jnmrr commited on
Commit
d5038e0
·
verified ·
1 Parent(s): 3a530fe

Upload RT-DETRv2 voucher classifier

Browse files
Files changed (44) hide show
  1. README.md +15 -15
  2. checkpoint-3520/config.json +129 -0
  3. checkpoint-3520/model.safetensors +3 -0
  4. checkpoint-3520/optimizer.pt +3 -0
  5. checkpoint-3520/preprocessor_config.json +26 -0
  6. checkpoint-3520/rng_state.pth +3 -0
  7. checkpoint-3520/scheduler.pt +3 -0
  8. checkpoint-3520/trainer_state.json +482 -0
  9. checkpoint-3520/training_args.bin +3 -0
  10. checkpoint-3740/config.json +129 -0
  11. checkpoint-3740/model.safetensors +3 -0
  12. checkpoint-3740/optimizer.pt +3 -0
  13. checkpoint-3740/preprocessor_config.json +26 -0
  14. checkpoint-3740/rng_state.pth +3 -0
  15. checkpoint-3740/scheduler.pt +3 -0
  16. checkpoint-3740/trainer_state.json +510 -0
  17. checkpoint-3740/training_args.bin +3 -0
  18. checkpoint-3960/config.json +129 -0
  19. checkpoint-3960/model.safetensors +3 -0
  20. checkpoint-3960/optimizer.pt +3 -0
  21. checkpoint-3960/preprocessor_config.json +26 -0
  22. checkpoint-3960/rng_state.pth +3 -0
  23. checkpoint-3960/scheduler.pt +3 -0
  24. checkpoint-3960/trainer_state.json +538 -0
  25. checkpoint-3960/training_args.bin +3 -0
  26. checkpoint-4180/config.json +129 -0
  27. checkpoint-4180/model.safetensors +3 -0
  28. checkpoint-4180/optimizer.pt +3 -0
  29. checkpoint-4180/preprocessor_config.json +26 -0
  30. checkpoint-4180/rng_state.pth +3 -0
  31. checkpoint-4180/scheduler.pt +3 -0
  32. checkpoint-4180/trainer_state.json +566 -0
  33. checkpoint-4180/training_args.bin +3 -0
  34. checkpoint-4235/config.json +129 -0
  35. checkpoint-4235/model.safetensors +3 -0
  36. checkpoint-4235/optimizer.pt +3 -0
  37. checkpoint-4235/preprocessor_config.json +26 -0
  38. checkpoint-4235/rng_state.pth +3 -0
  39. checkpoint-4235/scheduler.pt +3 -0
  40. checkpoint-4235/trainer_state.json +573 -0
  41. checkpoint-4235/training_args.bin +3 -0
  42. config.json +1 -1
  43. model.safetensors +1 -1
  44. training_args.bin +2 -2
README.md CHANGED
@@ -44,11 +44,11 @@ This model is a fine-tuned version of [PekingU/rtdetr_v2_r101vd](https://hugging
44
 
45
 
46
  **Training Configuration:**
47
- - **Image Size**: 800x800
48
- - **Batch Size**: 24
49
- - **Learning Rate**: 1.5e-05
50
  - **Weight Decay**: 0.0001
51
- - **Epochs**: 50
52
  - **Validation Split**: 0.0
53
 
54
  **Data Processing:**
@@ -104,26 +104,26 @@ This model is a fine-tuned version of [PekingU/rtdetr_v2_r101vd](https://hugging
104
  **Model Configuration:**
105
  - **Base model**: PekingU/rtdetr_v2_r101vd
106
  - **Architecture**: rtdetr_v2_r101vd
107
- - **Input resolution**: 800×800 pixels
108
- - **Training epochs**: 50
109
- - **Batch size**: 24
110
 
111
  **Training Hardware:**
112
- - **GPU**: NVIDIA A100-SXM4-40GB
113
- - **VRAM**: 39.6 GB
114
- - **RAM**: 83.5 GB
115
- - **GPU configuration**: A100 optimized
116
 
117
- **Training Time**: 1.18 hours
118
 
119
  **Training Summary:**
120
- - **Final training loss**: 196.0164
121
- - **Final learning rate**: 6.00e-08
122
 
123
 
124
  ### MLflow Tracking
125
 
126
- - **MLflow Run ID**: fe7bd26bd1b344c292d9b485139be46c
127
  - **MLflow Experiment**: RT-DETRv2_Voucher_Classification
128
 
129
 
 
44
 
45
 
46
  **Training Configuration:**
47
+ - **Image Size**: 832x832
48
+ - **Batch Size**: 28
49
+ - **Learning Rate**: 1.4e-05
50
  - **Weight Decay**: 0.0001
51
+ - **Epochs**: 55
52
  - **Validation Split**: 0.0
53
 
54
  **Data Processing:**
 
104
  **Model Configuration:**
105
  - **Base model**: PekingU/rtdetr_v2_r101vd
106
  - **Architecture**: rtdetr_v2_r101vd
107
+ - **Input resolution**: 832×832 pixels
108
+ - **Training epochs**: 55
109
+ - **Batch size**: 28
110
 
111
  **Training Hardware:**
112
+ - **GPU**: NVIDIA L40S
113
+ - **VRAM**: 44.4 GB
114
+ - **RAM**: 62.8 GB
115
+ - **GPU configuration**: L40S optimized
116
 
117
+ **Training Time**: 1.36 hours
118
 
119
  **Training Summary:**
120
+ - **Final training loss**: 120.5679
121
+ - **Final learning rate**: 6.88e-09
122
 
123
 
124
  ### MLflow Tracking
125
 
126
+ - **MLflow Run ID**: f6bc3ebbbaa44f88bf184d07d93356ff
127
  - **MLflow Experiment**: RT-DETRv2_Voucher_Classification
128
 
129
 
checkpoint-3520/config.json ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "silu",
4
+ "anchor_image_size": null,
5
+ "architectures": [
6
+ "RTDetrV2ForObjectDetection"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auxiliary_loss": true,
10
+ "backbone": null,
11
+ "backbone_config": {
12
+ "depths": [
13
+ 3,
14
+ 4,
15
+ 23,
16
+ 3
17
+ ],
18
+ "downsample_in_bottleneck": false,
19
+ "downsample_in_first_stage": false,
20
+ "embedding_size": 64,
21
+ "hidden_act": "relu",
22
+ "hidden_sizes": [
23
+ 256,
24
+ 512,
25
+ 1024,
26
+ 2048
27
+ ],
28
+ "layer_type": "bottleneck",
29
+ "model_type": "rt_detr_resnet",
30
+ "num_channels": 3,
31
+ "out_features": [
32
+ "stage2",
33
+ "stage3",
34
+ "stage4"
35
+ ],
36
+ "out_indices": [
37
+ 2,
38
+ 3,
39
+ 4
40
+ ],
41
+ "stage_names": [
42
+ "stem",
43
+ "stage1",
44
+ "stage2",
45
+ "stage3",
46
+ "stage4"
47
+ ],
48
+ "torch_dtype": "float32"
49
+ },
50
+ "backbone_kwargs": null,
51
+ "batch_norm_eps": 1e-05,
52
+ "box_noise_scale": 1.0,
53
+ "d_model": 256,
54
+ "decoder_activation_function": "relu",
55
+ "decoder_attention_heads": 8,
56
+ "decoder_ffn_dim": 1024,
57
+ "decoder_in_channels": [
58
+ 384,
59
+ 384,
60
+ 384
61
+ ],
62
+ "decoder_layers": 6,
63
+ "decoder_method": "default",
64
+ "decoder_n_levels": 3,
65
+ "decoder_n_points": 4,
66
+ "decoder_offset_scale": 0.5,
67
+ "disable_custom_kernels": true,
68
+ "dropout": 0.0,
69
+ "encode_proj_layers": [
70
+ 2
71
+ ],
72
+ "encoder_activation_function": "gelu",
73
+ "encoder_attention_heads": 8,
74
+ "encoder_ffn_dim": 2048,
75
+ "encoder_hidden_dim": 384,
76
+ "encoder_in_channels": [
77
+ 512,
78
+ 1024,
79
+ 2048
80
+ ],
81
+ "encoder_layers": 1,
82
+ "eos_coefficient": 0.0001,
83
+ "eval_size": null,
84
+ "feat_strides": [
85
+ 8,
86
+ 16,
87
+ 32
88
+ ],
89
+ "focal_loss_alpha": 0.75,
90
+ "focal_loss_gamma": 2.0,
91
+ "freeze_backbone_batch_norms": true,
92
+ "hidden_expansion": 1.0,
93
+ "id2label": {
94
+ "0": "LABEL_0",
95
+ "1": "LABEL_1",
96
+ "2": "LABEL_2"
97
+ },
98
+ "initializer_bias_prior_prob": null,
99
+ "initializer_range": 0.01,
100
+ "is_encoder_decoder": true,
101
+ "label2id": {
102
+ "LABEL_0": 0,
103
+ "LABEL_1": 1,
104
+ "LABEL_2": 2
105
+ },
106
+ "label_noise_ratio": 0.5,
107
+ "layer_norm_eps": 1e-05,
108
+ "learn_initial_query": false,
109
+ "matcher_alpha": 0.25,
110
+ "matcher_bbox_cost": 5.0,
111
+ "matcher_class_cost": 2.0,
112
+ "matcher_gamma": 2.0,
113
+ "matcher_giou_cost": 2.0,
114
+ "model_type": "rt_detr_v2",
115
+ "normalize_before": false,
116
+ "num_denoising": 100,
117
+ "num_feature_levels": 3,
118
+ "num_queries": 300,
119
+ "positional_encoding_temperature": 10000,
120
+ "torch_dtype": "float32",
121
+ "transformers_version": "4.55.2",
122
+ "use_focal_loss": true,
123
+ "use_pretrained_backbone": false,
124
+ "use_timm_backbone": false,
125
+ "weight_loss_bbox": 5.0,
126
+ "weight_loss_giou": 2.0,
127
+ "weight_loss_vfl": 1.0,
128
+ "with_box_refine": true
129
+ }
checkpoint-3520/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2657b101f238e879c5f5c203f9376a3578df04ea7474324a53b19093b806f48
3
+ size 306699044
checkpoint-3520/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:433457ca3308317df40d36d8b025ac71ae26460ca55e09f798a6958e52f2893e
3
+ size 611580898
checkpoint-3520/preprocessor_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_annotations": true,
3
+ "do_normalize": false,
4
+ "do_pad": false,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "format": "coco_detection",
8
+ "image_mean": [
9
+ 0.485,
10
+ 0.456,
11
+ 0.406
12
+ ],
13
+ "image_processor_type": "RTDetrImageProcessor",
14
+ "image_std": [
15
+ 0.229,
16
+ 0.224,
17
+ 0.225
18
+ ],
19
+ "pad_size": null,
20
+ "resample": 2,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "height": 640,
24
+ "width": 640
25
+ }
26
+ }
checkpoint-3520/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26d8dbc68d586ebc03d440b6f91bc263da03830b540faac657b2b67604441fdb
3
+ size 14645
checkpoint-3520/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:36b280336c7f89ab668467fe9fe6ee1aa85c169847bd40455524a9bf18eea5dc
3
+ size 1465
checkpoint-3520/trainer_state.json ADDED
@@ -0,0 +1,482 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 45.714285714285715,
6
+ "eval_steps": 500,
7
+ "global_step": 3520,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.7142857142857143,
14
+ "grad_norm": 4405.81884765625,
15
+ "learning_rate": 3.436363636363636e-07,
16
+ "loss": 1357.5645,
17
+ "step": 55
18
+ },
19
+ {
20
+ "epoch": 1.4285714285714286,
21
+ "grad_norm": 5894.37841796875,
22
+ "learning_rate": 6.936363636363637e-07,
23
+ "loss": 1290.9688,
24
+ "step": 110
25
+ },
26
+ {
27
+ "epoch": 2.142857142857143,
28
+ "grad_norm": 3667.852294921875,
29
+ "learning_rate": 1.0436363636363635e-06,
30
+ "loss": 1140.9515,
31
+ "step": 165
32
+ },
33
+ {
34
+ "epoch": 2.857142857142857,
35
+ "grad_norm": 3578.294677734375,
36
+ "learning_rate": 1.3936363636363637e-06,
37
+ "loss": 942.618,
38
+ "step": 220
39
+ },
40
+ {
41
+ "epoch": 3.571428571428571,
42
+ "grad_norm": 2097.637939453125,
43
+ "learning_rate": 1.7436363636363636e-06,
44
+ "loss": 721.2856,
45
+ "step": 275
46
+ },
47
+ {
48
+ "epoch": 4.285714285714286,
49
+ "grad_norm": 1840.612548828125,
50
+ "learning_rate": 2.0936363636363636e-06,
51
+ "loss": 525.8542,
52
+ "step": 330
53
+ },
54
+ {
55
+ "epoch": 5.0,
56
+ "grad_norm": 1230.875,
57
+ "learning_rate": 2.4436363636363636e-06,
58
+ "loss": 397.7722,
59
+ "step": 385
60
+ },
61
+ {
62
+ "epoch": 5.714285714285714,
63
+ "grad_norm": 1280.699951171875,
64
+ "learning_rate": 2.7936363636363637e-06,
65
+ "loss": 313.8322,
66
+ "step": 440
67
+ },
68
+ {
69
+ "epoch": 6.428571428571429,
70
+ "grad_norm": 4265.28759765625,
71
+ "learning_rate": 3.1436363636363637e-06,
72
+ "loss": 254.1449,
73
+ "step": 495
74
+ },
75
+ {
76
+ "epoch": 7.142857142857143,
77
+ "grad_norm": 967.1757202148438,
78
+ "learning_rate": 3.4936363636363633e-06,
79
+ "loss": 218.4529,
80
+ "step": 550
81
+ },
82
+ {
83
+ "epoch": 7.857142857142857,
84
+ "grad_norm": 765.2302856445312,
85
+ "learning_rate": 3.843636363636364e-06,
86
+ "loss": 197.915,
87
+ "step": 605
88
+ },
89
+ {
90
+ "epoch": 8.571428571428571,
91
+ "grad_norm": 700.331787109375,
92
+ "learning_rate": 4.193636363636364e-06,
93
+ "loss": 170.9019,
94
+ "step": 660
95
+ },
96
+ {
97
+ "epoch": 9.285714285714286,
98
+ "grad_norm": 613.8758544921875,
99
+ "learning_rate": 4.543636363636363e-06,
100
+ "loss": 154.6195,
101
+ "step": 715
102
+ },
103
+ {
104
+ "epoch": 10.0,
105
+ "grad_norm": 3405.25732421875,
106
+ "learning_rate": 4.893636363636364e-06,
107
+ "loss": 136.8394,
108
+ "step": 770
109
+ },
110
+ {
111
+ "epoch": 10.714285714285714,
112
+ "grad_norm": 1219.8994140625,
113
+ "learning_rate": 5.243636363636364e-06,
114
+ "loss": 116.0246,
115
+ "step": 825
116
+ },
117
+ {
118
+ "epoch": 11.428571428571429,
119
+ "grad_norm": 306.2723083496094,
120
+ "learning_rate": 5.593636363636363e-06,
121
+ "loss": 104.2055,
122
+ "step": 880
123
+ },
124
+ {
125
+ "epoch": 12.142857142857142,
126
+ "grad_norm": 288.2547302246094,
127
+ "learning_rate": 5.943636363636364e-06,
128
+ "loss": 88.9812,
129
+ "step": 935
130
+ },
131
+ {
132
+ "epoch": 12.857142857142858,
133
+ "grad_norm": 1016.5924682617188,
134
+ "learning_rate": 6.293636363636363e-06,
135
+ "loss": 79.9017,
136
+ "step": 990
137
+ },
138
+ {
139
+ "epoch": 13.571428571428571,
140
+ "grad_norm": 239.49191284179688,
141
+ "learning_rate": 6.643636363636363e-06,
142
+ "loss": 73.6932,
143
+ "step": 1045
144
+ },
145
+ {
146
+ "epoch": 14.285714285714286,
147
+ "grad_norm": 193.92002868652344,
148
+ "learning_rate": 6.993636363636364e-06,
149
+ "loss": 62.2018,
150
+ "step": 1100
151
+ },
152
+ {
153
+ "epoch": 15.0,
154
+ "grad_norm": 179.0568084716797,
155
+ "learning_rate": 7.343636363636363e-06,
156
+ "loss": 54.0213,
157
+ "step": 1155
158
+ },
159
+ {
160
+ "epoch": 15.714285714285714,
161
+ "grad_norm": 166.02944946289062,
162
+ "learning_rate": 7.693636363636364e-06,
163
+ "loss": 48.1955,
164
+ "step": 1210
165
+ },
166
+ {
167
+ "epoch": 16.428571428571427,
168
+ "grad_norm": 126.53916931152344,
169
+ "learning_rate": 8.043636363636364e-06,
170
+ "loss": 43.5431,
171
+ "step": 1265
172
+ },
173
+ {
174
+ "epoch": 17.142857142857142,
175
+ "grad_norm": 145.61166381835938,
176
+ "learning_rate": 8.393636363636363e-06,
177
+ "loss": 41.4399,
178
+ "step": 1320
179
+ },
180
+ {
181
+ "epoch": 17.857142857142858,
182
+ "grad_norm": 122.2297134399414,
183
+ "learning_rate": 8.743636363636363e-06,
184
+ "loss": 35.2278,
185
+ "step": 1375
186
+ },
187
+ {
188
+ "epoch": 18.571428571428573,
189
+ "grad_norm": 117.88919067382812,
190
+ "learning_rate": 9.093636363636363e-06,
191
+ "loss": 31.2827,
192
+ "step": 1430
193
+ },
194
+ {
195
+ "epoch": 19.285714285714285,
196
+ "grad_norm": 88.52986907958984,
197
+ "learning_rate": 9.443636363636364e-06,
198
+ "loss": 28.8076,
199
+ "step": 1485
200
+ },
201
+ {
202
+ "epoch": 20.0,
203
+ "grad_norm": 88.49090576171875,
204
+ "learning_rate": 9.793636363636364e-06,
205
+ "loss": 28.232,
206
+ "step": 1540
207
+ },
208
+ {
209
+ "epoch": 20.714285714285715,
210
+ "grad_norm": 114.45001983642578,
211
+ "learning_rate": 1.0143636363636363e-05,
212
+ "loss": 26.0885,
213
+ "step": 1595
214
+ },
215
+ {
216
+ "epoch": 21.428571428571427,
217
+ "grad_norm": 74.16987609863281,
218
+ "learning_rate": 1.0493636363636363e-05,
219
+ "loss": 24.2689,
220
+ "step": 1650
221
+ },
222
+ {
223
+ "epoch": 22.142857142857142,
224
+ "grad_norm": 82.25133514404297,
225
+ "learning_rate": 1.0843636363636363e-05,
226
+ "loss": 22.6963,
227
+ "step": 1705
228
+ },
229
+ {
230
+ "epoch": 22.857142857142858,
231
+ "grad_norm": 74.76679229736328,
232
+ "learning_rate": 1.1193636363636363e-05,
233
+ "loss": 21.3257,
234
+ "step": 1760
235
+ },
236
+ {
237
+ "epoch": 23.571428571428573,
238
+ "grad_norm": 68.91163635253906,
239
+ "learning_rate": 1.1543636363636365e-05,
240
+ "loss": 20.2199,
241
+ "step": 1815
242
+ },
243
+ {
244
+ "epoch": 24.285714285714285,
245
+ "grad_norm": 59.86214828491211,
246
+ "learning_rate": 1.1893636363636363e-05,
247
+ "loss": 19.3769,
248
+ "step": 1870
249
+ },
250
+ {
251
+ "epoch": 25.0,
252
+ "grad_norm": 76.9339370727539,
253
+ "learning_rate": 1.2243636363636363e-05,
254
+ "loss": 18.0471,
255
+ "step": 1925
256
+ },
257
+ {
258
+ "epoch": 25.714285714285715,
259
+ "grad_norm": 66.16152954101562,
260
+ "learning_rate": 1.2593636363636363e-05,
261
+ "loss": 17.496,
262
+ "step": 1980
263
+ },
264
+ {
265
+ "epoch": 26.428571428571427,
266
+ "grad_norm": 58.77778244018555,
267
+ "learning_rate": 1.2943636363636363e-05,
268
+ "loss": 16.9468,
269
+ "step": 2035
270
+ },
271
+ {
272
+ "epoch": 27.142857142857142,
273
+ "grad_norm": 71.95783233642578,
274
+ "learning_rate": 1.3293636363636363e-05,
275
+ "loss": 16.3026,
276
+ "step": 2090
277
+ },
278
+ {
279
+ "epoch": 27.857142857142858,
280
+ "grad_norm": 131.7191619873047,
281
+ "learning_rate": 1.3643636363636363e-05,
282
+ "loss": 16.008,
283
+ "step": 2145
284
+ },
285
+ {
286
+ "epoch": 28.571428571428573,
287
+ "grad_norm": 81.78388977050781,
288
+ "learning_rate": 1.3993636363636363e-05,
289
+ "loss": 15.4616,
290
+ "step": 2200
291
+ },
292
+ {
293
+ "epoch": 29.285714285714285,
294
+ "grad_norm": 60.10112380981445,
295
+ "learning_rate": 1.3628501228501228e-05,
296
+ "loss": 15.195,
297
+ "step": 2255
298
+ },
299
+ {
300
+ "epoch": 30.0,
301
+ "grad_norm": 77.59001922607422,
302
+ "learning_rate": 1.325012285012285e-05,
303
+ "loss": 14.7557,
304
+ "step": 2310
305
+ },
306
+ {
307
+ "epoch": 30.714285714285715,
308
+ "grad_norm": 77.3856201171875,
309
+ "learning_rate": 1.287174447174447e-05,
310
+ "loss": 14.2851,
311
+ "step": 2365
312
+ },
313
+ {
314
+ "epoch": 31.428571428571427,
315
+ "grad_norm": 73.75492095947266,
316
+ "learning_rate": 1.2493366093366094e-05,
317
+ "loss": 13.867,
318
+ "step": 2420
319
+ },
320
+ {
321
+ "epoch": 32.142857142857146,
322
+ "grad_norm": 78.76831817626953,
323
+ "learning_rate": 1.2114987714987713e-05,
324
+ "loss": 13.6677,
325
+ "step": 2475
326
+ },
327
+ {
328
+ "epoch": 32.857142857142854,
329
+ "grad_norm": 92.99346160888672,
330
+ "learning_rate": 1.1736609336609336e-05,
331
+ "loss": 13.1851,
332
+ "step": 2530
333
+ },
334
+ {
335
+ "epoch": 33.57142857142857,
336
+ "grad_norm": 97.57978820800781,
337
+ "learning_rate": 1.1358230958230958e-05,
338
+ "loss": 13.5215,
339
+ "step": 2585
340
+ },
341
+ {
342
+ "epoch": 34.285714285714285,
343
+ "grad_norm": 63.118141174316406,
344
+ "learning_rate": 1.0979852579852579e-05,
345
+ "loss": 12.8525,
346
+ "step": 2640
347
+ },
348
+ {
349
+ "epoch": 35.0,
350
+ "grad_norm": 63.68766403198242,
351
+ "learning_rate": 1.06014742014742e-05,
352
+ "loss": 12.441,
353
+ "step": 2695
354
+ },
355
+ {
356
+ "epoch": 35.714285714285715,
357
+ "grad_norm": 70.33533477783203,
358
+ "learning_rate": 1.0223095823095823e-05,
359
+ "loss": 12.2331,
360
+ "step": 2750
361
+ },
362
+ {
363
+ "epoch": 36.42857142857143,
364
+ "grad_norm": 64.16197967529297,
365
+ "learning_rate": 9.844717444717445e-06,
366
+ "loss": 11.7547,
367
+ "step": 2805
368
+ },
369
+ {
370
+ "epoch": 37.142857142857146,
371
+ "grad_norm": 85.83612823486328,
372
+ "learning_rate": 9.466339066339066e-06,
373
+ "loss": 11.8633,
374
+ "step": 2860
375
+ },
376
+ {
377
+ "epoch": 37.857142857142854,
378
+ "grad_norm": 89.40387725830078,
379
+ "learning_rate": 9.087960687960689e-06,
380
+ "loss": 11.5795,
381
+ "step": 2915
382
+ },
383
+ {
384
+ "epoch": 38.57142857142857,
385
+ "grad_norm": 76.38651275634766,
386
+ "learning_rate": 8.709582309582309e-06,
387
+ "loss": 11.334,
388
+ "step": 2970
389
+ },
390
+ {
391
+ "epoch": 39.285714285714285,
392
+ "grad_norm": 110.11483764648438,
393
+ "learning_rate": 8.331203931203932e-06,
394
+ "loss": 10.805,
395
+ "step": 3025
396
+ },
397
+ {
398
+ "epoch": 40.0,
399
+ "grad_norm": 60.31315612792969,
400
+ "learning_rate": 7.952825552825553e-06,
401
+ "loss": 10.9056,
402
+ "step": 3080
403
+ },
404
+ {
405
+ "epoch": 40.714285714285715,
406
+ "grad_norm": 107.96589660644531,
407
+ "learning_rate": 7.574447174447175e-06,
408
+ "loss": 10.5828,
409
+ "step": 3135
410
+ },
411
+ {
412
+ "epoch": 41.42857142857143,
413
+ "grad_norm": 74.8004379272461,
414
+ "learning_rate": 7.1960687960687955e-06,
415
+ "loss": 10.2818,
416
+ "step": 3190
417
+ },
418
+ {
419
+ "epoch": 42.142857142857146,
420
+ "grad_norm": 54.093475341796875,
421
+ "learning_rate": 6.817690417690418e-06,
422
+ "loss": 10.0785,
423
+ "step": 3245
424
+ },
425
+ {
426
+ "epoch": 42.857142857142854,
427
+ "grad_norm": 115.28564453125,
428
+ "learning_rate": 6.439312039312039e-06,
429
+ "loss": 9.8794,
430
+ "step": 3300
431
+ },
432
+ {
433
+ "epoch": 43.57142857142857,
434
+ "grad_norm": 52.57551956176758,
435
+ "learning_rate": 6.06093366093366e-06,
436
+ "loss": 9.7072,
437
+ "step": 3355
438
+ },
439
+ {
440
+ "epoch": 44.285714285714285,
441
+ "grad_norm": 51.994075775146484,
442
+ "learning_rate": 5.682555282555282e-06,
443
+ "loss": 9.6015,
444
+ "step": 3410
445
+ },
446
+ {
447
+ "epoch": 45.0,
448
+ "grad_norm": 73.08104705810547,
449
+ "learning_rate": 5.304176904176904e-06,
450
+ "loss": 9.305,
451
+ "step": 3465
452
+ },
453
+ {
454
+ "epoch": 45.714285714285715,
455
+ "grad_norm": 75.5807876586914,
456
+ "learning_rate": 4.925798525798525e-06,
457
+ "loss": 9.2281,
458
+ "step": 3520
459
+ }
460
+ ],
461
+ "logging_steps": 55,
462
+ "max_steps": 4235,
463
+ "num_input_tokens_seen": 0,
464
+ "num_train_epochs": 55,
465
+ "save_steps": 220,
466
+ "stateful_callbacks": {
467
+ "TrainerControl": {
468
+ "args": {
469
+ "should_epoch_stop": false,
470
+ "should_evaluate": false,
471
+ "should_log": false,
472
+ "should_save": true,
473
+ "should_training_stop": false
474
+ },
475
+ "attributes": {}
476
+ }
477
+ },
478
+ "total_flos": 5.541350580943258e+19,
479
+ "train_batch_size": 28,
480
+ "trial_name": null,
481
+ "trial_params": null
482
+ }
checkpoint-3520/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99e55f6c266fb80fff7a194cd0925963d93d119d60f4443cb609bfd824b16fd5
3
+ size 5841
checkpoint-3740/config.json ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "silu",
4
+ "anchor_image_size": null,
5
+ "architectures": [
6
+ "RTDetrV2ForObjectDetection"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auxiliary_loss": true,
10
+ "backbone": null,
11
+ "backbone_config": {
12
+ "depths": [
13
+ 3,
14
+ 4,
15
+ 23,
16
+ 3
17
+ ],
18
+ "downsample_in_bottleneck": false,
19
+ "downsample_in_first_stage": false,
20
+ "embedding_size": 64,
21
+ "hidden_act": "relu",
22
+ "hidden_sizes": [
23
+ 256,
24
+ 512,
25
+ 1024,
26
+ 2048
27
+ ],
28
+ "layer_type": "bottleneck",
29
+ "model_type": "rt_detr_resnet",
30
+ "num_channels": 3,
31
+ "out_features": [
32
+ "stage2",
33
+ "stage3",
34
+ "stage4"
35
+ ],
36
+ "out_indices": [
37
+ 2,
38
+ 3,
39
+ 4
40
+ ],
41
+ "stage_names": [
42
+ "stem",
43
+ "stage1",
44
+ "stage2",
45
+ "stage3",
46
+ "stage4"
47
+ ],
48
+ "torch_dtype": "float32"
49
+ },
50
+ "backbone_kwargs": null,
51
+ "batch_norm_eps": 1e-05,
52
+ "box_noise_scale": 1.0,
53
+ "d_model": 256,
54
+ "decoder_activation_function": "relu",
55
+ "decoder_attention_heads": 8,
56
+ "decoder_ffn_dim": 1024,
57
+ "decoder_in_channels": [
58
+ 384,
59
+ 384,
60
+ 384
61
+ ],
62
+ "decoder_layers": 6,
63
+ "decoder_method": "default",
64
+ "decoder_n_levels": 3,
65
+ "decoder_n_points": 4,
66
+ "decoder_offset_scale": 0.5,
67
+ "disable_custom_kernels": true,
68
+ "dropout": 0.0,
69
+ "encode_proj_layers": [
70
+ 2
71
+ ],
72
+ "encoder_activation_function": "gelu",
73
+ "encoder_attention_heads": 8,
74
+ "encoder_ffn_dim": 2048,
75
+ "encoder_hidden_dim": 384,
76
+ "encoder_in_channels": [
77
+ 512,
78
+ 1024,
79
+ 2048
80
+ ],
81
+ "encoder_layers": 1,
82
+ "eos_coefficient": 0.0001,
83
+ "eval_size": null,
84
+ "feat_strides": [
85
+ 8,
86
+ 16,
87
+ 32
88
+ ],
89
+ "focal_loss_alpha": 0.75,
90
+ "focal_loss_gamma": 2.0,
91
+ "freeze_backbone_batch_norms": true,
92
+ "hidden_expansion": 1.0,
93
+ "id2label": {
94
+ "0": "LABEL_0",
95
+ "1": "LABEL_1",
96
+ "2": "LABEL_2"
97
+ },
98
+ "initializer_bias_prior_prob": null,
99
+ "initializer_range": 0.01,
100
+ "is_encoder_decoder": true,
101
+ "label2id": {
102
+ "LABEL_0": 0,
103
+ "LABEL_1": 1,
104
+ "LABEL_2": 2
105
+ },
106
+ "label_noise_ratio": 0.5,
107
+ "layer_norm_eps": 1e-05,
108
+ "learn_initial_query": false,
109
+ "matcher_alpha": 0.25,
110
+ "matcher_bbox_cost": 5.0,
111
+ "matcher_class_cost": 2.0,
112
+ "matcher_gamma": 2.0,
113
+ "matcher_giou_cost": 2.0,
114
+ "model_type": "rt_detr_v2",
115
+ "normalize_before": false,
116
+ "num_denoising": 100,
117
+ "num_feature_levels": 3,
118
+ "num_queries": 300,
119
+ "positional_encoding_temperature": 10000,
120
+ "torch_dtype": "float32",
121
+ "transformers_version": "4.55.2",
122
+ "use_focal_loss": true,
123
+ "use_pretrained_backbone": false,
124
+ "use_timm_backbone": false,
125
+ "weight_loss_bbox": 5.0,
126
+ "weight_loss_giou": 2.0,
127
+ "weight_loss_vfl": 1.0,
128
+ "with_box_refine": true
129
+ }
checkpoint-3740/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a2115dbeecbf83273dba36b4193b8f3570757801340eaeb7c866bfbffb09037
3
+ size 306699044
checkpoint-3740/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f365a587dc2edb4b370686f19f37461e6aafc04d8a63d69d08c1fac32defe3f
3
+ size 611580898
checkpoint-3740/preprocessor_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_annotations": true,
3
+ "do_normalize": false,
4
+ "do_pad": false,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "format": "coco_detection",
8
+ "image_mean": [
9
+ 0.485,
10
+ 0.456,
11
+ 0.406
12
+ ],
13
+ "image_processor_type": "RTDetrImageProcessor",
14
+ "image_std": [
15
+ 0.229,
16
+ 0.224,
17
+ 0.225
18
+ ],
19
+ "pad_size": null,
20
+ "resample": 2,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "height": 640,
24
+ "width": 640
25
+ }
26
+ }
checkpoint-3740/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6316449de2d4e20c95cd8aa4a3ea3cbf53eb2fa90c64079ff35a7c512a4b676
3
+ size 14645
checkpoint-3740/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6406f311d941e8c7c17e04f9170710735679be7f5cfce6463778674b141d08ce
3
+ size 1465
checkpoint-3740/trainer_state.json ADDED
@@ -0,0 +1,510 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 48.57142857142857,
6
+ "eval_steps": 500,
7
+ "global_step": 3740,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.7142857142857143,
14
+ "grad_norm": 4405.81884765625,
15
+ "learning_rate": 3.436363636363636e-07,
16
+ "loss": 1357.5645,
17
+ "step": 55
18
+ },
19
+ {
20
+ "epoch": 1.4285714285714286,
21
+ "grad_norm": 5894.37841796875,
22
+ "learning_rate": 6.936363636363637e-07,
23
+ "loss": 1290.9688,
24
+ "step": 110
25
+ },
26
+ {
27
+ "epoch": 2.142857142857143,
28
+ "grad_norm": 3667.852294921875,
29
+ "learning_rate": 1.0436363636363635e-06,
30
+ "loss": 1140.9515,
31
+ "step": 165
32
+ },
33
+ {
34
+ "epoch": 2.857142857142857,
35
+ "grad_norm": 3578.294677734375,
36
+ "learning_rate": 1.3936363636363637e-06,
37
+ "loss": 942.618,
38
+ "step": 220
39
+ },
40
+ {
41
+ "epoch": 3.571428571428571,
42
+ "grad_norm": 2097.637939453125,
43
+ "learning_rate": 1.7436363636363636e-06,
44
+ "loss": 721.2856,
45
+ "step": 275
46
+ },
47
+ {
48
+ "epoch": 4.285714285714286,
49
+ "grad_norm": 1840.612548828125,
50
+ "learning_rate": 2.0936363636363636e-06,
51
+ "loss": 525.8542,
52
+ "step": 330
53
+ },
54
+ {
55
+ "epoch": 5.0,
56
+ "grad_norm": 1230.875,
57
+ "learning_rate": 2.4436363636363636e-06,
58
+ "loss": 397.7722,
59
+ "step": 385
60
+ },
61
+ {
62
+ "epoch": 5.714285714285714,
63
+ "grad_norm": 1280.699951171875,
64
+ "learning_rate": 2.7936363636363637e-06,
65
+ "loss": 313.8322,
66
+ "step": 440
67
+ },
68
+ {
69
+ "epoch": 6.428571428571429,
70
+ "grad_norm": 4265.28759765625,
71
+ "learning_rate": 3.1436363636363637e-06,
72
+ "loss": 254.1449,
73
+ "step": 495
74
+ },
75
+ {
76
+ "epoch": 7.142857142857143,
77
+ "grad_norm": 967.1757202148438,
78
+ "learning_rate": 3.4936363636363633e-06,
79
+ "loss": 218.4529,
80
+ "step": 550
81
+ },
82
+ {
83
+ "epoch": 7.857142857142857,
84
+ "grad_norm": 765.2302856445312,
85
+ "learning_rate": 3.843636363636364e-06,
86
+ "loss": 197.915,
87
+ "step": 605
88
+ },
89
+ {
90
+ "epoch": 8.571428571428571,
91
+ "grad_norm": 700.331787109375,
92
+ "learning_rate": 4.193636363636364e-06,
93
+ "loss": 170.9019,
94
+ "step": 660
95
+ },
96
+ {
97
+ "epoch": 9.285714285714286,
98
+ "grad_norm": 613.8758544921875,
99
+ "learning_rate": 4.543636363636363e-06,
100
+ "loss": 154.6195,
101
+ "step": 715
102
+ },
103
+ {
104
+ "epoch": 10.0,
105
+ "grad_norm": 3405.25732421875,
106
+ "learning_rate": 4.893636363636364e-06,
107
+ "loss": 136.8394,
108
+ "step": 770
109
+ },
110
+ {
111
+ "epoch": 10.714285714285714,
112
+ "grad_norm": 1219.8994140625,
113
+ "learning_rate": 5.243636363636364e-06,
114
+ "loss": 116.0246,
115
+ "step": 825
116
+ },
117
+ {
118
+ "epoch": 11.428571428571429,
119
+ "grad_norm": 306.2723083496094,
120
+ "learning_rate": 5.593636363636363e-06,
121
+ "loss": 104.2055,
122
+ "step": 880
123
+ },
124
+ {
125
+ "epoch": 12.142857142857142,
126
+ "grad_norm": 288.2547302246094,
127
+ "learning_rate": 5.943636363636364e-06,
128
+ "loss": 88.9812,
129
+ "step": 935
130
+ },
131
+ {
132
+ "epoch": 12.857142857142858,
133
+ "grad_norm": 1016.5924682617188,
134
+ "learning_rate": 6.293636363636363e-06,
135
+ "loss": 79.9017,
136
+ "step": 990
137
+ },
138
+ {
139
+ "epoch": 13.571428571428571,
140
+ "grad_norm": 239.49191284179688,
141
+ "learning_rate": 6.643636363636363e-06,
142
+ "loss": 73.6932,
143
+ "step": 1045
144
+ },
145
+ {
146
+ "epoch": 14.285714285714286,
147
+ "grad_norm": 193.92002868652344,
148
+ "learning_rate": 6.993636363636364e-06,
149
+ "loss": 62.2018,
150
+ "step": 1100
151
+ },
152
+ {
153
+ "epoch": 15.0,
154
+ "grad_norm": 179.0568084716797,
155
+ "learning_rate": 7.343636363636363e-06,
156
+ "loss": 54.0213,
157
+ "step": 1155
158
+ },
159
+ {
160
+ "epoch": 15.714285714285714,
161
+ "grad_norm": 166.02944946289062,
162
+ "learning_rate": 7.693636363636364e-06,
163
+ "loss": 48.1955,
164
+ "step": 1210
165
+ },
166
+ {
167
+ "epoch": 16.428571428571427,
168
+ "grad_norm": 126.53916931152344,
169
+ "learning_rate": 8.043636363636364e-06,
170
+ "loss": 43.5431,
171
+ "step": 1265
172
+ },
173
+ {
174
+ "epoch": 17.142857142857142,
175
+ "grad_norm": 145.61166381835938,
176
+ "learning_rate": 8.393636363636363e-06,
177
+ "loss": 41.4399,
178
+ "step": 1320
179
+ },
180
+ {
181
+ "epoch": 17.857142857142858,
182
+ "grad_norm": 122.2297134399414,
183
+ "learning_rate": 8.743636363636363e-06,
184
+ "loss": 35.2278,
185
+ "step": 1375
186
+ },
187
+ {
188
+ "epoch": 18.571428571428573,
189
+ "grad_norm": 117.88919067382812,
190
+ "learning_rate": 9.093636363636363e-06,
191
+ "loss": 31.2827,
192
+ "step": 1430
193
+ },
194
+ {
195
+ "epoch": 19.285714285714285,
196
+ "grad_norm": 88.52986907958984,
197
+ "learning_rate": 9.443636363636364e-06,
198
+ "loss": 28.8076,
199
+ "step": 1485
200
+ },
201
+ {
202
+ "epoch": 20.0,
203
+ "grad_norm": 88.49090576171875,
204
+ "learning_rate": 9.793636363636364e-06,
205
+ "loss": 28.232,
206
+ "step": 1540
207
+ },
208
+ {
209
+ "epoch": 20.714285714285715,
210
+ "grad_norm": 114.45001983642578,
211
+ "learning_rate": 1.0143636363636363e-05,
212
+ "loss": 26.0885,
213
+ "step": 1595
214
+ },
215
+ {
216
+ "epoch": 21.428571428571427,
217
+ "grad_norm": 74.16987609863281,
218
+ "learning_rate": 1.0493636363636363e-05,
219
+ "loss": 24.2689,
220
+ "step": 1650
221
+ },
222
+ {
223
+ "epoch": 22.142857142857142,
224
+ "grad_norm": 82.25133514404297,
225
+ "learning_rate": 1.0843636363636363e-05,
226
+ "loss": 22.6963,
227
+ "step": 1705
228
+ },
229
+ {
230
+ "epoch": 22.857142857142858,
231
+ "grad_norm": 74.76679229736328,
232
+ "learning_rate": 1.1193636363636363e-05,
233
+ "loss": 21.3257,
234
+ "step": 1760
235
+ },
236
+ {
237
+ "epoch": 23.571428571428573,
238
+ "grad_norm": 68.91163635253906,
239
+ "learning_rate": 1.1543636363636365e-05,
240
+ "loss": 20.2199,
241
+ "step": 1815
242
+ },
243
+ {
244
+ "epoch": 24.285714285714285,
245
+ "grad_norm": 59.86214828491211,
246
+ "learning_rate": 1.1893636363636363e-05,
247
+ "loss": 19.3769,
248
+ "step": 1870
249
+ },
250
+ {
251
+ "epoch": 25.0,
252
+ "grad_norm": 76.9339370727539,
253
+ "learning_rate": 1.2243636363636363e-05,
254
+ "loss": 18.0471,
255
+ "step": 1925
256
+ },
257
+ {
258
+ "epoch": 25.714285714285715,
259
+ "grad_norm": 66.16152954101562,
260
+ "learning_rate": 1.2593636363636363e-05,
261
+ "loss": 17.496,
262
+ "step": 1980
263
+ },
264
+ {
265
+ "epoch": 26.428571428571427,
266
+ "grad_norm": 58.77778244018555,
267
+ "learning_rate": 1.2943636363636363e-05,
268
+ "loss": 16.9468,
269
+ "step": 2035
270
+ },
271
+ {
272
+ "epoch": 27.142857142857142,
273
+ "grad_norm": 71.95783233642578,
274
+ "learning_rate": 1.3293636363636363e-05,
275
+ "loss": 16.3026,
276
+ "step": 2090
277
+ },
278
+ {
279
+ "epoch": 27.857142857142858,
280
+ "grad_norm": 131.7191619873047,
281
+ "learning_rate": 1.3643636363636363e-05,
282
+ "loss": 16.008,
283
+ "step": 2145
284
+ },
285
+ {
286
+ "epoch": 28.571428571428573,
287
+ "grad_norm": 81.78388977050781,
288
+ "learning_rate": 1.3993636363636363e-05,
289
+ "loss": 15.4616,
290
+ "step": 2200
291
+ },
292
+ {
293
+ "epoch": 29.285714285714285,
294
+ "grad_norm": 60.10112380981445,
295
+ "learning_rate": 1.3628501228501228e-05,
296
+ "loss": 15.195,
297
+ "step": 2255
298
+ },
299
+ {
300
+ "epoch": 30.0,
301
+ "grad_norm": 77.59001922607422,
302
+ "learning_rate": 1.325012285012285e-05,
303
+ "loss": 14.7557,
304
+ "step": 2310
305
+ },
306
+ {
307
+ "epoch": 30.714285714285715,
308
+ "grad_norm": 77.3856201171875,
309
+ "learning_rate": 1.287174447174447e-05,
310
+ "loss": 14.2851,
311
+ "step": 2365
312
+ },
313
+ {
314
+ "epoch": 31.428571428571427,
315
+ "grad_norm": 73.75492095947266,
316
+ "learning_rate": 1.2493366093366094e-05,
317
+ "loss": 13.867,
318
+ "step": 2420
319
+ },
320
+ {
321
+ "epoch": 32.142857142857146,
322
+ "grad_norm": 78.76831817626953,
323
+ "learning_rate": 1.2114987714987713e-05,
324
+ "loss": 13.6677,
325
+ "step": 2475
326
+ },
327
+ {
328
+ "epoch": 32.857142857142854,
329
+ "grad_norm": 92.99346160888672,
330
+ "learning_rate": 1.1736609336609336e-05,
331
+ "loss": 13.1851,
332
+ "step": 2530
333
+ },
334
+ {
335
+ "epoch": 33.57142857142857,
336
+ "grad_norm": 97.57978820800781,
337
+ "learning_rate": 1.1358230958230958e-05,
338
+ "loss": 13.5215,
339
+ "step": 2585
340
+ },
341
+ {
342
+ "epoch": 34.285714285714285,
343
+ "grad_norm": 63.118141174316406,
344
+ "learning_rate": 1.0979852579852579e-05,
345
+ "loss": 12.8525,
346
+ "step": 2640
347
+ },
348
+ {
349
+ "epoch": 35.0,
350
+ "grad_norm": 63.68766403198242,
351
+ "learning_rate": 1.06014742014742e-05,
352
+ "loss": 12.441,
353
+ "step": 2695
354
+ },
355
+ {
356
+ "epoch": 35.714285714285715,
357
+ "grad_norm": 70.33533477783203,
358
+ "learning_rate": 1.0223095823095823e-05,
359
+ "loss": 12.2331,
360
+ "step": 2750
361
+ },
362
+ {
363
+ "epoch": 36.42857142857143,
364
+ "grad_norm": 64.16197967529297,
365
+ "learning_rate": 9.844717444717445e-06,
366
+ "loss": 11.7547,
367
+ "step": 2805
368
+ },
369
+ {
370
+ "epoch": 37.142857142857146,
371
+ "grad_norm": 85.83612823486328,
372
+ "learning_rate": 9.466339066339066e-06,
373
+ "loss": 11.8633,
374
+ "step": 2860
375
+ },
376
+ {
377
+ "epoch": 37.857142857142854,
378
+ "grad_norm": 89.40387725830078,
379
+ "learning_rate": 9.087960687960689e-06,
380
+ "loss": 11.5795,
381
+ "step": 2915
382
+ },
383
+ {
384
+ "epoch": 38.57142857142857,
385
+ "grad_norm": 76.38651275634766,
386
+ "learning_rate": 8.709582309582309e-06,
387
+ "loss": 11.334,
388
+ "step": 2970
389
+ },
390
+ {
391
+ "epoch": 39.285714285714285,
392
+ "grad_norm": 110.11483764648438,
393
+ "learning_rate": 8.331203931203932e-06,
394
+ "loss": 10.805,
395
+ "step": 3025
396
+ },
397
+ {
398
+ "epoch": 40.0,
399
+ "grad_norm": 60.31315612792969,
400
+ "learning_rate": 7.952825552825553e-06,
401
+ "loss": 10.9056,
402
+ "step": 3080
403
+ },
404
+ {
405
+ "epoch": 40.714285714285715,
406
+ "grad_norm": 107.96589660644531,
407
+ "learning_rate": 7.574447174447175e-06,
408
+ "loss": 10.5828,
409
+ "step": 3135
410
+ },
411
+ {
412
+ "epoch": 41.42857142857143,
413
+ "grad_norm": 74.8004379272461,
414
+ "learning_rate": 7.1960687960687955e-06,
415
+ "loss": 10.2818,
416
+ "step": 3190
417
+ },
418
+ {
419
+ "epoch": 42.142857142857146,
420
+ "grad_norm": 54.093475341796875,
421
+ "learning_rate": 6.817690417690418e-06,
422
+ "loss": 10.0785,
423
+ "step": 3245
424
+ },
425
+ {
426
+ "epoch": 42.857142857142854,
427
+ "grad_norm": 115.28564453125,
428
+ "learning_rate": 6.439312039312039e-06,
429
+ "loss": 9.8794,
430
+ "step": 3300
431
+ },
432
+ {
433
+ "epoch": 43.57142857142857,
434
+ "grad_norm": 52.57551956176758,
435
+ "learning_rate": 6.06093366093366e-06,
436
+ "loss": 9.7072,
437
+ "step": 3355
438
+ },
439
+ {
440
+ "epoch": 44.285714285714285,
441
+ "grad_norm": 51.994075775146484,
442
+ "learning_rate": 5.682555282555282e-06,
443
+ "loss": 9.6015,
444
+ "step": 3410
445
+ },
446
+ {
447
+ "epoch": 45.0,
448
+ "grad_norm": 73.08104705810547,
449
+ "learning_rate": 5.304176904176904e-06,
450
+ "loss": 9.305,
451
+ "step": 3465
452
+ },
453
+ {
454
+ "epoch": 45.714285714285715,
455
+ "grad_norm": 75.5807876586914,
456
+ "learning_rate": 4.925798525798525e-06,
457
+ "loss": 9.2281,
458
+ "step": 3520
459
+ },
460
+ {
461
+ "epoch": 46.42857142857143,
462
+ "grad_norm": 53.319637298583984,
463
+ "learning_rate": 4.547420147420147e-06,
464
+ "loss": 8.996,
465
+ "step": 3575
466
+ },
467
+ {
468
+ "epoch": 47.142857142857146,
469
+ "grad_norm": 72.28289794921875,
470
+ "learning_rate": 4.1690417690417685e-06,
471
+ "loss": 9.0072,
472
+ "step": 3630
473
+ },
474
+ {
475
+ "epoch": 47.857142857142854,
476
+ "grad_norm": 66.79710388183594,
477
+ "learning_rate": 3.7906633906633902e-06,
478
+ "loss": 8.7311,
479
+ "step": 3685
480
+ },
481
+ {
482
+ "epoch": 48.57142857142857,
483
+ "grad_norm": 83.55674743652344,
484
+ "learning_rate": 3.4122850122850124e-06,
485
+ "loss": 8.5959,
486
+ "step": 3740
487
+ }
488
+ ],
489
+ "logging_steps": 55,
490
+ "max_steps": 4235,
491
+ "num_input_tokens_seen": 0,
492
+ "num_train_epochs": 55,
493
+ "save_steps": 220,
494
+ "stateful_callbacks": {
495
+ "TrainerControl": {
496
+ "args": {
497
+ "should_epoch_stop": false,
498
+ "should_evaluate": false,
499
+ "should_log": false,
500
+ "should_save": true,
501
+ "should_training_stop": false
502
+ },
503
+ "attributes": {}
504
+ }
505
+ },
506
+ "total_flos": 5.8876427477621146e+19,
507
+ "train_batch_size": 28,
508
+ "trial_name": null,
509
+ "trial_params": null
510
+ }
checkpoint-3740/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99e55f6c266fb80fff7a194cd0925963d93d119d60f4443cb609bfd824b16fd5
3
+ size 5841
checkpoint-3960/config.json ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "silu",
4
+ "anchor_image_size": null,
5
+ "architectures": [
6
+ "RTDetrV2ForObjectDetection"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auxiliary_loss": true,
10
+ "backbone": null,
11
+ "backbone_config": {
12
+ "depths": [
13
+ 3,
14
+ 4,
15
+ 23,
16
+ 3
17
+ ],
18
+ "downsample_in_bottleneck": false,
19
+ "downsample_in_first_stage": false,
20
+ "embedding_size": 64,
21
+ "hidden_act": "relu",
22
+ "hidden_sizes": [
23
+ 256,
24
+ 512,
25
+ 1024,
26
+ 2048
27
+ ],
28
+ "layer_type": "bottleneck",
29
+ "model_type": "rt_detr_resnet",
30
+ "num_channels": 3,
31
+ "out_features": [
32
+ "stage2",
33
+ "stage3",
34
+ "stage4"
35
+ ],
36
+ "out_indices": [
37
+ 2,
38
+ 3,
39
+ 4
40
+ ],
41
+ "stage_names": [
42
+ "stem",
43
+ "stage1",
44
+ "stage2",
45
+ "stage3",
46
+ "stage4"
47
+ ],
48
+ "torch_dtype": "float32"
49
+ },
50
+ "backbone_kwargs": null,
51
+ "batch_norm_eps": 1e-05,
52
+ "box_noise_scale": 1.0,
53
+ "d_model": 256,
54
+ "decoder_activation_function": "relu",
55
+ "decoder_attention_heads": 8,
56
+ "decoder_ffn_dim": 1024,
57
+ "decoder_in_channels": [
58
+ 384,
59
+ 384,
60
+ 384
61
+ ],
62
+ "decoder_layers": 6,
63
+ "decoder_method": "default",
64
+ "decoder_n_levels": 3,
65
+ "decoder_n_points": 4,
66
+ "decoder_offset_scale": 0.5,
67
+ "disable_custom_kernels": true,
68
+ "dropout": 0.0,
69
+ "encode_proj_layers": [
70
+ 2
71
+ ],
72
+ "encoder_activation_function": "gelu",
73
+ "encoder_attention_heads": 8,
74
+ "encoder_ffn_dim": 2048,
75
+ "encoder_hidden_dim": 384,
76
+ "encoder_in_channels": [
77
+ 512,
78
+ 1024,
79
+ 2048
80
+ ],
81
+ "encoder_layers": 1,
82
+ "eos_coefficient": 0.0001,
83
+ "eval_size": null,
84
+ "feat_strides": [
85
+ 8,
86
+ 16,
87
+ 32
88
+ ],
89
+ "focal_loss_alpha": 0.75,
90
+ "focal_loss_gamma": 2.0,
91
+ "freeze_backbone_batch_norms": true,
92
+ "hidden_expansion": 1.0,
93
+ "id2label": {
94
+ "0": "LABEL_0",
95
+ "1": "LABEL_1",
96
+ "2": "LABEL_2"
97
+ },
98
+ "initializer_bias_prior_prob": null,
99
+ "initializer_range": 0.01,
100
+ "is_encoder_decoder": true,
101
+ "label2id": {
102
+ "LABEL_0": 0,
103
+ "LABEL_1": 1,
104
+ "LABEL_2": 2
105
+ },
106
+ "label_noise_ratio": 0.5,
107
+ "layer_norm_eps": 1e-05,
108
+ "learn_initial_query": false,
109
+ "matcher_alpha": 0.25,
110
+ "matcher_bbox_cost": 5.0,
111
+ "matcher_class_cost": 2.0,
112
+ "matcher_gamma": 2.0,
113
+ "matcher_giou_cost": 2.0,
114
+ "model_type": "rt_detr_v2",
115
+ "normalize_before": false,
116
+ "num_denoising": 100,
117
+ "num_feature_levels": 3,
118
+ "num_queries": 300,
119
+ "positional_encoding_temperature": 10000,
120
+ "torch_dtype": "float32",
121
+ "transformers_version": "4.55.2",
122
+ "use_focal_loss": true,
123
+ "use_pretrained_backbone": false,
124
+ "use_timm_backbone": false,
125
+ "weight_loss_bbox": 5.0,
126
+ "weight_loss_giou": 2.0,
127
+ "weight_loss_vfl": 1.0,
128
+ "with_box_refine": true
129
+ }
checkpoint-3960/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0da8aa9a9dbf156ff61271506250c72336d8a6bd77ebdb2f058032615c467db
3
+ size 306699044
checkpoint-3960/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:627c9c79f2991c364b1cd821ca8ca24dca3b10481c0b54ce39491f788f5e4659
3
+ size 611580898
checkpoint-3960/preprocessor_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_annotations": true,
3
+ "do_normalize": false,
4
+ "do_pad": false,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "format": "coco_detection",
8
+ "image_mean": [
9
+ 0.485,
10
+ 0.456,
11
+ 0.406
12
+ ],
13
+ "image_processor_type": "RTDetrImageProcessor",
14
+ "image_std": [
15
+ 0.229,
16
+ 0.224,
17
+ 0.225
18
+ ],
19
+ "pad_size": null,
20
+ "resample": 2,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "height": 640,
24
+ "width": 640
25
+ }
26
+ }
checkpoint-3960/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff0ae4594f12a938934516f599b96c8fa1d052ac793f26e44ed839202dd42016
3
+ size 14645
checkpoint-3960/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d866a08f2c164179018beb811aecb078e91850415968e1005b6ef3bd8128de66
3
+ size 1465
checkpoint-3960/trainer_state.json ADDED
@@ -0,0 +1,538 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 51.42857142857143,
6
+ "eval_steps": 500,
7
+ "global_step": 3960,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.7142857142857143,
14
+ "grad_norm": 4405.81884765625,
15
+ "learning_rate": 3.436363636363636e-07,
16
+ "loss": 1357.5645,
17
+ "step": 55
18
+ },
19
+ {
20
+ "epoch": 1.4285714285714286,
21
+ "grad_norm": 5894.37841796875,
22
+ "learning_rate": 6.936363636363637e-07,
23
+ "loss": 1290.9688,
24
+ "step": 110
25
+ },
26
+ {
27
+ "epoch": 2.142857142857143,
28
+ "grad_norm": 3667.852294921875,
29
+ "learning_rate": 1.0436363636363635e-06,
30
+ "loss": 1140.9515,
31
+ "step": 165
32
+ },
33
+ {
34
+ "epoch": 2.857142857142857,
35
+ "grad_norm": 3578.294677734375,
36
+ "learning_rate": 1.3936363636363637e-06,
37
+ "loss": 942.618,
38
+ "step": 220
39
+ },
40
+ {
41
+ "epoch": 3.571428571428571,
42
+ "grad_norm": 2097.637939453125,
43
+ "learning_rate": 1.7436363636363636e-06,
44
+ "loss": 721.2856,
45
+ "step": 275
46
+ },
47
+ {
48
+ "epoch": 4.285714285714286,
49
+ "grad_norm": 1840.612548828125,
50
+ "learning_rate": 2.0936363636363636e-06,
51
+ "loss": 525.8542,
52
+ "step": 330
53
+ },
54
+ {
55
+ "epoch": 5.0,
56
+ "grad_norm": 1230.875,
57
+ "learning_rate": 2.4436363636363636e-06,
58
+ "loss": 397.7722,
59
+ "step": 385
60
+ },
61
+ {
62
+ "epoch": 5.714285714285714,
63
+ "grad_norm": 1280.699951171875,
64
+ "learning_rate": 2.7936363636363637e-06,
65
+ "loss": 313.8322,
66
+ "step": 440
67
+ },
68
+ {
69
+ "epoch": 6.428571428571429,
70
+ "grad_norm": 4265.28759765625,
71
+ "learning_rate": 3.1436363636363637e-06,
72
+ "loss": 254.1449,
73
+ "step": 495
74
+ },
75
+ {
76
+ "epoch": 7.142857142857143,
77
+ "grad_norm": 967.1757202148438,
78
+ "learning_rate": 3.4936363636363633e-06,
79
+ "loss": 218.4529,
80
+ "step": 550
81
+ },
82
+ {
83
+ "epoch": 7.857142857142857,
84
+ "grad_norm": 765.2302856445312,
85
+ "learning_rate": 3.843636363636364e-06,
86
+ "loss": 197.915,
87
+ "step": 605
88
+ },
89
+ {
90
+ "epoch": 8.571428571428571,
91
+ "grad_norm": 700.331787109375,
92
+ "learning_rate": 4.193636363636364e-06,
93
+ "loss": 170.9019,
94
+ "step": 660
95
+ },
96
+ {
97
+ "epoch": 9.285714285714286,
98
+ "grad_norm": 613.8758544921875,
99
+ "learning_rate": 4.543636363636363e-06,
100
+ "loss": 154.6195,
101
+ "step": 715
102
+ },
103
+ {
104
+ "epoch": 10.0,
105
+ "grad_norm": 3405.25732421875,
106
+ "learning_rate": 4.893636363636364e-06,
107
+ "loss": 136.8394,
108
+ "step": 770
109
+ },
110
+ {
111
+ "epoch": 10.714285714285714,
112
+ "grad_norm": 1219.8994140625,
113
+ "learning_rate": 5.243636363636364e-06,
114
+ "loss": 116.0246,
115
+ "step": 825
116
+ },
117
+ {
118
+ "epoch": 11.428571428571429,
119
+ "grad_norm": 306.2723083496094,
120
+ "learning_rate": 5.593636363636363e-06,
121
+ "loss": 104.2055,
122
+ "step": 880
123
+ },
124
+ {
125
+ "epoch": 12.142857142857142,
126
+ "grad_norm": 288.2547302246094,
127
+ "learning_rate": 5.943636363636364e-06,
128
+ "loss": 88.9812,
129
+ "step": 935
130
+ },
131
+ {
132
+ "epoch": 12.857142857142858,
133
+ "grad_norm": 1016.5924682617188,
134
+ "learning_rate": 6.293636363636363e-06,
135
+ "loss": 79.9017,
136
+ "step": 990
137
+ },
138
+ {
139
+ "epoch": 13.571428571428571,
140
+ "grad_norm": 239.49191284179688,
141
+ "learning_rate": 6.643636363636363e-06,
142
+ "loss": 73.6932,
143
+ "step": 1045
144
+ },
145
+ {
146
+ "epoch": 14.285714285714286,
147
+ "grad_norm": 193.92002868652344,
148
+ "learning_rate": 6.993636363636364e-06,
149
+ "loss": 62.2018,
150
+ "step": 1100
151
+ },
152
+ {
153
+ "epoch": 15.0,
154
+ "grad_norm": 179.0568084716797,
155
+ "learning_rate": 7.343636363636363e-06,
156
+ "loss": 54.0213,
157
+ "step": 1155
158
+ },
159
+ {
160
+ "epoch": 15.714285714285714,
161
+ "grad_norm": 166.02944946289062,
162
+ "learning_rate": 7.693636363636364e-06,
163
+ "loss": 48.1955,
164
+ "step": 1210
165
+ },
166
+ {
167
+ "epoch": 16.428571428571427,
168
+ "grad_norm": 126.53916931152344,
169
+ "learning_rate": 8.043636363636364e-06,
170
+ "loss": 43.5431,
171
+ "step": 1265
172
+ },
173
+ {
174
+ "epoch": 17.142857142857142,
175
+ "grad_norm": 145.61166381835938,
176
+ "learning_rate": 8.393636363636363e-06,
177
+ "loss": 41.4399,
178
+ "step": 1320
179
+ },
180
+ {
181
+ "epoch": 17.857142857142858,
182
+ "grad_norm": 122.2297134399414,
183
+ "learning_rate": 8.743636363636363e-06,
184
+ "loss": 35.2278,
185
+ "step": 1375
186
+ },
187
+ {
188
+ "epoch": 18.571428571428573,
189
+ "grad_norm": 117.88919067382812,
190
+ "learning_rate": 9.093636363636363e-06,
191
+ "loss": 31.2827,
192
+ "step": 1430
193
+ },
194
+ {
195
+ "epoch": 19.285714285714285,
196
+ "grad_norm": 88.52986907958984,
197
+ "learning_rate": 9.443636363636364e-06,
198
+ "loss": 28.8076,
199
+ "step": 1485
200
+ },
201
+ {
202
+ "epoch": 20.0,
203
+ "grad_norm": 88.49090576171875,
204
+ "learning_rate": 9.793636363636364e-06,
205
+ "loss": 28.232,
206
+ "step": 1540
207
+ },
208
+ {
209
+ "epoch": 20.714285714285715,
210
+ "grad_norm": 114.45001983642578,
211
+ "learning_rate": 1.0143636363636363e-05,
212
+ "loss": 26.0885,
213
+ "step": 1595
214
+ },
215
+ {
216
+ "epoch": 21.428571428571427,
217
+ "grad_norm": 74.16987609863281,
218
+ "learning_rate": 1.0493636363636363e-05,
219
+ "loss": 24.2689,
220
+ "step": 1650
221
+ },
222
+ {
223
+ "epoch": 22.142857142857142,
224
+ "grad_norm": 82.25133514404297,
225
+ "learning_rate": 1.0843636363636363e-05,
226
+ "loss": 22.6963,
227
+ "step": 1705
228
+ },
229
+ {
230
+ "epoch": 22.857142857142858,
231
+ "grad_norm": 74.76679229736328,
232
+ "learning_rate": 1.1193636363636363e-05,
233
+ "loss": 21.3257,
234
+ "step": 1760
235
+ },
236
+ {
237
+ "epoch": 23.571428571428573,
238
+ "grad_norm": 68.91163635253906,
239
+ "learning_rate": 1.1543636363636365e-05,
240
+ "loss": 20.2199,
241
+ "step": 1815
242
+ },
243
+ {
244
+ "epoch": 24.285714285714285,
245
+ "grad_norm": 59.86214828491211,
246
+ "learning_rate": 1.1893636363636363e-05,
247
+ "loss": 19.3769,
248
+ "step": 1870
249
+ },
250
+ {
251
+ "epoch": 25.0,
252
+ "grad_norm": 76.9339370727539,
253
+ "learning_rate": 1.2243636363636363e-05,
254
+ "loss": 18.0471,
255
+ "step": 1925
256
+ },
257
+ {
258
+ "epoch": 25.714285714285715,
259
+ "grad_norm": 66.16152954101562,
260
+ "learning_rate": 1.2593636363636363e-05,
261
+ "loss": 17.496,
262
+ "step": 1980
263
+ },
264
+ {
265
+ "epoch": 26.428571428571427,
266
+ "grad_norm": 58.77778244018555,
267
+ "learning_rate": 1.2943636363636363e-05,
268
+ "loss": 16.9468,
269
+ "step": 2035
270
+ },
271
+ {
272
+ "epoch": 27.142857142857142,
273
+ "grad_norm": 71.95783233642578,
274
+ "learning_rate": 1.3293636363636363e-05,
275
+ "loss": 16.3026,
276
+ "step": 2090
277
+ },
278
+ {
279
+ "epoch": 27.857142857142858,
280
+ "grad_norm": 131.7191619873047,
281
+ "learning_rate": 1.3643636363636363e-05,
282
+ "loss": 16.008,
283
+ "step": 2145
284
+ },
285
+ {
286
+ "epoch": 28.571428571428573,
287
+ "grad_norm": 81.78388977050781,
288
+ "learning_rate": 1.3993636363636363e-05,
289
+ "loss": 15.4616,
290
+ "step": 2200
291
+ },
292
+ {
293
+ "epoch": 29.285714285714285,
294
+ "grad_norm": 60.10112380981445,
295
+ "learning_rate": 1.3628501228501228e-05,
296
+ "loss": 15.195,
297
+ "step": 2255
298
+ },
299
+ {
300
+ "epoch": 30.0,
301
+ "grad_norm": 77.59001922607422,
302
+ "learning_rate": 1.325012285012285e-05,
303
+ "loss": 14.7557,
304
+ "step": 2310
305
+ },
306
+ {
307
+ "epoch": 30.714285714285715,
308
+ "grad_norm": 77.3856201171875,
309
+ "learning_rate": 1.287174447174447e-05,
310
+ "loss": 14.2851,
311
+ "step": 2365
312
+ },
313
+ {
314
+ "epoch": 31.428571428571427,
315
+ "grad_norm": 73.75492095947266,
316
+ "learning_rate": 1.2493366093366094e-05,
317
+ "loss": 13.867,
318
+ "step": 2420
319
+ },
320
+ {
321
+ "epoch": 32.142857142857146,
322
+ "grad_norm": 78.76831817626953,
323
+ "learning_rate": 1.2114987714987713e-05,
324
+ "loss": 13.6677,
325
+ "step": 2475
326
+ },
327
+ {
328
+ "epoch": 32.857142857142854,
329
+ "grad_norm": 92.99346160888672,
330
+ "learning_rate": 1.1736609336609336e-05,
331
+ "loss": 13.1851,
332
+ "step": 2530
333
+ },
334
+ {
335
+ "epoch": 33.57142857142857,
336
+ "grad_norm": 97.57978820800781,
337
+ "learning_rate": 1.1358230958230958e-05,
338
+ "loss": 13.5215,
339
+ "step": 2585
340
+ },
341
+ {
342
+ "epoch": 34.285714285714285,
343
+ "grad_norm": 63.118141174316406,
344
+ "learning_rate": 1.0979852579852579e-05,
345
+ "loss": 12.8525,
346
+ "step": 2640
347
+ },
348
+ {
349
+ "epoch": 35.0,
350
+ "grad_norm": 63.68766403198242,
351
+ "learning_rate": 1.06014742014742e-05,
352
+ "loss": 12.441,
353
+ "step": 2695
354
+ },
355
+ {
356
+ "epoch": 35.714285714285715,
357
+ "grad_norm": 70.33533477783203,
358
+ "learning_rate": 1.0223095823095823e-05,
359
+ "loss": 12.2331,
360
+ "step": 2750
361
+ },
362
+ {
363
+ "epoch": 36.42857142857143,
364
+ "grad_norm": 64.16197967529297,
365
+ "learning_rate": 9.844717444717445e-06,
366
+ "loss": 11.7547,
367
+ "step": 2805
368
+ },
369
+ {
370
+ "epoch": 37.142857142857146,
371
+ "grad_norm": 85.83612823486328,
372
+ "learning_rate": 9.466339066339066e-06,
373
+ "loss": 11.8633,
374
+ "step": 2860
375
+ },
376
+ {
377
+ "epoch": 37.857142857142854,
378
+ "grad_norm": 89.40387725830078,
379
+ "learning_rate": 9.087960687960689e-06,
380
+ "loss": 11.5795,
381
+ "step": 2915
382
+ },
383
+ {
384
+ "epoch": 38.57142857142857,
385
+ "grad_norm": 76.38651275634766,
386
+ "learning_rate": 8.709582309582309e-06,
387
+ "loss": 11.334,
388
+ "step": 2970
389
+ },
390
+ {
391
+ "epoch": 39.285714285714285,
392
+ "grad_norm": 110.11483764648438,
393
+ "learning_rate": 8.331203931203932e-06,
394
+ "loss": 10.805,
395
+ "step": 3025
396
+ },
397
+ {
398
+ "epoch": 40.0,
399
+ "grad_norm": 60.31315612792969,
400
+ "learning_rate": 7.952825552825553e-06,
401
+ "loss": 10.9056,
402
+ "step": 3080
403
+ },
404
+ {
405
+ "epoch": 40.714285714285715,
406
+ "grad_norm": 107.96589660644531,
407
+ "learning_rate": 7.574447174447175e-06,
408
+ "loss": 10.5828,
409
+ "step": 3135
410
+ },
411
+ {
412
+ "epoch": 41.42857142857143,
413
+ "grad_norm": 74.8004379272461,
414
+ "learning_rate": 7.1960687960687955e-06,
415
+ "loss": 10.2818,
416
+ "step": 3190
417
+ },
418
+ {
419
+ "epoch": 42.142857142857146,
420
+ "grad_norm": 54.093475341796875,
421
+ "learning_rate": 6.817690417690418e-06,
422
+ "loss": 10.0785,
423
+ "step": 3245
424
+ },
425
+ {
426
+ "epoch": 42.857142857142854,
427
+ "grad_norm": 115.28564453125,
428
+ "learning_rate": 6.439312039312039e-06,
429
+ "loss": 9.8794,
430
+ "step": 3300
431
+ },
432
+ {
433
+ "epoch": 43.57142857142857,
434
+ "grad_norm": 52.57551956176758,
435
+ "learning_rate": 6.06093366093366e-06,
436
+ "loss": 9.7072,
437
+ "step": 3355
438
+ },
439
+ {
440
+ "epoch": 44.285714285714285,
441
+ "grad_norm": 51.994075775146484,
442
+ "learning_rate": 5.682555282555282e-06,
443
+ "loss": 9.6015,
444
+ "step": 3410
445
+ },
446
+ {
447
+ "epoch": 45.0,
448
+ "grad_norm": 73.08104705810547,
449
+ "learning_rate": 5.304176904176904e-06,
450
+ "loss": 9.305,
451
+ "step": 3465
452
+ },
453
+ {
454
+ "epoch": 45.714285714285715,
455
+ "grad_norm": 75.5807876586914,
456
+ "learning_rate": 4.925798525798525e-06,
457
+ "loss": 9.2281,
458
+ "step": 3520
459
+ },
460
+ {
461
+ "epoch": 46.42857142857143,
462
+ "grad_norm": 53.319637298583984,
463
+ "learning_rate": 4.547420147420147e-06,
464
+ "loss": 8.996,
465
+ "step": 3575
466
+ },
467
+ {
468
+ "epoch": 47.142857142857146,
469
+ "grad_norm": 72.28289794921875,
470
+ "learning_rate": 4.1690417690417685e-06,
471
+ "loss": 9.0072,
472
+ "step": 3630
473
+ },
474
+ {
475
+ "epoch": 47.857142857142854,
476
+ "grad_norm": 66.79710388183594,
477
+ "learning_rate": 3.7906633906633902e-06,
478
+ "loss": 8.7311,
479
+ "step": 3685
480
+ },
481
+ {
482
+ "epoch": 48.57142857142857,
483
+ "grad_norm": 83.55674743652344,
484
+ "learning_rate": 3.4122850122850124e-06,
485
+ "loss": 8.5959,
486
+ "step": 3740
487
+ },
488
+ {
489
+ "epoch": 49.285714285714285,
490
+ "grad_norm": 78.8072738647461,
491
+ "learning_rate": 3.0339066339066337e-06,
492
+ "loss": 8.5462,
493
+ "step": 3795
494
+ },
495
+ {
496
+ "epoch": 50.0,
497
+ "grad_norm": 102.47228240966797,
498
+ "learning_rate": 2.6555282555282554e-06,
499
+ "loss": 8.2928,
500
+ "step": 3850
501
+ },
502
+ {
503
+ "epoch": 50.714285714285715,
504
+ "grad_norm": 93.27884674072266,
505
+ "learning_rate": 2.277149877149877e-06,
506
+ "loss": 8.1427,
507
+ "step": 3905
508
+ },
509
+ {
510
+ "epoch": 51.42857142857143,
511
+ "grad_norm": 137.4209747314453,
512
+ "learning_rate": 1.8987714987714987e-06,
513
+ "loss": 8.164,
514
+ "step": 3960
515
+ }
516
+ ],
517
+ "logging_steps": 55,
518
+ "max_steps": 4235,
519
+ "num_input_tokens_seen": 0,
520
+ "num_train_epochs": 55,
521
+ "save_steps": 220,
522
+ "stateful_callbacks": {
523
+ "TrainerControl": {
524
+ "args": {
525
+ "should_epoch_stop": false,
526
+ "should_evaluate": false,
527
+ "should_log": false,
528
+ "should_save": true,
529
+ "should_training_stop": false
530
+ },
531
+ "attributes": {}
532
+ }
533
+ },
534
+ "total_flos": 6.2339349145809715e+19,
535
+ "train_batch_size": 28,
536
+ "trial_name": null,
537
+ "trial_params": null
538
+ }
checkpoint-3960/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99e55f6c266fb80fff7a194cd0925963d93d119d60f4443cb609bfd824b16fd5
3
+ size 5841
checkpoint-4180/config.json ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "silu",
4
+ "anchor_image_size": null,
5
+ "architectures": [
6
+ "RTDetrV2ForObjectDetection"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auxiliary_loss": true,
10
+ "backbone": null,
11
+ "backbone_config": {
12
+ "depths": [
13
+ 3,
14
+ 4,
15
+ 23,
16
+ 3
17
+ ],
18
+ "downsample_in_bottleneck": false,
19
+ "downsample_in_first_stage": false,
20
+ "embedding_size": 64,
21
+ "hidden_act": "relu",
22
+ "hidden_sizes": [
23
+ 256,
24
+ 512,
25
+ 1024,
26
+ 2048
27
+ ],
28
+ "layer_type": "bottleneck",
29
+ "model_type": "rt_detr_resnet",
30
+ "num_channels": 3,
31
+ "out_features": [
32
+ "stage2",
33
+ "stage3",
34
+ "stage4"
35
+ ],
36
+ "out_indices": [
37
+ 2,
38
+ 3,
39
+ 4
40
+ ],
41
+ "stage_names": [
42
+ "stem",
43
+ "stage1",
44
+ "stage2",
45
+ "stage3",
46
+ "stage4"
47
+ ],
48
+ "torch_dtype": "float32"
49
+ },
50
+ "backbone_kwargs": null,
51
+ "batch_norm_eps": 1e-05,
52
+ "box_noise_scale": 1.0,
53
+ "d_model": 256,
54
+ "decoder_activation_function": "relu",
55
+ "decoder_attention_heads": 8,
56
+ "decoder_ffn_dim": 1024,
57
+ "decoder_in_channels": [
58
+ 384,
59
+ 384,
60
+ 384
61
+ ],
62
+ "decoder_layers": 6,
63
+ "decoder_method": "default",
64
+ "decoder_n_levels": 3,
65
+ "decoder_n_points": 4,
66
+ "decoder_offset_scale": 0.5,
67
+ "disable_custom_kernels": true,
68
+ "dropout": 0.0,
69
+ "encode_proj_layers": [
70
+ 2
71
+ ],
72
+ "encoder_activation_function": "gelu",
73
+ "encoder_attention_heads": 8,
74
+ "encoder_ffn_dim": 2048,
75
+ "encoder_hidden_dim": 384,
76
+ "encoder_in_channels": [
77
+ 512,
78
+ 1024,
79
+ 2048
80
+ ],
81
+ "encoder_layers": 1,
82
+ "eos_coefficient": 0.0001,
83
+ "eval_size": null,
84
+ "feat_strides": [
85
+ 8,
86
+ 16,
87
+ 32
88
+ ],
89
+ "focal_loss_alpha": 0.75,
90
+ "focal_loss_gamma": 2.0,
91
+ "freeze_backbone_batch_norms": true,
92
+ "hidden_expansion": 1.0,
93
+ "id2label": {
94
+ "0": "LABEL_0",
95
+ "1": "LABEL_1",
96
+ "2": "LABEL_2"
97
+ },
98
+ "initializer_bias_prior_prob": null,
99
+ "initializer_range": 0.01,
100
+ "is_encoder_decoder": true,
101
+ "label2id": {
102
+ "LABEL_0": 0,
103
+ "LABEL_1": 1,
104
+ "LABEL_2": 2
105
+ },
106
+ "label_noise_ratio": 0.5,
107
+ "layer_norm_eps": 1e-05,
108
+ "learn_initial_query": false,
109
+ "matcher_alpha": 0.25,
110
+ "matcher_bbox_cost": 5.0,
111
+ "matcher_class_cost": 2.0,
112
+ "matcher_gamma": 2.0,
113
+ "matcher_giou_cost": 2.0,
114
+ "model_type": "rt_detr_v2",
115
+ "normalize_before": false,
116
+ "num_denoising": 100,
117
+ "num_feature_levels": 3,
118
+ "num_queries": 300,
119
+ "positional_encoding_temperature": 10000,
120
+ "torch_dtype": "float32",
121
+ "transformers_version": "4.55.2",
122
+ "use_focal_loss": true,
123
+ "use_pretrained_backbone": false,
124
+ "use_timm_backbone": false,
125
+ "weight_loss_bbox": 5.0,
126
+ "weight_loss_giou": 2.0,
127
+ "weight_loss_vfl": 1.0,
128
+ "with_box_refine": true
129
+ }
checkpoint-4180/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a3708ff7eb478f4d78afc03d6682682b88855ad74bd90d61a1f44bd6cbd9031
3
+ size 306699044
checkpoint-4180/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c37e9d87c40ec6bec2836994bbdc85c72273474444714cbc9174eedffff8cbce
3
+ size 611580898
checkpoint-4180/preprocessor_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_annotations": true,
3
+ "do_normalize": false,
4
+ "do_pad": false,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "format": "coco_detection",
8
+ "image_mean": [
9
+ 0.485,
10
+ 0.456,
11
+ 0.406
12
+ ],
13
+ "image_processor_type": "RTDetrImageProcessor",
14
+ "image_std": [
15
+ 0.229,
16
+ 0.224,
17
+ 0.225
18
+ ],
19
+ "pad_size": null,
20
+ "resample": 2,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "height": 640,
24
+ "width": 640
25
+ }
26
+ }
checkpoint-4180/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa8987e65fa1b372b40e3601811f025fddf3af32827da76164d695e4a6697a78
3
+ size 14645
checkpoint-4180/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8a0d01e3953f2300825d40619ebf7431436db245ce8c643859c6c66bd4e6043
3
+ size 1465
checkpoint-4180/trainer_state.json ADDED
@@ -0,0 +1,566 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 54.285714285714285,
6
+ "eval_steps": 500,
7
+ "global_step": 4180,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.7142857142857143,
14
+ "grad_norm": 4405.81884765625,
15
+ "learning_rate": 3.436363636363636e-07,
16
+ "loss": 1357.5645,
17
+ "step": 55
18
+ },
19
+ {
20
+ "epoch": 1.4285714285714286,
21
+ "grad_norm": 5894.37841796875,
22
+ "learning_rate": 6.936363636363637e-07,
23
+ "loss": 1290.9688,
24
+ "step": 110
25
+ },
26
+ {
27
+ "epoch": 2.142857142857143,
28
+ "grad_norm": 3667.852294921875,
29
+ "learning_rate": 1.0436363636363635e-06,
30
+ "loss": 1140.9515,
31
+ "step": 165
32
+ },
33
+ {
34
+ "epoch": 2.857142857142857,
35
+ "grad_norm": 3578.294677734375,
36
+ "learning_rate": 1.3936363636363637e-06,
37
+ "loss": 942.618,
38
+ "step": 220
39
+ },
40
+ {
41
+ "epoch": 3.571428571428571,
42
+ "grad_norm": 2097.637939453125,
43
+ "learning_rate": 1.7436363636363636e-06,
44
+ "loss": 721.2856,
45
+ "step": 275
46
+ },
47
+ {
48
+ "epoch": 4.285714285714286,
49
+ "grad_norm": 1840.612548828125,
50
+ "learning_rate": 2.0936363636363636e-06,
51
+ "loss": 525.8542,
52
+ "step": 330
53
+ },
54
+ {
55
+ "epoch": 5.0,
56
+ "grad_norm": 1230.875,
57
+ "learning_rate": 2.4436363636363636e-06,
58
+ "loss": 397.7722,
59
+ "step": 385
60
+ },
61
+ {
62
+ "epoch": 5.714285714285714,
63
+ "grad_norm": 1280.699951171875,
64
+ "learning_rate": 2.7936363636363637e-06,
65
+ "loss": 313.8322,
66
+ "step": 440
67
+ },
68
+ {
69
+ "epoch": 6.428571428571429,
70
+ "grad_norm": 4265.28759765625,
71
+ "learning_rate": 3.1436363636363637e-06,
72
+ "loss": 254.1449,
73
+ "step": 495
74
+ },
75
+ {
76
+ "epoch": 7.142857142857143,
77
+ "grad_norm": 967.1757202148438,
78
+ "learning_rate": 3.4936363636363633e-06,
79
+ "loss": 218.4529,
80
+ "step": 550
81
+ },
82
+ {
83
+ "epoch": 7.857142857142857,
84
+ "grad_norm": 765.2302856445312,
85
+ "learning_rate": 3.843636363636364e-06,
86
+ "loss": 197.915,
87
+ "step": 605
88
+ },
89
+ {
90
+ "epoch": 8.571428571428571,
91
+ "grad_norm": 700.331787109375,
92
+ "learning_rate": 4.193636363636364e-06,
93
+ "loss": 170.9019,
94
+ "step": 660
95
+ },
96
+ {
97
+ "epoch": 9.285714285714286,
98
+ "grad_norm": 613.8758544921875,
99
+ "learning_rate": 4.543636363636363e-06,
100
+ "loss": 154.6195,
101
+ "step": 715
102
+ },
103
+ {
104
+ "epoch": 10.0,
105
+ "grad_norm": 3405.25732421875,
106
+ "learning_rate": 4.893636363636364e-06,
107
+ "loss": 136.8394,
108
+ "step": 770
109
+ },
110
+ {
111
+ "epoch": 10.714285714285714,
112
+ "grad_norm": 1219.8994140625,
113
+ "learning_rate": 5.243636363636364e-06,
114
+ "loss": 116.0246,
115
+ "step": 825
116
+ },
117
+ {
118
+ "epoch": 11.428571428571429,
119
+ "grad_norm": 306.2723083496094,
120
+ "learning_rate": 5.593636363636363e-06,
121
+ "loss": 104.2055,
122
+ "step": 880
123
+ },
124
+ {
125
+ "epoch": 12.142857142857142,
126
+ "grad_norm": 288.2547302246094,
127
+ "learning_rate": 5.943636363636364e-06,
128
+ "loss": 88.9812,
129
+ "step": 935
130
+ },
131
+ {
132
+ "epoch": 12.857142857142858,
133
+ "grad_norm": 1016.5924682617188,
134
+ "learning_rate": 6.293636363636363e-06,
135
+ "loss": 79.9017,
136
+ "step": 990
137
+ },
138
+ {
139
+ "epoch": 13.571428571428571,
140
+ "grad_norm": 239.49191284179688,
141
+ "learning_rate": 6.643636363636363e-06,
142
+ "loss": 73.6932,
143
+ "step": 1045
144
+ },
145
+ {
146
+ "epoch": 14.285714285714286,
147
+ "grad_norm": 193.92002868652344,
148
+ "learning_rate": 6.993636363636364e-06,
149
+ "loss": 62.2018,
150
+ "step": 1100
151
+ },
152
+ {
153
+ "epoch": 15.0,
154
+ "grad_norm": 179.0568084716797,
155
+ "learning_rate": 7.343636363636363e-06,
156
+ "loss": 54.0213,
157
+ "step": 1155
158
+ },
159
+ {
160
+ "epoch": 15.714285714285714,
161
+ "grad_norm": 166.02944946289062,
162
+ "learning_rate": 7.693636363636364e-06,
163
+ "loss": 48.1955,
164
+ "step": 1210
165
+ },
166
+ {
167
+ "epoch": 16.428571428571427,
168
+ "grad_norm": 126.53916931152344,
169
+ "learning_rate": 8.043636363636364e-06,
170
+ "loss": 43.5431,
171
+ "step": 1265
172
+ },
173
+ {
174
+ "epoch": 17.142857142857142,
175
+ "grad_norm": 145.61166381835938,
176
+ "learning_rate": 8.393636363636363e-06,
177
+ "loss": 41.4399,
178
+ "step": 1320
179
+ },
180
+ {
181
+ "epoch": 17.857142857142858,
182
+ "grad_norm": 122.2297134399414,
183
+ "learning_rate": 8.743636363636363e-06,
184
+ "loss": 35.2278,
185
+ "step": 1375
186
+ },
187
+ {
188
+ "epoch": 18.571428571428573,
189
+ "grad_norm": 117.88919067382812,
190
+ "learning_rate": 9.093636363636363e-06,
191
+ "loss": 31.2827,
192
+ "step": 1430
193
+ },
194
+ {
195
+ "epoch": 19.285714285714285,
196
+ "grad_norm": 88.52986907958984,
197
+ "learning_rate": 9.443636363636364e-06,
198
+ "loss": 28.8076,
199
+ "step": 1485
200
+ },
201
+ {
202
+ "epoch": 20.0,
203
+ "grad_norm": 88.49090576171875,
204
+ "learning_rate": 9.793636363636364e-06,
205
+ "loss": 28.232,
206
+ "step": 1540
207
+ },
208
+ {
209
+ "epoch": 20.714285714285715,
210
+ "grad_norm": 114.45001983642578,
211
+ "learning_rate": 1.0143636363636363e-05,
212
+ "loss": 26.0885,
213
+ "step": 1595
214
+ },
215
+ {
216
+ "epoch": 21.428571428571427,
217
+ "grad_norm": 74.16987609863281,
218
+ "learning_rate": 1.0493636363636363e-05,
219
+ "loss": 24.2689,
220
+ "step": 1650
221
+ },
222
+ {
223
+ "epoch": 22.142857142857142,
224
+ "grad_norm": 82.25133514404297,
225
+ "learning_rate": 1.0843636363636363e-05,
226
+ "loss": 22.6963,
227
+ "step": 1705
228
+ },
229
+ {
230
+ "epoch": 22.857142857142858,
231
+ "grad_norm": 74.76679229736328,
232
+ "learning_rate": 1.1193636363636363e-05,
233
+ "loss": 21.3257,
234
+ "step": 1760
235
+ },
236
+ {
237
+ "epoch": 23.571428571428573,
238
+ "grad_norm": 68.91163635253906,
239
+ "learning_rate": 1.1543636363636365e-05,
240
+ "loss": 20.2199,
241
+ "step": 1815
242
+ },
243
+ {
244
+ "epoch": 24.285714285714285,
245
+ "grad_norm": 59.86214828491211,
246
+ "learning_rate": 1.1893636363636363e-05,
247
+ "loss": 19.3769,
248
+ "step": 1870
249
+ },
250
+ {
251
+ "epoch": 25.0,
252
+ "grad_norm": 76.9339370727539,
253
+ "learning_rate": 1.2243636363636363e-05,
254
+ "loss": 18.0471,
255
+ "step": 1925
256
+ },
257
+ {
258
+ "epoch": 25.714285714285715,
259
+ "grad_norm": 66.16152954101562,
260
+ "learning_rate": 1.2593636363636363e-05,
261
+ "loss": 17.496,
262
+ "step": 1980
263
+ },
264
+ {
265
+ "epoch": 26.428571428571427,
266
+ "grad_norm": 58.77778244018555,
267
+ "learning_rate": 1.2943636363636363e-05,
268
+ "loss": 16.9468,
269
+ "step": 2035
270
+ },
271
+ {
272
+ "epoch": 27.142857142857142,
273
+ "grad_norm": 71.95783233642578,
274
+ "learning_rate": 1.3293636363636363e-05,
275
+ "loss": 16.3026,
276
+ "step": 2090
277
+ },
278
+ {
279
+ "epoch": 27.857142857142858,
280
+ "grad_norm": 131.7191619873047,
281
+ "learning_rate": 1.3643636363636363e-05,
282
+ "loss": 16.008,
283
+ "step": 2145
284
+ },
285
+ {
286
+ "epoch": 28.571428571428573,
287
+ "grad_norm": 81.78388977050781,
288
+ "learning_rate": 1.3993636363636363e-05,
289
+ "loss": 15.4616,
290
+ "step": 2200
291
+ },
292
+ {
293
+ "epoch": 29.285714285714285,
294
+ "grad_norm": 60.10112380981445,
295
+ "learning_rate": 1.3628501228501228e-05,
296
+ "loss": 15.195,
297
+ "step": 2255
298
+ },
299
+ {
300
+ "epoch": 30.0,
301
+ "grad_norm": 77.59001922607422,
302
+ "learning_rate": 1.325012285012285e-05,
303
+ "loss": 14.7557,
304
+ "step": 2310
305
+ },
306
+ {
307
+ "epoch": 30.714285714285715,
308
+ "grad_norm": 77.3856201171875,
309
+ "learning_rate": 1.287174447174447e-05,
310
+ "loss": 14.2851,
311
+ "step": 2365
312
+ },
313
+ {
314
+ "epoch": 31.428571428571427,
315
+ "grad_norm": 73.75492095947266,
316
+ "learning_rate": 1.2493366093366094e-05,
317
+ "loss": 13.867,
318
+ "step": 2420
319
+ },
320
+ {
321
+ "epoch": 32.142857142857146,
322
+ "grad_norm": 78.76831817626953,
323
+ "learning_rate": 1.2114987714987713e-05,
324
+ "loss": 13.6677,
325
+ "step": 2475
326
+ },
327
+ {
328
+ "epoch": 32.857142857142854,
329
+ "grad_norm": 92.99346160888672,
330
+ "learning_rate": 1.1736609336609336e-05,
331
+ "loss": 13.1851,
332
+ "step": 2530
333
+ },
334
+ {
335
+ "epoch": 33.57142857142857,
336
+ "grad_norm": 97.57978820800781,
337
+ "learning_rate": 1.1358230958230958e-05,
338
+ "loss": 13.5215,
339
+ "step": 2585
340
+ },
341
+ {
342
+ "epoch": 34.285714285714285,
343
+ "grad_norm": 63.118141174316406,
344
+ "learning_rate": 1.0979852579852579e-05,
345
+ "loss": 12.8525,
346
+ "step": 2640
347
+ },
348
+ {
349
+ "epoch": 35.0,
350
+ "grad_norm": 63.68766403198242,
351
+ "learning_rate": 1.06014742014742e-05,
352
+ "loss": 12.441,
353
+ "step": 2695
354
+ },
355
+ {
356
+ "epoch": 35.714285714285715,
357
+ "grad_norm": 70.33533477783203,
358
+ "learning_rate": 1.0223095823095823e-05,
359
+ "loss": 12.2331,
360
+ "step": 2750
361
+ },
362
+ {
363
+ "epoch": 36.42857142857143,
364
+ "grad_norm": 64.16197967529297,
365
+ "learning_rate": 9.844717444717445e-06,
366
+ "loss": 11.7547,
367
+ "step": 2805
368
+ },
369
+ {
370
+ "epoch": 37.142857142857146,
371
+ "grad_norm": 85.83612823486328,
372
+ "learning_rate": 9.466339066339066e-06,
373
+ "loss": 11.8633,
374
+ "step": 2860
375
+ },
376
+ {
377
+ "epoch": 37.857142857142854,
378
+ "grad_norm": 89.40387725830078,
379
+ "learning_rate": 9.087960687960689e-06,
380
+ "loss": 11.5795,
381
+ "step": 2915
382
+ },
383
+ {
384
+ "epoch": 38.57142857142857,
385
+ "grad_norm": 76.38651275634766,
386
+ "learning_rate": 8.709582309582309e-06,
387
+ "loss": 11.334,
388
+ "step": 2970
389
+ },
390
+ {
391
+ "epoch": 39.285714285714285,
392
+ "grad_norm": 110.11483764648438,
393
+ "learning_rate": 8.331203931203932e-06,
394
+ "loss": 10.805,
395
+ "step": 3025
396
+ },
397
+ {
398
+ "epoch": 40.0,
399
+ "grad_norm": 60.31315612792969,
400
+ "learning_rate": 7.952825552825553e-06,
401
+ "loss": 10.9056,
402
+ "step": 3080
403
+ },
404
+ {
405
+ "epoch": 40.714285714285715,
406
+ "grad_norm": 107.96589660644531,
407
+ "learning_rate": 7.574447174447175e-06,
408
+ "loss": 10.5828,
409
+ "step": 3135
410
+ },
411
+ {
412
+ "epoch": 41.42857142857143,
413
+ "grad_norm": 74.8004379272461,
414
+ "learning_rate": 7.1960687960687955e-06,
415
+ "loss": 10.2818,
416
+ "step": 3190
417
+ },
418
+ {
419
+ "epoch": 42.142857142857146,
420
+ "grad_norm": 54.093475341796875,
421
+ "learning_rate": 6.817690417690418e-06,
422
+ "loss": 10.0785,
423
+ "step": 3245
424
+ },
425
+ {
426
+ "epoch": 42.857142857142854,
427
+ "grad_norm": 115.28564453125,
428
+ "learning_rate": 6.439312039312039e-06,
429
+ "loss": 9.8794,
430
+ "step": 3300
431
+ },
432
+ {
433
+ "epoch": 43.57142857142857,
434
+ "grad_norm": 52.57551956176758,
435
+ "learning_rate": 6.06093366093366e-06,
436
+ "loss": 9.7072,
437
+ "step": 3355
438
+ },
439
+ {
440
+ "epoch": 44.285714285714285,
441
+ "grad_norm": 51.994075775146484,
442
+ "learning_rate": 5.682555282555282e-06,
443
+ "loss": 9.6015,
444
+ "step": 3410
445
+ },
446
+ {
447
+ "epoch": 45.0,
448
+ "grad_norm": 73.08104705810547,
449
+ "learning_rate": 5.304176904176904e-06,
450
+ "loss": 9.305,
451
+ "step": 3465
452
+ },
453
+ {
454
+ "epoch": 45.714285714285715,
455
+ "grad_norm": 75.5807876586914,
456
+ "learning_rate": 4.925798525798525e-06,
457
+ "loss": 9.2281,
458
+ "step": 3520
459
+ },
460
+ {
461
+ "epoch": 46.42857142857143,
462
+ "grad_norm": 53.319637298583984,
463
+ "learning_rate": 4.547420147420147e-06,
464
+ "loss": 8.996,
465
+ "step": 3575
466
+ },
467
+ {
468
+ "epoch": 47.142857142857146,
469
+ "grad_norm": 72.28289794921875,
470
+ "learning_rate": 4.1690417690417685e-06,
471
+ "loss": 9.0072,
472
+ "step": 3630
473
+ },
474
+ {
475
+ "epoch": 47.857142857142854,
476
+ "grad_norm": 66.79710388183594,
477
+ "learning_rate": 3.7906633906633902e-06,
478
+ "loss": 8.7311,
479
+ "step": 3685
480
+ },
481
+ {
482
+ "epoch": 48.57142857142857,
483
+ "grad_norm": 83.55674743652344,
484
+ "learning_rate": 3.4122850122850124e-06,
485
+ "loss": 8.5959,
486
+ "step": 3740
487
+ },
488
+ {
489
+ "epoch": 49.285714285714285,
490
+ "grad_norm": 78.8072738647461,
491
+ "learning_rate": 3.0339066339066337e-06,
492
+ "loss": 8.5462,
493
+ "step": 3795
494
+ },
495
+ {
496
+ "epoch": 50.0,
497
+ "grad_norm": 102.47228240966797,
498
+ "learning_rate": 2.6555282555282554e-06,
499
+ "loss": 8.2928,
500
+ "step": 3850
501
+ },
502
+ {
503
+ "epoch": 50.714285714285715,
504
+ "grad_norm": 93.27884674072266,
505
+ "learning_rate": 2.277149877149877e-06,
506
+ "loss": 8.1427,
507
+ "step": 3905
508
+ },
509
+ {
510
+ "epoch": 51.42857142857143,
511
+ "grad_norm": 137.4209747314453,
512
+ "learning_rate": 1.8987714987714987e-06,
513
+ "loss": 8.164,
514
+ "step": 3960
515
+ },
516
+ {
517
+ "epoch": 52.142857142857146,
518
+ "grad_norm": 38.38957214355469,
519
+ "learning_rate": 1.5203931203931202e-06,
520
+ "loss": 7.8464,
521
+ "step": 4015
522
+ },
523
+ {
524
+ "epoch": 52.857142857142854,
525
+ "grad_norm": 63.25676727294922,
526
+ "learning_rate": 1.1420147420147421e-06,
527
+ "loss": 7.8844,
528
+ "step": 4070
529
+ },
530
+ {
531
+ "epoch": 53.57142857142857,
532
+ "grad_norm": 46.60155487060547,
533
+ "learning_rate": 7.636363636363636e-07,
534
+ "loss": 7.5751,
535
+ "step": 4125
536
+ },
537
+ {
538
+ "epoch": 54.285714285714285,
539
+ "grad_norm": 56.5892448425293,
540
+ "learning_rate": 3.8525798525798526e-07,
541
+ "loss": 7.7376,
542
+ "step": 4180
543
+ }
544
+ ],
545
+ "logging_steps": 55,
546
+ "max_steps": 4235,
547
+ "num_input_tokens_seen": 0,
548
+ "num_train_epochs": 55,
549
+ "save_steps": 220,
550
+ "stateful_callbacks": {
551
+ "TrainerControl": {
552
+ "args": {
553
+ "should_epoch_stop": false,
554
+ "should_evaluate": false,
555
+ "should_log": false,
556
+ "should_save": true,
557
+ "should_training_stop": false
558
+ },
559
+ "attributes": {}
560
+ }
561
+ },
562
+ "total_flos": 6.5802270813998285e+19,
563
+ "train_batch_size": 28,
564
+ "trial_name": null,
565
+ "trial_params": null
566
+ }
checkpoint-4180/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99e55f6c266fb80fff7a194cd0925963d93d119d60f4443cb609bfd824b16fd5
3
+ size 5841
checkpoint-4235/config.json ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "silu",
4
+ "anchor_image_size": null,
5
+ "architectures": [
6
+ "RTDetrV2ForObjectDetection"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auxiliary_loss": true,
10
+ "backbone": null,
11
+ "backbone_config": {
12
+ "depths": [
13
+ 3,
14
+ 4,
15
+ 23,
16
+ 3
17
+ ],
18
+ "downsample_in_bottleneck": false,
19
+ "downsample_in_first_stage": false,
20
+ "embedding_size": 64,
21
+ "hidden_act": "relu",
22
+ "hidden_sizes": [
23
+ 256,
24
+ 512,
25
+ 1024,
26
+ 2048
27
+ ],
28
+ "layer_type": "bottleneck",
29
+ "model_type": "rt_detr_resnet",
30
+ "num_channels": 3,
31
+ "out_features": [
32
+ "stage2",
33
+ "stage3",
34
+ "stage4"
35
+ ],
36
+ "out_indices": [
37
+ 2,
38
+ 3,
39
+ 4
40
+ ],
41
+ "stage_names": [
42
+ "stem",
43
+ "stage1",
44
+ "stage2",
45
+ "stage3",
46
+ "stage4"
47
+ ],
48
+ "torch_dtype": "float32"
49
+ },
50
+ "backbone_kwargs": null,
51
+ "batch_norm_eps": 1e-05,
52
+ "box_noise_scale": 1.0,
53
+ "d_model": 256,
54
+ "decoder_activation_function": "relu",
55
+ "decoder_attention_heads": 8,
56
+ "decoder_ffn_dim": 1024,
57
+ "decoder_in_channels": [
58
+ 384,
59
+ 384,
60
+ 384
61
+ ],
62
+ "decoder_layers": 6,
63
+ "decoder_method": "default",
64
+ "decoder_n_levels": 3,
65
+ "decoder_n_points": 4,
66
+ "decoder_offset_scale": 0.5,
67
+ "disable_custom_kernels": true,
68
+ "dropout": 0.0,
69
+ "encode_proj_layers": [
70
+ 2
71
+ ],
72
+ "encoder_activation_function": "gelu",
73
+ "encoder_attention_heads": 8,
74
+ "encoder_ffn_dim": 2048,
75
+ "encoder_hidden_dim": 384,
76
+ "encoder_in_channels": [
77
+ 512,
78
+ 1024,
79
+ 2048
80
+ ],
81
+ "encoder_layers": 1,
82
+ "eos_coefficient": 0.0001,
83
+ "eval_size": null,
84
+ "feat_strides": [
85
+ 8,
86
+ 16,
87
+ 32
88
+ ],
89
+ "focal_loss_alpha": 0.75,
90
+ "focal_loss_gamma": 2.0,
91
+ "freeze_backbone_batch_norms": true,
92
+ "hidden_expansion": 1.0,
93
+ "id2label": {
94
+ "0": "LABEL_0",
95
+ "1": "LABEL_1",
96
+ "2": "LABEL_2"
97
+ },
98
+ "initializer_bias_prior_prob": null,
99
+ "initializer_range": 0.01,
100
+ "is_encoder_decoder": true,
101
+ "label2id": {
102
+ "LABEL_0": 0,
103
+ "LABEL_1": 1,
104
+ "LABEL_2": 2
105
+ },
106
+ "label_noise_ratio": 0.5,
107
+ "layer_norm_eps": 1e-05,
108
+ "learn_initial_query": false,
109
+ "matcher_alpha": 0.25,
110
+ "matcher_bbox_cost": 5.0,
111
+ "matcher_class_cost": 2.0,
112
+ "matcher_gamma": 2.0,
113
+ "matcher_giou_cost": 2.0,
114
+ "model_type": "rt_detr_v2",
115
+ "normalize_before": false,
116
+ "num_denoising": 100,
117
+ "num_feature_levels": 3,
118
+ "num_queries": 300,
119
+ "positional_encoding_temperature": 10000,
120
+ "torch_dtype": "float32",
121
+ "transformers_version": "4.55.2",
122
+ "use_focal_loss": true,
123
+ "use_pretrained_backbone": false,
124
+ "use_timm_backbone": false,
125
+ "weight_loss_bbox": 5.0,
126
+ "weight_loss_giou": 2.0,
127
+ "weight_loss_vfl": 1.0,
128
+ "with_box_refine": true
129
+ }
checkpoint-4235/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a93c80de10ffd31fe782054b3e02c6be5b97bb74ab13990ee1c9c50346f43fba
3
+ size 306699044
checkpoint-4235/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97305cabfe1039400d2343d85b38ae0e37694a3e342a6319f2687d14ec72953e
3
+ size 611580898
checkpoint-4235/preprocessor_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_annotations": true,
3
+ "do_normalize": false,
4
+ "do_pad": false,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "format": "coco_detection",
8
+ "image_mean": [
9
+ 0.485,
10
+ 0.456,
11
+ 0.406
12
+ ],
13
+ "image_processor_type": "RTDetrImageProcessor",
14
+ "image_std": [
15
+ 0.229,
16
+ 0.224,
17
+ 0.225
18
+ ],
19
+ "pad_size": null,
20
+ "resample": 2,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "height": 640,
24
+ "width": 640
25
+ }
26
+ }
checkpoint-4235/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52710090d6cfecdf10ea9486a80f6062200f2525471ac4971080093c461ef84c
3
+ size 14645
checkpoint-4235/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f961f56765e0678d862d847e28c6b1236181897dd1cc787bb5549277ecbd035c
3
+ size 1465
checkpoint-4235/trainer_state.json ADDED
@@ -0,0 +1,573 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 55.0,
6
+ "eval_steps": 500,
7
+ "global_step": 4235,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.7142857142857143,
14
+ "grad_norm": 4405.81884765625,
15
+ "learning_rate": 3.436363636363636e-07,
16
+ "loss": 1357.5645,
17
+ "step": 55
18
+ },
19
+ {
20
+ "epoch": 1.4285714285714286,
21
+ "grad_norm": 5894.37841796875,
22
+ "learning_rate": 6.936363636363637e-07,
23
+ "loss": 1290.9688,
24
+ "step": 110
25
+ },
26
+ {
27
+ "epoch": 2.142857142857143,
28
+ "grad_norm": 3667.852294921875,
29
+ "learning_rate": 1.0436363636363635e-06,
30
+ "loss": 1140.9515,
31
+ "step": 165
32
+ },
33
+ {
34
+ "epoch": 2.857142857142857,
35
+ "grad_norm": 3578.294677734375,
36
+ "learning_rate": 1.3936363636363637e-06,
37
+ "loss": 942.618,
38
+ "step": 220
39
+ },
40
+ {
41
+ "epoch": 3.571428571428571,
42
+ "grad_norm": 2097.637939453125,
43
+ "learning_rate": 1.7436363636363636e-06,
44
+ "loss": 721.2856,
45
+ "step": 275
46
+ },
47
+ {
48
+ "epoch": 4.285714285714286,
49
+ "grad_norm": 1840.612548828125,
50
+ "learning_rate": 2.0936363636363636e-06,
51
+ "loss": 525.8542,
52
+ "step": 330
53
+ },
54
+ {
55
+ "epoch": 5.0,
56
+ "grad_norm": 1230.875,
57
+ "learning_rate": 2.4436363636363636e-06,
58
+ "loss": 397.7722,
59
+ "step": 385
60
+ },
61
+ {
62
+ "epoch": 5.714285714285714,
63
+ "grad_norm": 1280.699951171875,
64
+ "learning_rate": 2.7936363636363637e-06,
65
+ "loss": 313.8322,
66
+ "step": 440
67
+ },
68
+ {
69
+ "epoch": 6.428571428571429,
70
+ "grad_norm": 4265.28759765625,
71
+ "learning_rate": 3.1436363636363637e-06,
72
+ "loss": 254.1449,
73
+ "step": 495
74
+ },
75
+ {
76
+ "epoch": 7.142857142857143,
77
+ "grad_norm": 967.1757202148438,
78
+ "learning_rate": 3.4936363636363633e-06,
79
+ "loss": 218.4529,
80
+ "step": 550
81
+ },
82
+ {
83
+ "epoch": 7.857142857142857,
84
+ "grad_norm": 765.2302856445312,
85
+ "learning_rate": 3.843636363636364e-06,
86
+ "loss": 197.915,
87
+ "step": 605
88
+ },
89
+ {
90
+ "epoch": 8.571428571428571,
91
+ "grad_norm": 700.331787109375,
92
+ "learning_rate": 4.193636363636364e-06,
93
+ "loss": 170.9019,
94
+ "step": 660
95
+ },
96
+ {
97
+ "epoch": 9.285714285714286,
98
+ "grad_norm": 613.8758544921875,
99
+ "learning_rate": 4.543636363636363e-06,
100
+ "loss": 154.6195,
101
+ "step": 715
102
+ },
103
+ {
104
+ "epoch": 10.0,
105
+ "grad_norm": 3405.25732421875,
106
+ "learning_rate": 4.893636363636364e-06,
107
+ "loss": 136.8394,
108
+ "step": 770
109
+ },
110
+ {
111
+ "epoch": 10.714285714285714,
112
+ "grad_norm": 1219.8994140625,
113
+ "learning_rate": 5.243636363636364e-06,
114
+ "loss": 116.0246,
115
+ "step": 825
116
+ },
117
+ {
118
+ "epoch": 11.428571428571429,
119
+ "grad_norm": 306.2723083496094,
120
+ "learning_rate": 5.593636363636363e-06,
121
+ "loss": 104.2055,
122
+ "step": 880
123
+ },
124
+ {
125
+ "epoch": 12.142857142857142,
126
+ "grad_norm": 288.2547302246094,
127
+ "learning_rate": 5.943636363636364e-06,
128
+ "loss": 88.9812,
129
+ "step": 935
130
+ },
131
+ {
132
+ "epoch": 12.857142857142858,
133
+ "grad_norm": 1016.5924682617188,
134
+ "learning_rate": 6.293636363636363e-06,
135
+ "loss": 79.9017,
136
+ "step": 990
137
+ },
138
+ {
139
+ "epoch": 13.571428571428571,
140
+ "grad_norm": 239.49191284179688,
141
+ "learning_rate": 6.643636363636363e-06,
142
+ "loss": 73.6932,
143
+ "step": 1045
144
+ },
145
+ {
146
+ "epoch": 14.285714285714286,
147
+ "grad_norm": 193.92002868652344,
148
+ "learning_rate": 6.993636363636364e-06,
149
+ "loss": 62.2018,
150
+ "step": 1100
151
+ },
152
+ {
153
+ "epoch": 15.0,
154
+ "grad_norm": 179.0568084716797,
155
+ "learning_rate": 7.343636363636363e-06,
156
+ "loss": 54.0213,
157
+ "step": 1155
158
+ },
159
+ {
160
+ "epoch": 15.714285714285714,
161
+ "grad_norm": 166.02944946289062,
162
+ "learning_rate": 7.693636363636364e-06,
163
+ "loss": 48.1955,
164
+ "step": 1210
165
+ },
166
+ {
167
+ "epoch": 16.428571428571427,
168
+ "grad_norm": 126.53916931152344,
169
+ "learning_rate": 8.043636363636364e-06,
170
+ "loss": 43.5431,
171
+ "step": 1265
172
+ },
173
+ {
174
+ "epoch": 17.142857142857142,
175
+ "grad_norm": 145.61166381835938,
176
+ "learning_rate": 8.393636363636363e-06,
177
+ "loss": 41.4399,
178
+ "step": 1320
179
+ },
180
+ {
181
+ "epoch": 17.857142857142858,
182
+ "grad_norm": 122.2297134399414,
183
+ "learning_rate": 8.743636363636363e-06,
184
+ "loss": 35.2278,
185
+ "step": 1375
186
+ },
187
+ {
188
+ "epoch": 18.571428571428573,
189
+ "grad_norm": 117.88919067382812,
190
+ "learning_rate": 9.093636363636363e-06,
191
+ "loss": 31.2827,
192
+ "step": 1430
193
+ },
194
+ {
195
+ "epoch": 19.285714285714285,
196
+ "grad_norm": 88.52986907958984,
197
+ "learning_rate": 9.443636363636364e-06,
198
+ "loss": 28.8076,
199
+ "step": 1485
200
+ },
201
+ {
202
+ "epoch": 20.0,
203
+ "grad_norm": 88.49090576171875,
204
+ "learning_rate": 9.793636363636364e-06,
205
+ "loss": 28.232,
206
+ "step": 1540
207
+ },
208
+ {
209
+ "epoch": 20.714285714285715,
210
+ "grad_norm": 114.45001983642578,
211
+ "learning_rate": 1.0143636363636363e-05,
212
+ "loss": 26.0885,
213
+ "step": 1595
214
+ },
215
+ {
216
+ "epoch": 21.428571428571427,
217
+ "grad_norm": 74.16987609863281,
218
+ "learning_rate": 1.0493636363636363e-05,
219
+ "loss": 24.2689,
220
+ "step": 1650
221
+ },
222
+ {
223
+ "epoch": 22.142857142857142,
224
+ "grad_norm": 82.25133514404297,
225
+ "learning_rate": 1.0843636363636363e-05,
226
+ "loss": 22.6963,
227
+ "step": 1705
228
+ },
229
+ {
230
+ "epoch": 22.857142857142858,
231
+ "grad_norm": 74.76679229736328,
232
+ "learning_rate": 1.1193636363636363e-05,
233
+ "loss": 21.3257,
234
+ "step": 1760
235
+ },
236
+ {
237
+ "epoch": 23.571428571428573,
238
+ "grad_norm": 68.91163635253906,
239
+ "learning_rate": 1.1543636363636365e-05,
240
+ "loss": 20.2199,
241
+ "step": 1815
242
+ },
243
+ {
244
+ "epoch": 24.285714285714285,
245
+ "grad_norm": 59.86214828491211,
246
+ "learning_rate": 1.1893636363636363e-05,
247
+ "loss": 19.3769,
248
+ "step": 1870
249
+ },
250
+ {
251
+ "epoch": 25.0,
252
+ "grad_norm": 76.9339370727539,
253
+ "learning_rate": 1.2243636363636363e-05,
254
+ "loss": 18.0471,
255
+ "step": 1925
256
+ },
257
+ {
258
+ "epoch": 25.714285714285715,
259
+ "grad_norm": 66.16152954101562,
260
+ "learning_rate": 1.2593636363636363e-05,
261
+ "loss": 17.496,
262
+ "step": 1980
263
+ },
264
+ {
265
+ "epoch": 26.428571428571427,
266
+ "grad_norm": 58.77778244018555,
267
+ "learning_rate": 1.2943636363636363e-05,
268
+ "loss": 16.9468,
269
+ "step": 2035
270
+ },
271
+ {
272
+ "epoch": 27.142857142857142,
273
+ "grad_norm": 71.95783233642578,
274
+ "learning_rate": 1.3293636363636363e-05,
275
+ "loss": 16.3026,
276
+ "step": 2090
277
+ },
278
+ {
279
+ "epoch": 27.857142857142858,
280
+ "grad_norm": 131.7191619873047,
281
+ "learning_rate": 1.3643636363636363e-05,
282
+ "loss": 16.008,
283
+ "step": 2145
284
+ },
285
+ {
286
+ "epoch": 28.571428571428573,
287
+ "grad_norm": 81.78388977050781,
288
+ "learning_rate": 1.3993636363636363e-05,
289
+ "loss": 15.4616,
290
+ "step": 2200
291
+ },
292
+ {
293
+ "epoch": 29.285714285714285,
294
+ "grad_norm": 60.10112380981445,
295
+ "learning_rate": 1.3628501228501228e-05,
296
+ "loss": 15.195,
297
+ "step": 2255
298
+ },
299
+ {
300
+ "epoch": 30.0,
301
+ "grad_norm": 77.59001922607422,
302
+ "learning_rate": 1.325012285012285e-05,
303
+ "loss": 14.7557,
304
+ "step": 2310
305
+ },
306
+ {
307
+ "epoch": 30.714285714285715,
308
+ "grad_norm": 77.3856201171875,
309
+ "learning_rate": 1.287174447174447e-05,
310
+ "loss": 14.2851,
311
+ "step": 2365
312
+ },
313
+ {
314
+ "epoch": 31.428571428571427,
315
+ "grad_norm": 73.75492095947266,
316
+ "learning_rate": 1.2493366093366094e-05,
317
+ "loss": 13.867,
318
+ "step": 2420
319
+ },
320
+ {
321
+ "epoch": 32.142857142857146,
322
+ "grad_norm": 78.76831817626953,
323
+ "learning_rate": 1.2114987714987713e-05,
324
+ "loss": 13.6677,
325
+ "step": 2475
326
+ },
327
+ {
328
+ "epoch": 32.857142857142854,
329
+ "grad_norm": 92.99346160888672,
330
+ "learning_rate": 1.1736609336609336e-05,
331
+ "loss": 13.1851,
332
+ "step": 2530
333
+ },
334
+ {
335
+ "epoch": 33.57142857142857,
336
+ "grad_norm": 97.57978820800781,
337
+ "learning_rate": 1.1358230958230958e-05,
338
+ "loss": 13.5215,
339
+ "step": 2585
340
+ },
341
+ {
342
+ "epoch": 34.285714285714285,
343
+ "grad_norm": 63.118141174316406,
344
+ "learning_rate": 1.0979852579852579e-05,
345
+ "loss": 12.8525,
346
+ "step": 2640
347
+ },
348
+ {
349
+ "epoch": 35.0,
350
+ "grad_norm": 63.68766403198242,
351
+ "learning_rate": 1.06014742014742e-05,
352
+ "loss": 12.441,
353
+ "step": 2695
354
+ },
355
+ {
356
+ "epoch": 35.714285714285715,
357
+ "grad_norm": 70.33533477783203,
358
+ "learning_rate": 1.0223095823095823e-05,
359
+ "loss": 12.2331,
360
+ "step": 2750
361
+ },
362
+ {
363
+ "epoch": 36.42857142857143,
364
+ "grad_norm": 64.16197967529297,
365
+ "learning_rate": 9.844717444717445e-06,
366
+ "loss": 11.7547,
367
+ "step": 2805
368
+ },
369
+ {
370
+ "epoch": 37.142857142857146,
371
+ "grad_norm": 85.83612823486328,
372
+ "learning_rate": 9.466339066339066e-06,
373
+ "loss": 11.8633,
374
+ "step": 2860
375
+ },
376
+ {
377
+ "epoch": 37.857142857142854,
378
+ "grad_norm": 89.40387725830078,
379
+ "learning_rate": 9.087960687960689e-06,
380
+ "loss": 11.5795,
381
+ "step": 2915
382
+ },
383
+ {
384
+ "epoch": 38.57142857142857,
385
+ "grad_norm": 76.38651275634766,
386
+ "learning_rate": 8.709582309582309e-06,
387
+ "loss": 11.334,
388
+ "step": 2970
389
+ },
390
+ {
391
+ "epoch": 39.285714285714285,
392
+ "grad_norm": 110.11483764648438,
393
+ "learning_rate": 8.331203931203932e-06,
394
+ "loss": 10.805,
395
+ "step": 3025
396
+ },
397
+ {
398
+ "epoch": 40.0,
399
+ "grad_norm": 60.31315612792969,
400
+ "learning_rate": 7.952825552825553e-06,
401
+ "loss": 10.9056,
402
+ "step": 3080
403
+ },
404
+ {
405
+ "epoch": 40.714285714285715,
406
+ "grad_norm": 107.96589660644531,
407
+ "learning_rate": 7.574447174447175e-06,
408
+ "loss": 10.5828,
409
+ "step": 3135
410
+ },
411
+ {
412
+ "epoch": 41.42857142857143,
413
+ "grad_norm": 74.8004379272461,
414
+ "learning_rate": 7.1960687960687955e-06,
415
+ "loss": 10.2818,
416
+ "step": 3190
417
+ },
418
+ {
419
+ "epoch": 42.142857142857146,
420
+ "grad_norm": 54.093475341796875,
421
+ "learning_rate": 6.817690417690418e-06,
422
+ "loss": 10.0785,
423
+ "step": 3245
424
+ },
425
+ {
426
+ "epoch": 42.857142857142854,
427
+ "grad_norm": 115.28564453125,
428
+ "learning_rate": 6.439312039312039e-06,
429
+ "loss": 9.8794,
430
+ "step": 3300
431
+ },
432
+ {
433
+ "epoch": 43.57142857142857,
434
+ "grad_norm": 52.57551956176758,
435
+ "learning_rate": 6.06093366093366e-06,
436
+ "loss": 9.7072,
437
+ "step": 3355
438
+ },
439
+ {
440
+ "epoch": 44.285714285714285,
441
+ "grad_norm": 51.994075775146484,
442
+ "learning_rate": 5.682555282555282e-06,
443
+ "loss": 9.6015,
444
+ "step": 3410
445
+ },
446
+ {
447
+ "epoch": 45.0,
448
+ "grad_norm": 73.08104705810547,
449
+ "learning_rate": 5.304176904176904e-06,
450
+ "loss": 9.305,
451
+ "step": 3465
452
+ },
453
+ {
454
+ "epoch": 45.714285714285715,
455
+ "grad_norm": 75.5807876586914,
456
+ "learning_rate": 4.925798525798525e-06,
457
+ "loss": 9.2281,
458
+ "step": 3520
459
+ },
460
+ {
461
+ "epoch": 46.42857142857143,
462
+ "grad_norm": 53.319637298583984,
463
+ "learning_rate": 4.547420147420147e-06,
464
+ "loss": 8.996,
465
+ "step": 3575
466
+ },
467
+ {
468
+ "epoch": 47.142857142857146,
469
+ "grad_norm": 72.28289794921875,
470
+ "learning_rate": 4.1690417690417685e-06,
471
+ "loss": 9.0072,
472
+ "step": 3630
473
+ },
474
+ {
475
+ "epoch": 47.857142857142854,
476
+ "grad_norm": 66.79710388183594,
477
+ "learning_rate": 3.7906633906633902e-06,
478
+ "loss": 8.7311,
479
+ "step": 3685
480
+ },
481
+ {
482
+ "epoch": 48.57142857142857,
483
+ "grad_norm": 83.55674743652344,
484
+ "learning_rate": 3.4122850122850124e-06,
485
+ "loss": 8.5959,
486
+ "step": 3740
487
+ },
488
+ {
489
+ "epoch": 49.285714285714285,
490
+ "grad_norm": 78.8072738647461,
491
+ "learning_rate": 3.0339066339066337e-06,
492
+ "loss": 8.5462,
493
+ "step": 3795
494
+ },
495
+ {
496
+ "epoch": 50.0,
497
+ "grad_norm": 102.47228240966797,
498
+ "learning_rate": 2.6555282555282554e-06,
499
+ "loss": 8.2928,
500
+ "step": 3850
501
+ },
502
+ {
503
+ "epoch": 50.714285714285715,
504
+ "grad_norm": 93.27884674072266,
505
+ "learning_rate": 2.277149877149877e-06,
506
+ "loss": 8.1427,
507
+ "step": 3905
508
+ },
509
+ {
510
+ "epoch": 51.42857142857143,
511
+ "grad_norm": 137.4209747314453,
512
+ "learning_rate": 1.8987714987714987e-06,
513
+ "loss": 8.164,
514
+ "step": 3960
515
+ },
516
+ {
517
+ "epoch": 52.142857142857146,
518
+ "grad_norm": 38.38957214355469,
519
+ "learning_rate": 1.5203931203931202e-06,
520
+ "loss": 7.8464,
521
+ "step": 4015
522
+ },
523
+ {
524
+ "epoch": 52.857142857142854,
525
+ "grad_norm": 63.25676727294922,
526
+ "learning_rate": 1.1420147420147421e-06,
527
+ "loss": 7.8844,
528
+ "step": 4070
529
+ },
530
+ {
531
+ "epoch": 53.57142857142857,
532
+ "grad_norm": 46.60155487060547,
533
+ "learning_rate": 7.636363636363636e-07,
534
+ "loss": 7.5751,
535
+ "step": 4125
536
+ },
537
+ {
538
+ "epoch": 54.285714285714285,
539
+ "grad_norm": 56.5892448425293,
540
+ "learning_rate": 3.8525798525798526e-07,
541
+ "loss": 7.7376,
542
+ "step": 4180
543
+ },
544
+ {
545
+ "epoch": 55.0,
546
+ "grad_norm": 87.8992919921875,
547
+ "learning_rate": 6.879606879606879e-09,
548
+ "loss": 7.5818,
549
+ "step": 4235
550
+ }
551
+ ],
552
+ "logging_steps": 55,
553
+ "max_steps": 4235,
554
+ "num_input_tokens_seen": 0,
555
+ "num_train_epochs": 55,
556
+ "save_steps": 220,
557
+ "stateful_callbacks": {
558
+ "TrainerControl": {
559
+ "args": {
560
+ "should_epoch_stop": false,
561
+ "should_evaluate": false,
562
+ "should_log": false,
563
+ "should_save": true,
564
+ "should_training_stop": true
565
+ },
566
+ "attributes": {}
567
+ }
568
+ },
569
+ "total_flos": 6.666743797117747e+19,
570
+ "train_batch_size": 28,
571
+ "trial_name": null,
572
+ "trial_params": null
573
+ }
checkpoint-4235/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99e55f6c266fb80fff7a194cd0925963d93d119d60f4443cb609bfd824b16fd5
3
+ size 5841
config.json CHANGED
@@ -118,7 +118,7 @@
118
  "num_queries": 300,
119
  "positional_encoding_temperature": 10000,
120
  "torch_dtype": "float32",
121
- "transformers_version": "4.55.0",
122
  "use_focal_loss": true,
123
  "use_pretrained_backbone": false,
124
  "use_timm_backbone": false,
 
118
  "num_queries": 300,
119
  "positional_encoding_temperature": 10000,
120
  "torch_dtype": "float32",
121
+ "transformers_version": "4.55.2",
122
  "use_focal_loss": true,
123
  "use_pretrained_backbone": false,
124
  "use_timm_backbone": false,
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d3a52d4cd4386295eedbfb267bc679eca4b27864d745fff06694c0f9dbf823a6
3
  size 306699044
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a93c80de10ffd31fe782054b3e02c6be5b97bb74ab13990ee1c9c50346f43fba
3
  size 306699044
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a7b84367094b7487f77de50fba614a6c6667e9cf018b77ee5bfc158268fc5eaf
3
- size 5368
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99e55f6c266fb80fff7a194cd0925963d93d119d60f4443cb609bfd824b16fd5
3
+ size 5841