Model save

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 6.6440677966101696,
+    "eval_loss": 2.6779351234436035,
+    "eval_runtime": 16.1015,
+    "eval_samples_per_second": 43.971,
+    "eval_steps_per_second": 5.527,
+    "total_flos": 5.749859310737818e+16,
+    "train_loss": 0.9609878695740992,
+    "train_runtime": 1272.9991,
+    "train_samples_per_second": 15.556,
+    "train_steps_per_second": 0.077
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 6.6440677966101696,
+    "eval_loss": 2.6779351234436035,
+    "eval_runtime": 16.1015,
+    "eval_samples_per_second": 43.971,
+    "eval_steps_per_second": 5.527
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 6.6440677966101696,
+    "total_flos": 5.749859310737818e+16,
+    "train_loss": 0.9609878695740992,
+    "train_runtime": 1272.9991,
+    "train_samples_per_second": 15.556,
+    "train_steps_per_second": 0.077
+}

trainer_state.json

@@ -0,0 +1,194 @@
+{
+  "best_metric": 2.6791086196899414,
+  "best_model_checkpoint": "./results/checkpoint-50",
+  "epoch": 6.6440677966101696,
+  "eval_steps": 10,
+  "global_step": 98,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 0.3754754960536957,
+      "learning_rate": 0.0002,
+      "loss": 1.4178,
+      "step": 10
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "eval_loss": 2.233613967895508,
+      "eval_runtime": 16.1303,
+      "eval_samples_per_second": 43.893,
+      "eval_steps_per_second": 5.518,
+      "step": 10
+    },
+    {
+      "epoch": 1.3559322033898304,
+      "grad_norm": 0.5044791102409363,
+      "learning_rate": 0.0002,
+      "loss": 1.3907,
+      "step": 20
+    },
+    {
+      "epoch": 1.3559322033898304,
+      "eval_loss": 2.3868091106414795,
+      "eval_runtime": 16.1072,
+      "eval_samples_per_second": 43.956,
+      "eval_steps_per_second": 5.525,
+      "step": 20
+    },
+    {
+      "epoch": 2.0338983050847457,
+      "grad_norm": 0.6073395609855652,
+      "learning_rate": 0.0002,
+      "loss": 1.2803,
+      "step": 30
+    },
+    {
+      "epoch": 2.0338983050847457,
+      "eval_loss": 2.409191846847534,
+      "eval_runtime": 16.1264,
+      "eval_samples_per_second": 43.903,
+      "eval_steps_per_second": 5.519,
+      "step": 30
+    },
+    {
+      "epoch": 2.711864406779661,
+      "grad_norm": 0.6424660682678223,
+      "learning_rate": 0.0002,
+      "loss": 1.1003,
+      "step": 40
+    },
+    {
+      "epoch": 2.711864406779661,
+      "eval_loss": 2.614776372909546,
+      "eval_runtime": 16.1817,
+      "eval_samples_per_second": 43.753,
+      "eval_steps_per_second": 5.5,
+      "step": 40
+    },
+    {
+      "epoch": 3.389830508474576,
+      "grad_norm": 0.8125430345535278,
+      "learning_rate": 0.0002,
+      "loss": 0.9902,
+      "step": 50
+    },
+    {
+      "epoch": 3.389830508474576,
+      "eval_loss": 2.6791086196899414,
+      "eval_runtime": 16.1831,
+      "eval_samples_per_second": 43.749,
+      "eval_steps_per_second": 5.5,
+      "step": 50
+    },
+    {
+      "epoch": 4.067796610169491,
+      "grad_norm": 0.7281854748725891,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 60
+    },
+    {
+      "epoch": 4.067796610169491,
+      "eval_loss": 2.880143404006958,
+      "eval_runtime": 16.1,
+      "eval_samples_per_second": 43.975,
+      "eval_steps_per_second": 5.528,
+      "step": 60
+    },
+    {
+      "epoch": 4.745762711864407,
+      "grad_norm": 0.7661570310592651,
+      "learning_rate": 0.0002,
+      "loss": 0.7266,
+      "step": 70
+    },
+    {
+      "epoch": 4.745762711864407,
+      "eval_loss": 2.900453805923462,
+      "eval_runtime": 16.041,
+      "eval_samples_per_second": 44.137,
+      "eval_steps_per_second": 5.548,
+      "step": 70
+    },
+    {
+      "epoch": 5.423728813559322,
+      "grad_norm": 0.7970860600471497,
+      "learning_rate": 0.0002,
+      "loss": 0.66,
+      "step": 80
+    },
+    {
+      "epoch": 5.423728813559322,
+      "eval_loss": 3.0775656700134277,
+      "eval_runtime": 16.0539,
+      "eval_samples_per_second": 44.101,
+      "eval_steps_per_second": 5.544,
+      "step": 80
+    },
+    {
+      "epoch": 6.101694915254237,
+      "grad_norm": 0.7553396224975586,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 90
+    },
+    {
+      "epoch": 6.101694915254237,
+      "eval_loss": 3.3246374130249023,
+      "eval_runtime": 16.0984,
+      "eval_samples_per_second": 43.979,
+      "eval_steps_per_second": 5.528,
+      "step": 90
+    },
+    {
+      "epoch": 6.6440677966101696,
+      "step": 98,
+      "total_flos": 5.749859310737818e+16,
+      "train_loss": 0.9609878695740992,
+      "train_runtime": 1272.9991,
+      "train_samples_per_second": 15.556,
+      "train_steps_per_second": 0.077
+    },
+    {
+      "epoch": 6.6440677966101696,
+      "eval_loss": 2.679919481277466,
+      "eval_runtime": 16.1806,
+      "eval_samples_per_second": 43.756,
+      "eval_steps_per_second": 5.5,
+      "step": 98
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 98,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 7,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.01
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 4
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.749859310737818e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}