Training in progress, step 2000, checkpoint

checkpoint-2000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fdbe2454b5a4b7954cd8c80488092d4a71b8e30baa9eefc2e153f84383450bcd
+oid sha256:6dbaa4a0e5441d379dc6454d81f608c57b1afad59996b3e3769b26b416b0408b
 size 1711848436

checkpoint-2000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3b2afef50395c1b100767c3b2d015915d99465b148e4a408eb5a3790cb72f3da
+oid sha256:9ad69267022abd5c076fa8211eada06def9b19a93f8ac9f9d3b6ef635b4a7c72
 size 3424043887

checkpoint-2000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9fedc9d05d8c53e20b8390de6f1f8081f1a35d4ebd835fabc1f6569e6c78c6e0
+oid sha256:503c1f86f0ef5e725d6adbda6d7362a3646fd00b433b16a4df7ce6f007cad7e4
 size 623

checkpoint-2000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.239119052886963,
+  "best_metric": 1.2890639305114746,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-2000",
-  "epoch": 0.36845983787767134,
+  "epoch": 0.21331058020477817,
   "eval_steps": 500,
   "global_step": 2000,
   "is_hyper_param_search": false,
@@ -9,68 +9,68 @@
   "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
     },
     {
-      "epoch": 0.18422991893883567,
-      "grad_norm": 74.25025939941406,
-      "learning_rate": 4.7697126013264556e-08,
-      "loss": 0.4103,
+      "epoch": 0.10665529010238908,
+      "grad_norm": 73.73340606689453,
+      "learning_rate": 4.8666808873720136e-08,
+      "loss": 0.4532,
       "step": 1000
     },
     {
-      "epoch": 0.18422991893883567,
-      "eval_loss": 1.3500770330429077,
-      "eval_runtime": 75.7013,
-      "eval_samples_per_second": 15.944,
-      "eval_steps_per_second": 1.995,
+      "epoch": 0.10665529010238908,
+      "eval_loss": 1.3815597295761108,
+      "eval_runtime": 62.4624,
+      "eval_samples_per_second": 15.801,
+      "eval_steps_per_second": 1.985,
       "step": 1000
     },
     {
-      "epoch": 0.2763448784082535,
-      "grad_norm": 0.5102410912513733,
-      "learning_rate": 4.6545689019896826e-08,
-      "loss": 0.433,
+      "epoch": 0.1599829351535836,
+      "grad_norm": 471.4011535644531,
+      "learning_rate": 4.80002133105802e-08,
+      "loss": 0.3749,
       "step": 1500
     },
     {
-      "epoch": 0.2763448784082535,
-      "eval_loss": 1.2885024547576904,
-      "eval_runtime": 75.197,
-      "eval_samples_per_second": 16.051,
-      "eval_steps_per_second": 2.008,
+      "epoch": 0.1599829351535836,
+      "eval_loss": 1.3310539722442627,
+      "eval_runtime": 63.3918,
+      "eval_samples_per_second": 15.57,
+      "eval_steps_per_second": 1.956,
       "step": 1500
     },
     {
-      "epoch": 0.36845983787767134,
-      "grad_norm": 0.1667678952217102,
-      "learning_rate": 4.539425202652911e-08,
-      "loss": 0.3424,
+      "epoch": 0.21331058020477817,
+      "grad_norm": 11.962693214416504,
+      "learning_rate": 4.733361774744027e-08,
+      "loss": 0.336,
       "step": 2000
     },
     {
-      "epoch": 0.36845983787767134,
-      "eval_loss": 1.239119052886963,
-      "eval_runtime": 75.387,
-      "eval_samples_per_second": 16.011,
-      "eval_steps_per_second": 2.003,
+      "epoch": 0.21331058020477817,
+      "eval_loss": 1.2890639305114746,
+      "eval_runtime": 63.177,
+      "eval_samples_per_second": 15.623,
+      "eval_steps_per_second": 1.963,
       "step": 2000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 21712,
+  "max_steps": 37504,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,

checkpoint-2000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a29c2374e59ce3f57b34db9cb20d2d1bdc870c3dad128e9a4d462636c37e8b0
+oid sha256:87e35dca972e13467c750106c19301b098fbd2918fda1087487e227bac3aca37
 size 4847