Training in progress, step 500, checkpoint

checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c721bc7ae861f75163ea461855741324940b8a117a8b33871156fcb46d2f545f
+oid sha256:1c394ef32bad0da7182b6da4fe9a4c404fdac7efe4d993e10f91eb9a2dcca110
 size 1711848436

checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d8f3314bd1c0e15796eaa8a74ee2f395850ac71a2416ab6979f3f9880e5eec30
+oid sha256:d75856d80a4753b5c4e7c0e28e1c04e6d1c291a8406cfcaa92d87b9f4d6d6c29
 size 3424043887

checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7d437d27261ee9193fe564e196242ccf13107ef626a115f7aa6d8c69f4ee32f6
+oid sha256:6c0d56903807f766a4bec5c194a762961a70804a7aa52532b1ba59ac19c7b303
 size 623

checkpoint-500/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.418845295906067,
+  "best_metric": 1.4427576065063477,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l58-l/checkpoint-500",
-  "epoch": 0.09211495946941783,
+  "epoch": 0.05332764505119454,
   "eval_steps": 500,
   "global_step": 500,
   "is_hyper_param_search": false,
@@ -9,23 +9,23 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.09211495946941783,
-      "grad_norm": 19.808820724487305,
-      "learning_rate": 4.84647506755097e-08,
-      "loss": 0.3758,
+      "epoch": 0.05332764505119454,
+      "grad_norm": 217.40109252929688,
+      "learning_rate": 4.911120591581342e-08,
+      "loss": 0.4667,
       "step": 500
     },
     {
-      "epoch": 0.09211495946941783,
-      "eval_loss": 1.418845295906067,
-      "eval_runtime": 75.2377,
-      "eval_samples_per_second": 16.042,
-      "eval_steps_per_second": 2.007,
+      "epoch": 0.05332764505119454,
+      "eval_loss": 1.4427576065063477,
+      "eval_runtime": 60.328,
+      "eval_samples_per_second": 16.361,
+      "eval_steps_per_second": 2.055,
       "step": 500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 16284,
+  "max_steps": 28128,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,

checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cbcece3d46fd39b0ba80eb3e495fdf46c9f891af5b4bddce9de77ad5a467b625
+oid sha256:590a3eed3f4b2ba9fcb0b831d5e4c3f2410a46be8f57155ebf9fdd97c6c96524
 size 4847