End of training

README.md

@@ -3,6 +3,9 @@ base_model: meta-llama/Llama-3.1-8B
 library_name: peft
 license: llama3.1
 tags:
+- question-answering
+- QA
+- text-generation
 - trl
 - sft
 - generated_from_trainer
@@ -16,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Llama-3.1-8B-medquad-V2
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on the None dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on the MedQuAD: Ben-Abacha and Demner-Fushman (2019) dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.8959
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.630937880633374,
+    "eval_loss": 0.8958781361579895,
+    "eval_runtime": 294.9962,
+    "eval_samples_per_second": 5.563,
+    "eval_steps_per_second": 0.698,
+    "total_flos": 3.658397647269151e+17,
+    "train_loss": 0.9498361190160115,
+    "train_runtime": 17857.7741,
+    "train_samples_per_second": 5.145,
+    "train_steps_per_second": 0.027
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.630937880633374,
+    "eval_loss": 0.8958781361579895,
+    "eval_runtime": 294.9962,
+    "eval_samples_per_second": 5.563,
+    "eval_steps_per_second": 0.698
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.630937880633374,
+    "total_flos": 3.658397647269151e+17,
+    "train_loss": 0.9498361190160115,
+    "train_runtime": 17857.7741,
+    "train_samples_per_second": 5.145,
+    "train_steps_per_second": 0.027
+}

trainer_state.json

@@ -0,0 +1,321 @@
+{
+  "best_metric": 0.8958781361579895,
+  "best_model_checkpoint": "Llama-3.1-8B-medquad-V2/checkpoint-180",
+  "epoch": 2.630937880633374,
+  "eval_steps": 10,
+  "global_step": 180,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.146163215590743,
+      "grad_norm": 0.3197868764400482,
+      "learning_rate": 0.0002,
+      "loss": 1.2503,
+      "step": 10
+    },
+    {
+      "epoch": 0.146163215590743,
+      "eval_loss": 1.1359219551086426,
+      "eval_runtime": 294.7811,
+      "eval_samples_per_second": 5.567,
+      "eval_steps_per_second": 0.699,
+      "step": 10
+    },
+    {
+      "epoch": 0.292326431181486,
+      "grad_norm": 0.24248264729976654,
+      "learning_rate": 0.0002,
+      "loss": 1.1182,
+      "step": 20
+    },
+    {
+      "epoch": 0.292326431181486,
+      "eval_loss": 1.0198906660079956,
+      "eval_runtime": 294.7536,
+      "eval_samples_per_second": 5.567,
+      "eval_steps_per_second": 0.699,
+      "step": 20
+    },
+    {
+      "epoch": 0.438489646772229,
+      "grad_norm": 0.21424424648284912,
+      "learning_rate": 0.0002,
+      "loss": 1.0864,
+      "step": 30
+    },
+    {
+      "epoch": 0.438489646772229,
+      "eval_loss": 0.9855759143829346,
+      "eval_runtime": 294.9371,
+      "eval_samples_per_second": 5.564,
+      "eval_steps_per_second": 0.698,
+      "step": 30
+    },
+    {
+      "epoch": 0.584652862362972,
+      "grad_norm": 0.1644379198551178,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 40
+    },
+    {
+      "epoch": 0.584652862362972,
+      "eval_loss": 0.9680554866790771,
+      "eval_runtime": 294.8273,
+      "eval_samples_per_second": 5.566,
+      "eval_steps_per_second": 0.699,
+      "step": 40
+    },
+    {
+      "epoch": 0.730816077953715,
+      "grad_norm": 0.3253229260444641,
+      "learning_rate": 0.0002,
+      "loss": 1.0773,
+      "step": 50
+    },
+    {
+      "epoch": 0.730816077953715,
+      "eval_loss": 0.9498887062072754,
+      "eval_runtime": 294.9458,
+      "eval_samples_per_second": 5.564,
+      "eval_steps_per_second": 0.698,
+      "step": 50
+    },
+    {
+      "epoch": 0.876979293544458,
+      "grad_norm": 0.17308901250362396,
+      "learning_rate": 0.0002,
+      "loss": 0.9575,
+      "step": 60
+    },
+    {
+      "epoch": 0.876979293544458,
+      "eval_loss": 0.9426676630973816,
+      "eval_runtime": 294.8574,
+      "eval_samples_per_second": 5.565,
+      "eval_steps_per_second": 0.699,
+      "step": 60
+    },
+    {
+      "epoch": 1.0231425091352009,
+      "grad_norm": 0.16445040702819824,
+      "learning_rate": 0.0002,
+      "loss": 0.9768,
+      "step": 70
+    },
+    {
+      "epoch": 1.0231425091352009,
+      "eval_loss": 0.9452133178710938,
+      "eval_runtime": 294.8787,
+      "eval_samples_per_second": 5.565,
+      "eval_steps_per_second": 0.699,
+      "step": 70
+    },
+    {
+      "epoch": 1.169305724725944,
+      "grad_norm": 0.16264739632606506,
+      "learning_rate": 0.0002,
+      "loss": 0.9673,
+      "step": 80
+    },
+    {
+      "epoch": 1.169305724725944,
+      "eval_loss": 0.9263980984687805,
+      "eval_runtime": 294.9078,
+      "eval_samples_per_second": 5.564,
+      "eval_steps_per_second": 0.699,
+      "step": 80
+    },
+    {
+      "epoch": 1.315468940316687,
+      "grad_norm": 0.1224495992064476,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 90
+    },
+    {
+      "epoch": 1.315468940316687,
+      "eval_loss": 0.9281975626945496,
+      "eval_runtime": 294.9471,
+      "eval_samples_per_second": 5.564,
+      "eval_steps_per_second": 0.698,
+      "step": 90
+    },
+    {
+      "epoch": 1.46163215590743,
+      "grad_norm": 0.20104917883872986,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 100
+    },
+    {
+      "epoch": 1.46163215590743,
+      "eval_loss": 0.918040931224823,
+      "eval_runtime": 294.8931,
+      "eval_samples_per_second": 5.565,
+      "eval_steps_per_second": 0.699,
+      "step": 100
+    },
+    {
+      "epoch": 1.607795371498173,
+      "grad_norm": 0.1838410645723343,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 110
+    },
+    {
+      "epoch": 1.607795371498173,
+      "eval_loss": 0.9211308360099792,
+      "eval_runtime": 294.9367,
+      "eval_samples_per_second": 5.564,
+      "eval_steps_per_second": 0.698,
+      "step": 110
+    },
+    {
+      "epoch": 1.753958587088916,
+      "grad_norm": 0.2129116952419281,
+      "learning_rate": 0.0002,
+      "loss": 0.9317,
+      "step": 120
+    },
+    {
+      "epoch": 1.753958587088916,
+      "eval_loss": 0.9142090678215027,
+      "eval_runtime": 294.8478,
+      "eval_samples_per_second": 5.566,
+      "eval_steps_per_second": 0.699,
+      "step": 120
+    },
+    {
+      "epoch": 1.900121802679659,
+      "grad_norm": 0.15467825531959534,
+      "learning_rate": 0.0002,
+      "loss": 0.9498,
+      "step": 130
+    },
+    {
+      "epoch": 1.900121802679659,
+      "eval_loss": 0.901090681552887,
+      "eval_runtime": 294.95,
+      "eval_samples_per_second": 5.564,
+      "eval_steps_per_second": 0.698,
+      "step": 130
+    },
+    {
+      "epoch": 2.0462850182704018,
+      "grad_norm": 0.13313396275043488,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 140
+    },
+    {
+      "epoch": 2.0462850182704018,
+      "eval_loss": 0.9035805463790894,
+      "eval_runtime": 294.9282,
+      "eval_samples_per_second": 5.564,
+      "eval_steps_per_second": 0.698,
+      "step": 140
+    },
+    {
+      "epoch": 2.192448233861145,
+      "grad_norm": 0.18854761123657227,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 150
+    },
+    {
+      "epoch": 2.192448233861145,
+      "eval_loss": 0.9030548334121704,
+      "eval_runtime": 294.8581,
+      "eval_samples_per_second": 5.565,
+      "eval_steps_per_second": 0.699,
+      "step": 150
+    },
+    {
+      "epoch": 2.338611449451888,
+      "grad_norm": 0.14641670882701874,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 160
+    },
+    {
+      "epoch": 2.338611449451888,
+      "eval_loss": 0.8989503383636475,
+      "eval_runtime": 294.8764,
+      "eval_samples_per_second": 5.565,
+      "eval_steps_per_second": 0.699,
+      "step": 160
+    },
+    {
+      "epoch": 2.484774665042631,
+      "grad_norm": 0.24440337717533112,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 170
+    },
+    {
+      "epoch": 2.484774665042631,
+      "eval_loss": 0.9033491611480713,
+      "eval_runtime": 294.8116,
+      "eval_samples_per_second": 5.566,
+      "eval_steps_per_second": 0.699,
+      "step": 170
+    },
+    {
+      "epoch": 2.630937880633374,
+      "grad_norm": 0.19349223375320435,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 180
+    },
+    {
+      "epoch": 2.630937880633374,
+      "eval_loss": 0.8958781361579895,
+      "eval_runtime": 294.9602,
+      "eval_samples_per_second": 5.563,
+      "eval_steps_per_second": 0.698,
+      "step": 180
+    },
+    {
+      "epoch": 2.630937880633374,
+      "step": 180,
+      "total_flos": 3.658397647269151e+17,
+      "train_loss": 0.9498361190160115,
+      "train_runtime": 17857.7741,
+      "train_samples_per_second": 5.145,
+      "train_steps_per_second": 0.027
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 476,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 7,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.01
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 5
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.658397647269151e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}