.935522004810426e-05,
|
111 |
+
"loss": 0.4558,
|
112 |
+
"step": 8000
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"epoch": 0.74,
|
116 |
+
"learning_rate": 9.91728653305258e-05,
|
117 |
+
"loss": 0.4534,
|
118 |
+
"step": 8500
|
119 |
+
},
|
120 |
+
{
|
121 |
+
"epoch": 0.78,
|
122 |
+
"learning_rate": 9.897935339304003e-05,
|
123 |
+
"loss": 0.4529,
|
124 |
+
"step": 9000
|
125 |
+
},
|
126 |
+
{
|
127 |
+
"epoch": 0.83,
|
128 |
+
"learning_rate": 9.87750865890554e-05,
|
129 |
+
"loss": 0.4511,
|
130 |
+
"step": 9500
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"epoch": 0.87,
|
134 |
+
"learning_rate": 9.856055590913237e-05,
|
135 |
+
"loss": 0.45,
|
136 |
+
"step": 10000
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.87,
|
140 |
+
"eval_loss": 0.42864668369293213,
|
141 |
+
"eval_runtime": 74.102,
|
142 |
+
"eval_samples_per_second": 67.475,
|
143 |
+
"eval_steps_per_second": 1.066,
|
144 |
+
"step": 10000
|
145 |
+
}
|
146 |
+
],
|
147 |
+
"max_steps": 100000,
|
148 |
+
"num_train_epochs": 9,
|
149 |
+
"total_flos": 4.709861347295232e+20,
|
150 |
+
"trial_name": null,
|
151 |
+
"trial_params": null
|
152 |
+
}
|
last-checkpoint/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5338e8a2c0a904652b6d8f7e355dcea8351d2f08d02d32f0f504d2e3f53f6aa5
|
3 |
+
size 5551
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5abd226a9e0cc358d3b07162345d74dc2fab3322999fae6a6de0ab8221263680
|
3 |
+
size 449471589
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5338e8a2c0a904652b6d8f7e355dcea8351d2f08d02d32f0f504d2e3f53f6aa5
|
3 |
+
size 5551
|