Training in progress, step 500, checkpoint

checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:33bf090ca0bd2461884b36e6d1605d91cc5b8976d68e35be28f30cdd758eedc3
+oid sha256:825cba45d25a650cea0a0e03c6049e07df257e03a5f755758293bace233bd686
 size 1711848436

checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0aa9a6e1f4f1d296ddb1b26332eb06c7a63552f2f5c25016065420e9d986cd87
+oid sha256:69a42cefb582c4fe5b20fcdef56430a24937f92fb57e1373bc37aba606a8e659
 size 3424043887

checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cbdc4dbb33632d2bae03a8a45932156c19761e16dbcfe2d06603895693890b8c
+oid sha256:f84bd46537ef272b6c9aa04d96003f0a8243652a06e83d15e664a08212aaacf7
 size 623

checkpoint-500/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.069305419921875,
+  "best_metric": 1.1265727281570435,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-500",
-  "epoch": 0.09211495946941783,
+  "epoch": 0.056287290329843524,
   "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": 3.0293657779693604,
-      "learning_rate": 4.884856300663227e-07,
-      "loss": 0.271,
+      "epoch": 0.056287290329843524,
+      "grad_norm": 150.1642303466797,
+      "learning_rate": 4.929640887087696e-07,
+      "loss": 0.3486,
       "step": 500
     },
     {
-      "epoch": 0.09211495946941783,
-      "eval_loss": 1.069305419921875,
-      "eval_runtime": 75.1409,
-      "eval_samples_per_second": 16.063,
-      "eval_steps_per_second": 2.01,
+      "epoch": 0.056287290329843524,
+      "eval_loss": 1.1265727281570435,
+      "eval_runtime": 123.7806,
+      "eval_samples_per_second": 15.948,
+      "eval_steps_per_second": 1.995,
       "step": 500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 21712,
+  "max_steps": 35532,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,

checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bf76416e4a857628175385db7f617e54451cd7ffb62c1f9590ee9f30268f8c5b
+oid sha256:319be2a769ef7117b6826908976f080f9082f159f4f0ad75da1a151ad29d9bff
 size 4847