End of training

README.md

@@ -3,6 +3,9 @@ library_name: transformers
 license: other
 base_model: facebook/mask2former-swin-tiny-coco-instance
 tags:
+- image-segmentation
+- instance-segmentation
+- vision
 - generated_from_trainer
 model-index:
 - name: finetune-instance-segmentation-ade20k-mini-mask2former
@@ -14,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # finetune-instance-segmentation-ade20k-mini-mask2former
 
-This model is a fine-tuned version of [facebook/mask2former-swin-tiny-coco-instance](https://huggingface.co/facebook/mask2former-swin-tiny-coco-instance) on an unknown dataset.
+This model is a fine-tuned version of [facebook/mask2former-swin-tiny-coco-instance](https://huggingface.co/facebook/mask2former-swin-tiny-coco-instance) on the qubvel-hf/ade20k-mini dataset.
 It achieves the following results on the evaluation set:
 - Loss: 31.0067
 - Map: 0.2061

all_results.json

@@ -1,28 +1,28 @@
 {
-    "epoch": 4.0,
-    "test_loss": 28.440181732177734,
-    "test_map": 0.2172,
-    "test_map_50": 0.4234,
-    "test_map_75": 0.2041,
-    "test_map_car": 0.2867,
-    "test_map_large": 0.8076,
-    "test_map_medium": 0.6353,
-    "test_map_person": 0.1476,
-    "test_map_small": 0.1458,
-    "test_mar_1": 0.0953,
-    "test_mar_10": 0.254,
-    "test_mar_100": 0.2903,
-    "test_mar_100_car": 0.3755,
-    "test_mar_100_person": 0.205,
-    "test_mar_large": 0.8594,
-    "test_mar_medium": 0.7113,
-    "test_mar_small": 0.2169,
-    "test_runtime": 61.8717,
-    "test_samples_per_second": 5.334,
-    "test_steps_per_second": 0.679,
-    "total_flos": 3.574777289637888e+17,
-    "train_loss": 28.362115478515626,
-    "train_runtime": 686.6708,
-    "train_samples_per_second": 9.32,
-    "train_steps_per_second": 0.583
+    "epoch": 2.0,
+    "test_loss": 31.059144973754883,
+    "test_map": 0.2061,
+    "test_map_50": 0.4076,
+    "test_map_75": 0.1946,
+    "test_map_car": 0.2709,
+    "test_map_large": 0.82,
+    "test_map_medium": 0.623,
+    "test_map_person": 0.1413,
+    "test_map_small": 0.1368,
+    "test_mar_1": 0.0921,
+    "test_mar_10": 0.2488,
+    "test_mar_100": 0.2856,
+    "test_mar_100_car": 0.3678,
+    "test_mar_100_person": 0.2035,
+    "test_mar_large": 0.8705,
+    "test_mar_medium": 0.7164,
+    "test_mar_small": 0.2105,
+    "test_runtime": 57.7838,
+    "test_samples_per_second": 5.711,
+    "test_steps_per_second": 0.727,
+    "total_flos": 1.787388644818944e+17,
+    "train_loss": 33.218726806640625,
+    "train_runtime": 300.7183,
+    "train_samples_per_second": 10.641,
+    "train_steps_per_second": 0.665
 }

runs/Jan02_18-52-26_mothership/events.out.tfevents.1735824523.mothership.226887.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fe6ad8d89adf9607a71d0a1911ee443d59ff55ae18b953431e7e63c02386bb2
+size 1194

test_results.json

@@ -1,23 +1,23 @@
 {
-    "epoch": 4.0,
-    "test_loss": 28.440181732177734,
-    "test_map": 0.2172,
-    "test_map_50": 0.4234,
-    "test_map_75": 0.2041,
-    "test_map_car": 0.2867,
-    "test_map_large": 0.8076,
-    "test_map_medium": 0.6353,
-    "test_map_person": 0.1476,
-    "test_map_small": 0.1458,
-    "test_mar_1": 0.0953,
-    "test_mar_10": 0.254,
-    "test_mar_100": 0.2903,
-    "test_mar_100_car": 0.3755,
-    "test_mar_100_person": 0.205,
-    "test_mar_large": 0.8594,
-    "test_mar_medium": 0.7113,
-    "test_mar_small": 0.2169,
-    "test_runtime": 61.8717,
-    "test_samples_per_second": 5.334,
-    "test_steps_per_second": 0.679
+    "epoch": 2.0,
+    "test_loss": 31.059144973754883,
+    "test_map": 0.2061,
+    "test_map_50": 0.4076,
+    "test_map_75": 0.1946,
+    "test_map_car": 0.2709,
+    "test_map_large": 0.82,
+    "test_map_medium": 0.623,
+    "test_map_person": 0.1413,
+    "test_map_small": 0.1368,
+    "test_mar_1": 0.0921,
+    "test_mar_10": 0.2488,
+    "test_mar_100": 0.2856,
+    "test_mar_100_car": 0.3678,
+    "test_mar_100_person": 0.2035,
+    "test_mar_large": 0.8705,
+    "test_mar_medium": 0.7164,
+    "test_mar_small": 0.2105,
+    "test_runtime": 57.7838,
+    "test_samples_per_second": 5.711,
+    "test_steps_per_second": 0.727
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.0,
-    "total_flos": 3.574777289637888e+17,
-    "train_loss": 28.362115478515626,
-    "train_runtime": 686.6708,
-    "train_samples_per_second": 9.32,
-    "train_steps_per_second": 0.583
+    "epoch": 2.0,
+    "total_flos": 1.787388644818944e+17,
+    "train_loss": 33.218726806640625,
+    "train_runtime": 300.7183,
+    "train_samples_per_second": 10.641,
+    "train_steps_per_second": 0.665
 }

trainer_state.json

@@ -1,151 +1,89 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.0,
+  "epoch": 2.0,
   "eval_steps": 500,
-  "global_step": 400,
+  "global_step": 200,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 294.3598327636719,
+      "grad_norm": 618.759521484375,
       "learning_rate": 1e-05,
-      "loss": 33.5412,
+      "loss": 37.3934,
       "step": 100
     },
     {
       "epoch": 1.0,
-      "eval_loss": 31.532806396484375,
-      "eval_map": 0.1929,
-      "eval_map_50": 0.3918,
-      "eval_map_75": 0.1737,
-      "eval_map_car": 0.2558,
-      "eval_map_large": 0.7895,
-      "eval_map_medium": 0.6122,
-      "eval_map_person": 0.13,
-      "eval_map_small": 0.1281,
-      "eval_mar_1": 0.0904,
-      "eval_mar_10": 0.2473,
-      "eval_mar_100": 0.2836,
-      "eval_mar_100_car": 0.3672,
+      "eval_loss": 33.63319778442383,
+      "eval_map": 0.1981,
+      "eval_map_50": 0.3956,
+      "eval_map_75": 0.1799,
+      "eval_map_car": 0.2613,
+      "eval_map_large": 0.7992,
+      "eval_map_medium": 0.6173,
+      "eval_map_person": 0.1349,
+      "eval_map_small": 0.1285,
+      "eval_mar_1": 0.0896,
+      "eval_mar_10": 0.2453,
+      "eval_mar_100": 0.2821,
+      "eval_mar_100_car": 0.3641,
       "eval_mar_100_person": 0.2001,
-      "eval_mar_large": 0.8229,
-      "eval_mar_medium": 0.7063,
-      "eval_mar_small": 0.2105,
-      "eval_runtime": 55.4841,
-      "eval_samples_per_second": 5.948,
-      "eval_steps_per_second": 0.757,
+      "eval_mar_large": 0.8354,
+      "eval_mar_medium": 0.7135,
+      "eval_mar_small": 0.2074,
+      "eval_runtime": 54.9214,
+      "eval_samples_per_second": 6.009,
+      "eval_steps_per_second": 0.765,
       "step": 100
     },
     {
       "epoch": 2.0,
-      "grad_norm": 104.88582611083984,
+      "grad_norm": 113.26927947998047,
       "learning_rate": 1e-05,
-      "loss": 27.9471,
+      "loss": 29.0441,
       "step": 200
     },
     {
       "epoch": 2.0,
-      "eval_loss": 29.718120574951172,
-      "eval_map": 0.2053,
-      "eval_map_50": 0.4151,
-      "eval_map_75": 0.1851,
-      "eval_map_car": 0.2741,
-      "eval_map_large": 0.8018,
-      "eval_map_medium": 0.6192,
-      "eval_map_person": 0.1364,
-      "eval_map_small": 0.1387,
-      "eval_mar_1": 0.093,
-      "eval_mar_10": 0.2507,
-      "eval_mar_100": 0.2872,
-      "eval_mar_100_car": 0.3714,
-      "eval_mar_100_person": 0.2029,
-      "eval_mar_large": 0.8323,
-      "eval_mar_medium": 0.7079,
-      "eval_mar_small": 0.2142,
-      "eval_runtime": 61.2008,
-      "eval_samples_per_second": 5.392,
-      "eval_steps_per_second": 0.686,
+      "eval_loss": 31.006668090820312,
+      "eval_map": 0.2061,
+      "eval_map_50": 0.4076,
+      "eval_map_75": 0.1946,
+      "eval_map_car": 0.2709,
+      "eval_map_large": 0.82,
+      "eval_map_medium": 0.623,
+      "eval_map_person": 0.1413,
+      "eval_map_small": 0.1368,
+      "eval_mar_1": 0.0921,
+      "eval_mar_10": 0.2488,
+      "eval_mar_100": 0.2856,
+      "eval_mar_100_car": 0.3678,
+      "eval_mar_100_person": 0.2035,
+      "eval_mar_large": 0.8705,
+      "eval_mar_medium": 0.7164,
+      "eval_mar_small": 0.2105,
+      "eval_runtime": 58.669,
+      "eval_samples_per_second": 5.625,
+      "eval_steps_per_second": 0.716,
       "step": 200
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 156.72207641601562,
-      "learning_rate": 1e-05,
-      "loss": 26.4855,
-      "step": 300
-    },
-    {
-      "epoch": 3.0,
-      "eval_loss": 28.978605270385742,
-      "eval_map": 0.2134,
-      "eval_map_50": 0.4219,
-      "eval_map_75": 0.1945,
-      "eval_map_car": 0.2837,
-      "eval_map_large": 0.8047,
-      "eval_map_medium": 0.6255,
-      "eval_map_person": 0.143,
-      "eval_map_small": 0.1451,
-      "eval_mar_1": 0.0944,
-      "eval_mar_10": 0.2543,
-      "eval_mar_100": 0.2918,
-      "eval_mar_100_car": 0.3777,
-      "eval_mar_100_person": 0.2059,
-      "eval_mar_large": 0.8594,
-      "eval_mar_medium": 0.7045,
-      "eval_mar_small": 0.2198,
-      "eval_runtime": 61.9496,
-      "eval_samples_per_second": 5.327,
-      "eval_steps_per_second": 0.678,
-      "step": 300
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 159.6126708984375,
-      "learning_rate": 1e-05,
-      "loss": 25.4746,
-      "step": 400
-    },
-    {
-      "epoch": 4.0,
-      "eval_loss": 28.44811248779297,
-      "eval_map": 0.2172,
-      "eval_map_50": 0.4234,
-      "eval_map_75": 0.2041,
-      "eval_map_car": 0.2867,
-      "eval_map_large": 0.8076,
-      "eval_map_medium": 0.6353,
-      "eval_map_person": 0.1476,
-      "eval_map_small": 0.1458,
-      "eval_mar_1": 0.0953,
-      "eval_mar_10": 0.254,
-      "eval_mar_100": 0.2903,
-      "eval_mar_100_car": 0.3755,
-      "eval_mar_100_person": 0.205,
-      "eval_mar_large": 0.8594,
-      "eval_mar_medium": 0.7113,
-      "eval_mar_small": 0.2169,
-      "eval_runtime": 62.2166,
-      "eval_samples_per_second": 5.304,
-      "eval_steps_per_second": 0.675,
-      "step": 400
-    },
-    {
-      "epoch": 4.0,
-      "step": 400,
-      "total_flos": 3.574777289637888e+17,
-      "train_loss": 28.362115478515626,
-      "train_runtime": 686.6708,
-      "train_samples_per_second": 9.32,
-      "train_steps_per_second": 0.583
+      "epoch": 2.0,
+      "step": 200,
+      "total_flos": 1.787388644818944e+17,
+      "train_loss": 33.218726806640625,
+      "train_runtime": 300.7183,
+      "train_samples_per_second": 10.641,
+      "train_steps_per_second": 0.665
     }
   ],
   "logging_steps": 500,
-  "max_steps": 400,
+  "max_steps": 200,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 2,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -159,7 +97,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.574777289637888e+17,
+  "total_flos": 1.787388644818944e+17,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null