End of training

README.md

@@ -1,17 +1,21 @@
 ---
 base_model: HuggingFaceTB/SmolLM2-135M-Instruct
+datasets: Crimsoin/OTC_Medicine_PH_v2
 library_name: transformers
-model_name: HFTB-SmolLM2-135M-Instruct-OTCMedicinePHv2
+model_name: HuggingFaceTB/SmolLM2-135M-Instruct
 tags:
 - generated_from_trainer
+- question-answering
+- QA
+- text-generation
 - trl
 - sft
 licence: license
 ---
 
-# Model Card for HFTB-SmolLM2-135M-Instruct-OTCMedicinePHv2
+# Model Card for HuggingFaceTB/SmolLM2-135M-Instruct
 
-This model is a fine-tuned version of [HuggingFaceTB/SmolLM2-135M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM2-135M-Instruct).
+This model is a fine-tuned version of [HuggingFaceTB/SmolLM2-135M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM2-135M-Instruct) on the [Crimsoin/OTC_Medicine_PH_v2](https://huggingface.co/datasets/Crimsoin/OTC_Medicine_PH_v2) dataset.
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 3.883495145631068,
-    "eval_loss": 0.14737741649150848,
-    "eval_runtime": 14.0975,
-    "eval_samples_per_second": 37.595,
-    "eval_steps_per_second": 4.753,
-    "total_flos": 258162252195456.0,
-    "train_loss": 0.4854153060913086,
-    "train_runtime": 885.215,
-    "train_samples_per_second": 5.422,
-    "train_steps_per_second": 0.113
+    "epoch": 9.70873786407767,
+    "eval_loss": 0.11601576954126358,
+    "eval_runtime": 14.5695,
+    "eval_samples_per_second": 36.377,
+    "eval_steps_per_second": 4.599,
+    "total_flos": 634548176930304.0,
+    "train_loss": 0.2663491096496582,
+    "train_runtime": 2216.0805,
+    "train_samples_per_second": 21.66,
+    "train_steps_per_second": 0.451
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 3.883495145631068,
-    "eval_loss": 0.14737741649150848,
-    "eval_runtime": 14.0975,
-    "eval_samples_per_second": 37.595,
-    "eval_steps_per_second": 4.753
+    "epoch": 9.70873786407767,
+    "eval_loss": 0.11601576954126358,
+    "eval_runtime": 14.5695,
+    "eval_samples_per_second": 36.377,
+    "eval_steps_per_second": 4.599
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.883495145631068,
-    "total_flos": 258162252195456.0,
-    "train_loss": 0.4854153060913086,
-    "train_runtime": 885.215,
-    "train_samples_per_second": 5.422,
-    "train_steps_per_second": 0.113
+    "epoch": 9.70873786407767,
+    "total_flos": 634548176930304.0,
+    "train_loss": 0.2663491096496582,
+    "train_runtime": 2216.0805,
+    "train_samples_per_second": 21.66,
+    "train_steps_per_second": 0.451
 }

trainer_state.json

@@ -1,43 +1,103 @@
 {
-  "best_metric": 0.14742426574230194,
-  "best_model_checkpoint": "./HFTB-SmolLM2-135M-Instruct-OTCMedicinePHv2\\checkpoint-100",
-  "epoch": 3.883495145631068,
-  "eval_steps": 100,
-  "global_step": 100,
+  "best_metric": 0.11598550528287888,
+  "best_model_checkpoint": "./HFTB-SmolLM2-135M-Instruct-OTCMedicinePHv2\\checkpoint-250",
+  "epoch": 9.70873786407767,
+  "eval_steps": 50,
+  "global_step": 250,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.35462793707847595,
+      "learning_rate": 0.0005,
+      "loss": 0.8189,
+      "step": 50
+    },
+    {
+      "epoch": 1.941747572815534,
+      "eval_loss": 0.16038869321346283,
+      "eval_runtime": 13.1355,
+      "eval_samples_per_second": 40.349,
+      "eval_steps_per_second": 5.101,
+      "step": 50
+    },
     {
       "epoch": 3.883495145631068,
-      "grad_norm": 0.20863692462444305,
+      "grad_norm": 0.2993379235267639,
       "learning_rate": 0.0005,
-      "loss": 0.4854,
+      "loss": 0.1464,
       "step": 100
     },
     {
       "epoch": 3.883495145631068,
-      "eval_loss": 0.14742426574230194,
-      "eval_runtime": 13.1852,
-      "eval_samples_per_second": 40.197,
-      "eval_steps_per_second": 5.081,
+      "eval_loss": 0.12895184755325317,
+      "eval_runtime": 12.8791,
+      "eval_samples_per_second": 41.152,
+      "eval_steps_per_second": 5.202,
       "step": 100
     },
     {
-      "epoch": 3.883495145631068,
-      "step": 100,
-      "total_flos": 258162252195456.0,
-      "train_loss": 0.4854153060913086,
-      "train_runtime": 885.215,
-      "train_samples_per_second": 5.422,
-      "train_steps_per_second": 0.113
+      "epoch": 5.825242718446602,
+      "grad_norm": 0.14636453986167908,
+      "learning_rate": 0.0005,
+      "loss": 0.127,
+      "step": 150
+    },
+    {
+      "epoch": 5.825242718446602,
+      "eval_loss": 0.12111850827932358,
+      "eval_runtime": 12.7129,
+      "eval_samples_per_second": 41.69,
+      "eval_steps_per_second": 5.27,
+      "step": 150
+    },
+    {
+      "epoch": 7.766990291262136,
+      "grad_norm": 0.1845141053199768,
+      "learning_rate": 0.0005,
+      "loss": 0.1212,
+      "step": 200
+    },
+    {
+      "epoch": 7.766990291262136,
+      "eval_loss": 0.11958930641412735,
+      "eval_runtime": 14.1475,
+      "eval_samples_per_second": 37.463,
+      "eval_steps_per_second": 4.736,
+      "step": 200
+    },
+    {
+      "epoch": 9.70873786407767,
+      "grad_norm": 0.14870111644268036,
+      "learning_rate": 0.0005,
+      "loss": 0.1183,
+      "step": 250
+    },
+    {
+      "epoch": 9.70873786407767,
+      "eval_loss": 0.11598550528287888,
+      "eval_runtime": 13.0644,
+      "eval_samples_per_second": 40.568,
+      "eval_steps_per_second": 5.128,
+      "step": 250
+    },
+    {
+      "epoch": 9.70873786407767,
+      "step": 250,
+      "total_flos": 634548176930304.0,
+      "train_loss": 0.2663491096496582,
+      "train_runtime": 2216.0805,
+      "train_samples_per_second": 21.66,
+      "train_steps_per_second": 0.451
     }
   ],
-  "logging_steps": 100,
-  "max_steps": 100,
+  "logging_steps": 50,
+  "max_steps": 1000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
-  "save_steps": 100,
+  "num_train_epochs": 40,
+  "save_steps": 50,
   "stateful_callbacks": {
     "EarlyStoppingCallback": {
       "args": {
@@ -45,7 +105,7 @@
         "early_stopping_threshold": 0.01
       },
       "attributes": {
-        "early_stopping_patience_counter": 0
+        "early_stopping_patience_counter": 3
       }
     },
     "TrainerControl": {
@@ -59,7 +119,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 258162252195456.0,
+  "total_flos": 634548176930304.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null