Training in progress, step 4096

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8611b00fb559665ca26e68dae93dfb4ebdf3717ac3a1f548409581450a0a18a4
+oid sha256:4224574c9e85152ad061f5cfd88e5ade72f84f20aa6d778d16c5a0ab0ab115b4
 size 105666297

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:908dc25e857dbdefbfe009b9ed0b992ce31760b393bf814f450b0d3c701c2397
+oid sha256:bdf489f059ab44f9bc26200d314d3fc5954dabfb501b51ecc19cdc4d4be8a527
 size 3579

training_args.json

@@ -6,34 +6,34 @@
   "do_predict": false,
   "evaluation_strategy": "steps",
   "prediction_loss_only": false,
-  "per_device_train_batch_size": 10,
+  "per_device_train_batch_size": 9,
   "per_device_eval_batch_size": 8,
   "per_gpu_train_batch_size": null,
   "per_gpu_eval_batch_size": null,
   "gradient_accumulation_steps": 1,
   "eval_accumulation_steps": null,
   "eval_delay": 0,
-  "learning_rate": 0.0005,
-  "weight_decay": 0.1,
+  "learning_rate": 5e-05,
+  "weight_decay": 0.0,
   "adam_beta1": 0.9,
   "adam_beta2": 0.999,
   "adam_epsilon": 1e-08,
   "max_grad_norm": 1.0,
-  "num_train_epochs": 6,
+  "num_train_epochs": 5,
   "max_steps": -1,
-  "lr_scheduler_type": "cosine",
+  "lr_scheduler_type": "linear",
   "warmup_ratio": 0.0,
-  "warmup_steps": 200,
+  "warmup_steps": 0,
   "log_level": "passive",
   "log_level_replica": "passive",
   "log_on_each_node": true,
   "logging_dir": "models/elec-gmusic-familized/logs",
   "logging_strategy": "steps",
   "logging_first_step": false,
-  "logging_steps": 4096,
+  "logging_steps": 1024,
   "logging_nan_inf_filter": true,
   "save_strategy": "steps",
-  "save_steps": 16384,
+  "save_steps": 4096,
   "save_total_limit": 5,
   "save_on_each_node": false,
   "no_cuda": false,
@@ -55,7 +55,7 @@
   "tpu_metrics_debug": false,
   "debug": [],
   "dataloader_drop_last": false,
-  "eval_steps": 4096,
+  "eval_steps": 1024,
   "dataloader_num_workers": 0,
   "past_index": -1,
   "run_name": "models/elec-gmusic-familized",