Training in progress, epoch 1

adapter_config.json

@@ -13,15 +13,15 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 16,
-  "lora_dropout": 0.05,
+  "lora_alpha": 32,
+  "lora_dropout": 0.1,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": [
     "classifier"
   ],
   "peft_type": "LORA",
-  "r": 32,
+  "r": 8,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7327e969b05815e106811199049307f7a8b07bc6a554682b44a565dec9b6965b
-size 4731912
+oid sha256:18e9275c994e42478d67744083436f40b2e9dbca8acc3e43b81b8a9a67117116
+size 1192848

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 5.0,
-    "total_flos": 3.405838778276659e+17,
-    "train_loss": 0.6811685698372977,
-    "train_runtime": 162.9509,
-    "train_samples_per_second": 26.787,
-    "train_steps_per_second": 0.215
+    "epoch": 10.0,
+    "total_flos": 6.858183153733632e+17,
+    "train_loss": 0.6843075002942767,
+    "train_runtime": 315.6994,
+    "train_samples_per_second": 27.653,
+    "train_steps_per_second": 0.222
 }

runs/Aug15_14-14-13_c03d0dbfda2e/events.out.tfevents.1723731254.c03d0dbfda2e.1885.9

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b828ee4c7d6f0d28bb619362ff52d1f8d6efe3eab5c05ff0a7da379f9165c245
+size 5879

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 5.0,
-    "total_flos": 3.405838778276659e+17,
-    "train_loss": 0.6811685698372977,
-    "train_runtime": 162.9509,
-    "train_samples_per_second": 26.787,
-    "train_steps_per_second": 0.215
+    "epoch": 10.0,
+    "total_flos": 6.858183153733632e+17,
+    "train_loss": 0.6843075002942767,
+    "train_runtime": 315.6994,
+    "train_samples_per_second": 27.653,
+    "train_steps_per_second": 0.222
 }

trainer_state.json

@@ -1,93 +1,166 @@
 {
-  "best_metric": 0.5714285714285714,
-  "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-qlora-houseplant/checkpoint-35",
-  "epoch": 5.0,
+  "best_metric": 0.5204081632653061,
+  "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-qlora-houseplant/checkpoint-21",
+  "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 35,
+  "global_step": 70,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5204081632653061,
-      "eval_loss": 0.6893933415412903,
-      "eval_runtime": 2.0373,
-      "eval_samples_per_second": 48.103,
-      "eval_steps_per_second": 1.963,
+      "eval_accuracy": 0.5102040816326531,
+      "eval_loss": 0.6916877627372742,
+      "eval_runtime": 1.9776,
+      "eval_samples_per_second": 49.555,
+      "eval_steps_per_second": 2.023,
       "step": 7
     },
     {
       "epoch": 1.4285714285714286,
-      "grad_norm": 0.5420740246772766,
-      "learning_rate": 4.032258064516129e-05,
-      "loss": 0.6869,
+      "grad_norm": 0.5460406541824341,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.688,
       "step": 10
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5306122448979592,
-      "eval_loss": 0.6856390237808228,
-      "eval_runtime": 2.0142,
-      "eval_samples_per_second": 48.654,
-      "eval_steps_per_second": 1.986,
+      "eval_accuracy": 0.5102040816326531,
+      "eval_loss": 0.6909428834915161,
+      "eval_runtime": 2.0178,
+      "eval_samples_per_second": 48.568,
+      "eval_steps_per_second": 1.982,
       "step": 14
     },
     {
       "epoch": 2.857142857142857,
-      "grad_norm": 0.44206592440605164,
-      "learning_rate": 2.4193548387096777e-05,
-      "loss": 0.6797,
+      "grad_norm": 0.44764864444732666,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.6845,
       "step": 20
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.5612244897959183,
-      "eval_loss": 0.6827691793441772,
-      "eval_runtime": 2.3921,
-      "eval_samples_per_second": 40.969,
-      "eval_steps_per_second": 1.672,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.690205454826355,
+      "eval_runtime": 2.828,
+      "eval_samples_per_second": 34.654,
+      "eval_steps_per_second": 1.414,
       "step": 21
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.5612244897959183,
-      "eval_loss": 0.6809878945350647,
-      "eval_runtime": 2.0128,
-      "eval_samples_per_second": 48.688,
-      "eval_steps_per_second": 1.987,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.6895502805709839,
+      "eval_runtime": 1.9885,
+      "eval_samples_per_second": 49.284,
+      "eval_steps_per_second": 2.012,
       "step": 28
     },
     {
       "epoch": 4.285714285714286,
-      "grad_norm": 0.47674253582954407,
-      "learning_rate": 8.064516129032258e-06,
-      "loss": 0.6808,
+      "grad_norm": 0.48582252860069275,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.6874,
       "step": 30
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.5714285714285714,
-      "eval_loss": 0.6803476810455322,
-      "eval_runtime": 1.9919,
-      "eval_samples_per_second": 49.198,
-      "eval_steps_per_second": 2.008,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.6889798045158386,
+      "eval_runtime": 1.9771,
+      "eval_samples_per_second": 49.568,
+      "eval_steps_per_second": 2.023,
       "step": 35
     },
     {
-      "epoch": 5.0,
-      "step": 35,
-      "total_flos": 3.405838778276659e+17,
-      "train_loss": 0.6811685698372977,
-      "train_runtime": 162.9509,
-      "train_samples_per_second": 26.787,
-      "train_steps_per_second": 0.215
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.4863179922103882,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.6838,
+      "step": 40
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.6885014772415161,
+      "eval_runtime": 2.3711,
+      "eval_samples_per_second": 41.331,
+      "eval_steps_per_second": 1.687,
+      "step": 42
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.6881576776504517,
+      "eval_runtime": 1.9786,
+      "eval_samples_per_second": 49.529,
+      "eval_steps_per_second": 2.022,
+      "step": 49
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.5103937983512878,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.6817,
+      "step": 50
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.6879384517669678,
+      "eval_runtime": 1.9723,
+      "eval_samples_per_second": 49.687,
+      "eval_steps_per_second": 2.028,
+      "step": 56
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 0.45278120040893555,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 0.6813,
+      "step": 60
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.6877266764640808,
+      "eval_runtime": 1.9657,
+      "eval_samples_per_second": 49.856,
+      "eval_steps_per_second": 2.035,
+      "step": 63
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.6086100935935974,
+      "learning_rate": 0.0,
+      "loss": 0.6835,
+      "step": 70
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.5204081632653061,
+      "eval_loss": 0.6876619458198547,
+      "eval_runtime": 2.0199,
+      "eval_samples_per_second": 48.516,
+      "eval_steps_per_second": 1.98,
+      "step": 70
+    },
+    {
+      "epoch": 10.0,
+      "step": 70,
+      "total_flos": 6.858183153733632e+17,
+      "train_loss": 0.6843075002942767,
+      "train_runtime": 315.6994,
+      "train_samples_per_second": 27.653,
+      "train_steps_per_second": 0.222
     }
   ],
   "logging_steps": 10,
-  "max_steps": 35,
+  "max_steps": 70,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 10,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -101,7 +174,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.405838778276659e+17,
+  "total_flos": 6.858183153733632e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a2816daebe68496c7c177291686be8bff9b425ea88a7c6c51e2422304bdb6f5
+oid sha256:f10a19cc8dfab556e0dda2df9795ef3682fa534007bae539204011121769e8d1
 size 5304