End of training

README.md

@@ -4,6 +4,7 @@ license: apache-2.0
 base_model: AmberYifan/qwen3-4b-thinking-full-pretrain-mix-low-tweet-1m-en
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: qwen3-4b-thinking-full-pretrain-mix-low-tweet-1m-en-sft
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # qwen3-4b-thinking-full-pretrain-mix-low-tweet-1m-en-sft
 
-This model is a fine-tuned version of [AmberYifan/qwen3-4b-thinking-full-pretrain-mix-low-tweet-1m-en](https://huggingface.co/AmberYifan/qwen3-4b-thinking-full-pretrain-mix-low-tweet-1m-en) on an unknown dataset.
+This model is a fine-tuned version of [AmberYifan/qwen3-4b-thinking-full-pretrain-mix-low-tweet-1m-en](https://huggingface.co/AmberYifan/qwen3-4b-thinking-full-pretrain-mix-low-tweet-1m-en) on the alpaca_en dataset.
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 3713612709888.0,
+    "train_loss": 1.020559032758077,
+    "train_runtime": 200.6016,
+    "train_samples_per_second": 74.775,
+    "train_steps_per_second": 0.598
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 3713612709888.0,
+    "train_loss": 1.020559032758077,
+    "train_runtime": 200.6016,
+    "train_samples_per_second": 74.775,
+    "train_steps_per_second": 0.598
+}

trainer_state.json

@@ -0,0 +1,127 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 120,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.25316455696202533,
+      "grad_norm": 11.874071671786021,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.9812,
+      "step": 10
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.980111719643074,
+      "learning_rate": 9.896703108827758e-06,
+      "loss": 1.353,
+      "step": 20
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 1.8905454650238032,
+      "learning_rate": 9.401006955900555e-06,
+      "loss": 1.2479,
+      "step": 30
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8729820561469692,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 1.2144,
+      "step": 40
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 1.7342992975506955,
+      "learning_rate": 7.373001848738203e-06,
+      "loss": 0.9445,
+      "step": 50
+    },
+    {
+      "epoch": 1.5063291139240507,
+      "grad_norm": 1.8683455108997369,
+      "learning_rate": 6.011087861660191e-06,
+      "loss": 0.9147,
+      "step": 60
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 1.9494176703942474,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.9182,
+      "step": 70
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.7049715031754444,
+      "learning_rate": 3.1539692634365788e-06,
+      "loss": 0.9088,
+      "step": 80
+    },
+    {
+      "epoch": 2.2531645569620253,
+      "grad_norm": 1.7264154986690055,
+      "learning_rate": 1.8988225436587005e-06,
+      "loss": 0.7228,
+      "step": 90
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 1.6933940779071752,
+      "learning_rate": 9.042397785550405e-07,
+      "loss": 0.6609,
+      "step": 100
+    },
+    {
+      "epoch": 2.759493670886076,
+      "grad_norm": 1.7202795598604845,
+      "learning_rate": 2.5378678213483057e-07,
+      "loss": 0.7076,
+      "step": 110
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.7296483993190936,
+      "learning_rate": 2.1152495889970035e-09,
+      "loss": 0.6727,
+      "step": 120
+    },
+    {
+      "epoch": 3.0,
+      "step": 120,
+      "total_flos": 3713612709888.0,
+      "train_loss": 1.020559032758077,
+      "train_runtime": 200.6016,
+      "train_samples_per_second": 74.775,
+      "train_steps_per_second": 0.598
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 120,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3713612709888.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}