Training in progress, step 4000, checkpoint

checkpoint-4000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7782509cca1d07a21b8c556e21997e4446816b0885fe09b7b4501df0f55c8406
+oid sha256:c6469994a0d1fc083714cea684b90c729f20440980c9e3c369536781409ba2cc
 size 1711848436

checkpoint-4000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:662b1bb95d062a22928ca4dab2237e2b124d95d4a116054642669f0efe6430bc
+oid sha256:b05b37fb9f4d8fefe347d971c817a797b4260825a4e7d04837d0973be711f9d0
 size 3424043887

checkpoint-4000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1454cbc8ef4e92d532f823e5816216a6cd4ee18fc15f9b51dc40aa621c880224
+oid sha256:44c9fa5b803fb1f28770699a359bea30c519543808cc61771c36f033b35e9034
 size 623

checkpoint-4000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.0954023599624634,
+  "best_metric": 1.091786503791809,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-4000",
-  "epoch": 0.7367839381101492,
+  "epoch": 0.7369196757553427,
   "eval_steps": 500,
   "global_step": 4000,
   "is_hyper_param_search": false,
@@ -9,128 +9,128 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.09209799226376865,
-      "grad_norm": 0.0014697719598188996,
-      "learning_rate": 4.884877509670289e-08,
-      "loss": 0.4588,
+      "epoch": 0.09211495946941783,
+      "grad_norm": 20.28152847290039,
+      "learning_rate": 4.8848563006632274e-08,
+      "loss": 0.3758,
       "step": 500
     },
     {
-      "epoch": 0.09209799226376865,
-      "eval_loss": 1.4194883108139038,
-      "eval_runtime": 78.7703,
-      "eval_samples_per_second": 15.31,
-      "eval_steps_per_second": 1.917,
+      "epoch": 0.09211495946941783,
+      "eval_loss": 1.418367862701416,
+      "eval_runtime": 73.4439,
+      "eval_samples_per_second": 16.434,
+      "eval_steps_per_second": 2.056,
       "step": 500
     },
     {
-      "epoch": 0.1841959845275373,
-      "grad_norm": 301.76043701171875,
-      "learning_rate": 4.769755019340578e-08,
-      "loss": 0.4255,
+      "epoch": 0.18422991893883567,
+      "grad_norm": 47.29538345336914,
+      "learning_rate": 4.7697126013264556e-08,
+      "loss": 0.4104,
       "step": 1000
     },
     {
-      "epoch": 0.1841959845275373,
-      "eval_loss": 1.3416553735733032,
-      "eval_runtime": 77.0051,
-      "eval_samples_per_second": 15.661,
-      "eval_steps_per_second": 1.961,
+      "epoch": 0.18422991893883567,
+      "eval_loss": 1.3491536378860474,
+      "eval_runtime": 74.3404,
+      "eval_samples_per_second": 16.236,
+      "eval_steps_per_second": 2.031,
       "step": 1000
     },
     {
-      "epoch": 0.27629397679130596,
-      "grad_norm": 35.28853225708008,
-      "learning_rate": 4.654632529010867e-08,
-      "loss": 0.3724,
+      "epoch": 0.2763448784082535,
+      "grad_norm": 0.5450155138969421,
+      "learning_rate": 4.6545689019896826e-08,
+      "loss": 0.4325,
       "step": 1500
     },
     {
-      "epoch": 0.27629397679130596,
-      "eval_loss": 1.2872973680496216,
-      "eval_runtime": 77.5439,
-      "eval_samples_per_second": 15.552,
-      "eval_steps_per_second": 1.947,
+      "epoch": 0.2763448784082535,
+      "eval_loss": 1.2877168655395508,
+      "eval_runtime": 75.3303,
+      "eval_samples_per_second": 16.023,
+      "eval_steps_per_second": 2.005,
       "step": 1500
     },
     {
-      "epoch": 0.3683919690550746,
-      "grad_norm": 338.9349060058594,
-      "learning_rate": 4.539510038681156e-08,
-      "loss": 0.3251,
+      "epoch": 0.36845983787767134,
+      "grad_norm": 0.11449883133172989,
+      "learning_rate": 4.539425202652911e-08,
+      "loss": 0.3423,
       "step": 2000
     },
     {
-      "epoch": 0.3683919690550746,
-      "eval_loss": 1.234910249710083,
-      "eval_runtime": 78.7715,
-      "eval_samples_per_second": 15.31,
-      "eval_steps_per_second": 1.917,
+      "epoch": 0.36845983787767134,
+      "eval_loss": 1.2384389638900757,
+      "eval_runtime": 76.113,
+      "eval_samples_per_second": 15.858,
+      "eval_steps_per_second": 1.984,
       "step": 2000
     },
     {
-      "epoch": 0.46048996131884323,
-      "grad_norm": 325.85101318359375,
-      "learning_rate": 4.4243875483514457e-08,
-      "loss": 0.3308,
+      "epoch": 0.46057479734708917,
+      "grad_norm": 389.7380065917969,
+      "learning_rate": 4.4242815033161385e-08,
+      "loss": 0.3645,
       "step": 2500
     },
     {
-      "epoch": 0.46048996131884323,
-      "eval_loss": 1.1944907903671265,
-      "eval_runtime": 79.0397,
-      "eval_samples_per_second": 15.258,
-      "eval_steps_per_second": 1.91,
+      "epoch": 0.46057479734708917,
+      "eval_loss": 1.1896915435791016,
+      "eval_runtime": 76.23,
+      "eval_samples_per_second": 15.834,
+      "eval_steps_per_second": 1.981,
       "step": 2500
     },
     {
-      "epoch": 0.5525879535826119,
-      "grad_norm": 100.85855102539062,
-      "learning_rate": 4.309265058021735e-08,
-      "loss": 0.3017,
+      "epoch": 0.552689756816507,
+      "grad_norm": 340.1103820800781,
+      "learning_rate": 4.309137803979366e-08,
+      "loss": 0.3171,
       "step": 3000
     },
     {
-      "epoch": 0.5525879535826119,
-      "eval_loss": 1.1593009233474731,
-      "eval_runtime": 77.6354,
-      "eval_samples_per_second": 15.534,
-      "eval_steps_per_second": 1.945,
+      "epoch": 0.552689756816507,
+      "eval_loss": 1.149291753768921,
+      "eval_runtime": 76.1662,
+      "eval_samples_per_second": 15.847,
+      "eval_steps_per_second": 1.983,
       "step": 3000
     },
     {
-      "epoch": 0.6446859458463805,
-      "grad_norm": 8.865598678588867,
-      "learning_rate": 4.194142567692024e-08,
-      "loss": 0.2962,
+      "epoch": 0.6448047162859248,
+      "grad_norm": 196.7362518310547,
+      "learning_rate": 4.193994104642594e-08,
+      "loss": 0.275,
       "step": 3500
     },
     {
-      "epoch": 0.6446859458463805,
-      "eval_loss": 1.1259427070617676,
-      "eval_runtime": 78.2715,
-      "eval_samples_per_second": 15.408,
-      "eval_steps_per_second": 1.929,
+      "epoch": 0.6448047162859248,
+      "eval_loss": 1.1162768602371216,
+      "eval_runtime": 74.6569,
+      "eval_samples_per_second": 16.167,
+      "eval_steps_per_second": 2.023,
       "step": 3500
     },
     {
-      "epoch": 0.7367839381101492,
-      "grad_norm": 4.393447399139404,
-      "learning_rate": 4.079020077362314e-08,
-      "loss": 0.2919,
+      "epoch": 0.7369196757553427,
+      "grad_norm": 16.604211807250977,
+      "learning_rate": 4.078850405305821e-08,
+      "loss": 0.2907,
       "step": 4000
     },
     {
-      "epoch": 0.7367839381101492,
-      "eval_loss": 1.0954023599624634,
-      "eval_runtime": 78.5366,
-      "eval_samples_per_second": 15.356,
-      "eval_steps_per_second": 1.923,
+      "epoch": 0.7369196757553427,
+      "eval_loss": 1.091786503791809,
+      "eval_runtime": 76.3333,
+      "eval_samples_per_second": 15.812,
+      "eval_steps_per_second": 1.978,
       "step": 4000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 21716,
+  "max_steps": 21712,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,

checkpoint-4000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:417b0bacc3a8958dfa75621764230fb890db48c09ca241dff24e5264d406285b
+oid sha256:595bb6f39a89ec7030e395471c015665757cb5cacaafa6bba1613fd32b629686
 size 4847