Training in progress, step 4500, checkpoint

checkpoint-4500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:df4c8fc65b08faf7ffe95fcea36dc60fd997bb382afe6b5c6271b066a794328e
+oid sha256:7fe53fd40f951233c7a44fa7ac57779e5749ac4177502b54f7bf70d948671fe3
 size 1711848436

checkpoint-4500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:763842a72e3cd9176eb7d356c26697895a99eeff9d57be85df56a5f1e4477037
+oid sha256:9aa225279deb4ba7b2103bd043155153d6d73e36dab090719fa9e212edbe46ce
 size 3424043887

checkpoint-4500/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 1.0729899406433105,
+  "best_metric": 1.0721209049224854,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-4500",
   "epoch": 0.8290346352247605,
   "eval_steps": 500,
@@ -10,137 +10,137 @@
   "log_history": [
     {
       "epoch": 0.09211495946941783,
-      "grad_norm": 20.28152847290039,
+      "grad_norm": 20.04043960571289,
       "learning_rate": 4.8848563006632274e-08,
       "loss": 0.3758,
       "step": 500
     },
     {
       "epoch": 0.09211495946941783,
-      "eval_loss": 1.418367862701416,
-      "eval_runtime": 73.4439,
-      "eval_samples_per_second": 16.434,
-      "eval_steps_per_second": 2.056,
+      "eval_loss": 1.4184927940368652,
+      "eval_runtime": 74.3562,
+      "eval_samples_per_second": 16.233,
+      "eval_steps_per_second": 2.031,
       "step": 500
     },
     {
       "epoch": 0.18422991893883567,
-      "grad_norm": 47.29538345336914,
+      "grad_norm": 74.25025939941406,
       "learning_rate": 4.7697126013264556e-08,
-      "loss": 0.4104,
+      "loss": 0.4103,
       "step": 1000
     },
     {
       "epoch": 0.18422991893883567,
-      "eval_loss": 1.3491536378860474,
-      "eval_runtime": 74.3404,
-      "eval_samples_per_second": 16.236,
-      "eval_steps_per_second": 2.031,
+      "eval_loss": 1.3500770330429077,
+      "eval_runtime": 75.7013,
+      "eval_samples_per_second": 15.944,
+      "eval_steps_per_second": 1.995,
       "step": 1000
     },
     {
       "epoch": 0.2763448784082535,
-      "grad_norm": 0.5450155138969421,
+      "grad_norm": 0.5102410912513733,
       "learning_rate": 4.6545689019896826e-08,
-      "loss": 0.4325,
+      "loss": 0.433,
       "step": 1500
     },
     {
       "epoch": 0.2763448784082535,
-      "eval_loss": 1.2877168655395508,
-      "eval_runtime": 75.3303,
-      "eval_samples_per_second": 16.023,
-      "eval_steps_per_second": 2.005,
+      "eval_loss": 1.2885024547576904,
+      "eval_runtime": 75.197,
+      "eval_samples_per_second": 16.051,
+      "eval_steps_per_second": 2.008,
       "step": 1500
     },
     {
       "epoch": 0.36845983787767134,
-      "grad_norm": 0.11449883133172989,
+      "grad_norm": 0.1667678952217102,
       "learning_rate": 4.539425202652911e-08,
-      "loss": 0.3423,
+      "loss": 0.3424,
       "step": 2000
     },
     {
       "epoch": 0.36845983787767134,
-      "eval_loss": 1.2384389638900757,
-      "eval_runtime": 76.113,
-      "eval_samples_per_second": 15.858,
-      "eval_steps_per_second": 1.984,
+      "eval_loss": 1.239119052886963,
+      "eval_runtime": 75.387,
+      "eval_samples_per_second": 16.011,
+      "eval_steps_per_second": 2.003,
       "step": 2000
     },
     {
       "epoch": 0.46057479734708917,
-      "grad_norm": 389.7380065917969,
+      "grad_norm": 343.8175048828125,
       "learning_rate": 4.4242815033161385e-08,
       "loss": 0.3645,
       "step": 2500
     },
     {
       "epoch": 0.46057479734708917,
-      "eval_loss": 1.1896915435791016,
-      "eval_runtime": 76.23,
-      "eval_samples_per_second": 15.834,
-      "eval_steps_per_second": 1.981,
+      "eval_loss": 1.1902339458465576,
+      "eval_runtime": 75.9899,
+      "eval_samples_per_second": 15.884,
+      "eval_steps_per_second": 1.987,
       "step": 2500
     },
     {
       "epoch": 0.552689756816507,
-      "grad_norm": 340.1103820800781,
+      "grad_norm": 305.5626525878906,
       "learning_rate": 4.309137803979366e-08,
-      "loss": 0.3171,
+      "loss": 0.3172,
       "step": 3000
     },
     {
       "epoch": 0.552689756816507,
-      "eval_loss": 1.149291753768921,
-      "eval_runtime": 76.1662,
-      "eval_samples_per_second": 15.847,
-      "eval_steps_per_second": 1.983,
+      "eval_loss": 1.1506118774414062,
+      "eval_runtime": 74.8655,
+      "eval_samples_per_second": 16.122,
+      "eval_steps_per_second": 2.017,
       "step": 3000
     },
     {
       "epoch": 0.6448047162859248,
-      "grad_norm": 196.7362518310547,
+      "grad_norm": 65.59048461914062,
       "learning_rate": 4.193994104642594e-08,
-      "loss": 0.275,
+      "loss": 0.2751,
       "step": 3500
     },
     {
       "epoch": 0.6448047162859248,
-      "eval_loss": 1.1162768602371216,
-      "eval_runtime": 74.6569,
-      "eval_samples_per_second": 16.167,
-      "eval_steps_per_second": 2.023,
+      "eval_loss": 1.1169357299804688,
+      "eval_runtime": 76.2936,
+      "eval_samples_per_second": 15.82,
+      "eval_steps_per_second": 1.979,
       "step": 3500
     },
     {
       "epoch": 0.7369196757553427,
-      "grad_norm": 16.604211807250977,
+      "grad_norm": 18.106201171875,
       "learning_rate": 4.078850405305821e-08,
-      "loss": 0.2907,
+      "loss": 0.2919,
       "step": 4000
     },
     {
       "epoch": 0.7369196757553427,
-      "eval_loss": 1.091786503791809,
-      "eval_runtime": 76.3333,
-      "eval_samples_per_second": 15.812,
-      "eval_steps_per_second": 1.978,
+      "eval_loss": 1.0920671224594116,
+      "eval_runtime": 74.9101,
+      "eval_samples_per_second": 16.113,
+      "eval_steps_per_second": 2.016,
       "step": 4000
     },
     {
       "epoch": 0.8290346352247605,
-      "grad_norm": 527.0987548828125,
+      "grad_norm": 506.40570068359375,
       "learning_rate": 3.9637067059690496e-08,
-      "loss": 0.2581,
+      "loss": 0.2583,
       "step": 4500
     },
     {
       "epoch": 0.8290346352247605,
-      "eval_loss": 1.0729899406433105,
-      "eval_runtime": 74.4912,
-      "eval_samples_per_second": 16.203,
-      "eval_steps_per_second": 2.027,
+      "eval_loss": 1.0721209049224854,
+      "eval_runtime": 75.8053,
+      "eval_samples_per_second": 15.922,
+      "eval_steps_per_second": 1.992,
       "step": 4500
     }
   ],

checkpoint-4500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:595bb6f39a89ec7030e395471c015665757cb5cacaafa6bba1613fd32b629686
+oid sha256:9a29c2374e59ce3f57b34db9cb20d2d1bdc870c3dad128e9a4d462636c37e8b0
 size 4847