Training in progress, step 500, checkpoint

checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4dca7fc3cd3d0989b032cde2275d646f77adbb94a714e2c65009c12d5da693dd
+oid sha256:92794c01e17fb5d2e58f58a93e653d787c687f106903ee1a32d13a936d5a69c1
 size 1711848436

checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4941b5bd547acd9e9f6310def5e251770e2f02207c754218ea6229523709e43d
+oid sha256:723987b2ce3c78f994182c79dbcde2a55e61fd63b56f6870430dccbcfb91a792
 size 3424043887

checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c34c95efbb3e774ffebdb18cba7fdab3e47828c6f40b41baaf5bcea3c7763d1d
+oid sha256:1a4e19b58b280f583064f9384acd0ce9f1ada1d37c6782ec85c13c003d2fab8a
 size 623

checkpoint-500/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.4184927940368652,
+  "best_metric": 1.4426143169403076,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-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": 20.04043960571289,
-      "learning_rate": 4.8848563006632274e-08,
-      "loss": 0.3758,
+      "epoch": 0.05332764505119454,
+      "grad_norm": 217.55360412597656,
+      "learning_rate": 4.933340443686007e-08,
+      "loss": 0.4667,
       "step": 500
     },
     {
-      "epoch": 0.09211495946941783,
-      "eval_loss": 1.4184927940368652,
-      "eval_runtime": 74.3562,
-      "eval_samples_per_second": 16.233,
-      "eval_steps_per_second": 2.031,
+      "epoch": 0.05332764505119454,
+      "eval_loss": 1.4426143169403076,
+      "eval_runtime": 61.9901,
+      "eval_samples_per_second": 15.922,
+      "eval_steps_per_second": 2.0,
       "step": 500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 21712,
+  "max_steps": 37504,
   "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 sha256:9a29c2374e59ce3f57b34db9cb20d2d1bdc870c3dad128e9a4d462636c37e8b0
+oid sha256:87e35dca972e13467c750106c19301b098fbd2918fda1087487e227bac3aca37
 size 4847