End of training

README.md

@@ -3,11 +3,23 @@ license: other
 base_model: Qwen/Qwen1.5-4B
 tags:
 - generated_from_trainer
+datasets:
+- tyzhu/lmind_hotpot_train8000_eval7405_v1_reciteonly_qa
 metrics:
 - accuracy
 model-index:
 - name: lmind_hotpot_train8000_eval7405_v1_reciteonly_qa_Qwen_Qwen1.5-4B_lora2
-  results: []
+  results:
+  - task:
+      name: Causal Language Modeling
+      type: text-generation
+    dataset:
+      name: tyzhu/lmind_hotpot_train8000_eval7405_v1_reciteonly_qa
+      type: tyzhu/lmind_hotpot_train8000_eval7405_v1_reciteonly_qa
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.6657438136826783
 library_name: peft
 ---
 
@@ -16,7 +28,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # lmind_hotpot_train8000_eval7405_v1_reciteonly_qa_Qwen_Qwen1.5-4B_lora2
 
-This model is a fine-tuned version of [Qwen/Qwen1.5-4B](https://huggingface.co/Qwen/Qwen1.5-4B) on an unknown dataset.
+This model is a fine-tuned version of [Qwen/Qwen1.5-4B](https://huggingface.co/Qwen/Qwen1.5-4B) on the tyzhu/lmind_hotpot_train8000_eval7405_v1_reciteonly_qa dataset.
 It achieves the following results on the evaluation set:
 - Loss: 1.9636
 - Accuracy: 0.6657

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 10.0,
+    "eval_accuracy": 0.6657438136826783,
+    "eval_loss": 1.9635711908340454,
+    "eval_runtime": 11.6834,
+    "eval_samples": 500,
+    "eval_samples_per_second": 42.796,
+    "eval_steps_per_second": 5.392,
+    "perplexity": 7.1247254411303755,
+    "total_flos": 4.6107057930738074e+17,
+    "train_loss": 1.057360610961914,
+    "train_runtime": 8831.7737,
+    "train_samples": 8000,
+    "train_samples_per_second": 9.058,
+    "train_steps_per_second": 0.283
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 10.0,
+    "eval_accuracy": 0.6657438136826783,
+    "eval_loss": 1.9635711908340454,
+    "eval_runtime": 11.6834,
+    "eval_samples": 500,
+    "eval_samples_per_second": 42.796,
+    "eval_steps_per_second": 5.392,
+    "perplexity": 7.1247254411303755
+}

tokenizer.json

@@ -1,11 +1,6 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 1024,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
+  "truncation": null,
   "padding": null,
   "added_tokens": [
     {

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 4.6107057930738074e+17,
+    "train_loss": 1.057360610961914,
+    "train_runtime": 8831.7737,
+    "train_samples": 8000,
+    "train_samples_per_second": 9.058,
+    "train_steps_per_second": 0.283
+}

trainer_state.json

@@ -0,0 +1,295 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15738776326179504,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17586949467658997,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.6758602620087336,
+      "eval_loss": 1.5119266510009766,
+      "eval_runtime": 10.6538,
+      "eval_samples_per_second": 46.932,
+      "eval_steps_per_second": 5.913,
+      "step": 250
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2267732471227646,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.27053600549697876,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.24228893220424652,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6766957787481805,
+      "eval_loss": 1.5021401643753052,
+      "eval_runtime": 10.5752,
+      "eval_samples_per_second": 47.28,
+      "eval_steps_per_second": 5.957,
+      "step": 500
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.32249245047569275,
+      "learning_rate": 0.0001,
+      "loss": 1.333,
+      "step": 600
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.33233705163002014,
+      "learning_rate": 0.0001,
+      "loss": 1.3267,
+      "step": 700
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6766550218340611,
+      "eval_loss": 1.5067318677902222,
+      "eval_runtime": 10.7034,
+      "eval_samples_per_second": 46.714,
+      "eval_steps_per_second": 5.886,
+      "step": 750
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.3930753171443939,
+      "learning_rate": 0.0001,
+      "loss": 1.2809,
+      "step": 800
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.3908335268497467,
+      "learning_rate": 0.0001,
+      "loss": 1.2272,
+      "step": 900
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.4177429676055908,
+      "learning_rate": 0.0001,
+      "loss": 1.2354,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6760407569141194,
+      "eval_loss": 1.5289020538330078,
+      "eval_runtime": 10.763,
+      "eval_samples_per_second": 46.456,
+      "eval_steps_per_second": 5.853,
+      "step": 1000
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.49195700883865356,
+      "learning_rate": 0.0001,
+      "loss": 1.1186,
+      "step": 1100
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.5006465911865234,
+      "learning_rate": 0.0001,
+      "loss": 1.1245,
+      "step": 1200
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6744483260553129,
+      "eval_loss": 1.57330322265625,
+      "eval_runtime": 9.6986,
+      "eval_samples_per_second": 51.554,
+      "eval_steps_per_second": 6.496,
+      "step": 1250
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.5376680493354797,
+      "learning_rate": 0.0001,
+      "loss": 1.0669,
+      "step": 1300
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.5510832667350769,
+      "learning_rate": 0.0001,
+      "loss": 1.016,
+      "step": 1400
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.5245699286460876,
+      "learning_rate": 0.0001,
+      "loss": 1.0235,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6729868995633188,
+      "eval_loss": 1.6228069067001343,
+      "eval_runtime": 11.3386,
+      "eval_samples_per_second": 44.097,
+      "eval_steps_per_second": 5.556,
+      "step": 1500
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.6557231545448303,
+      "learning_rate": 0.0001,
+      "loss": 0.8934,
+      "step": 1600
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.668296217918396,
+      "learning_rate": 0.0001,
+      "loss": 0.9119,
+      "step": 1700
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6708500727802038,
+      "eval_loss": 1.6996169090270996,
+      "eval_runtime": 8.9777,
+      "eval_samples_per_second": 55.693,
+      "eval_steps_per_second": 7.017,
+      "step": 1750
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.7145330905914307,
+      "learning_rate": 0.0001,
+      "loss": 0.8453,
+      "step": 1800
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.785029411315918,
+      "learning_rate": 0.0001,
+      "loss": 0.7869,
+      "step": 1900
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.7093644142150879,
+      "learning_rate": 0.0001,
+      "loss": 0.8037,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6694614264919941,
+      "eval_loss": 1.771801233291626,
+      "eval_runtime": 12.0591,
+      "eval_samples_per_second": 41.462,
+      "eval_steps_per_second": 5.224,
+      "step": 2000
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.8164002895355225,
+      "learning_rate": 0.0001,
+      "loss": 0.6789,
+      "step": 2100
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.8319939374923706,
+      "learning_rate": 0.0001,
+      "loss": 0.6868,
+      "step": 2200
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6676448326055313,
+      "eval_loss": 1.849050521850586,
+      "eval_runtime": 10.8663,
+      "eval_samples_per_second": 46.014,
+      "eval_steps_per_second": 5.798,
+      "step": 2250
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 0.9181467294692993,
+      "learning_rate": 0.0001,
+      "loss": 0.6431,
+      "step": 2300
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 0.8400017023086548,
+      "learning_rate": 0.0001,
+      "loss": 0.5865,
+      "step": 2400
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.9538242816925049,
+      "learning_rate": 0.0001,
+      "loss": 0.6049,
+      "step": 2500
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.6657438136826783,
+      "eval_loss": 1.9635711908340454,
+      "eval_runtime": 9.0651,
+      "eval_samples_per_second": 55.157,
+      "eval_steps_per_second": 6.95,
+      "step": 2500
+    },
+    {
+      "epoch": 10.0,
+      "step": 2500,
+      "total_flos": 4.6107057930738074e+17,
+      "train_loss": 1.057360610961914,
+      "train_runtime": 8831.7737,
+      "train_samples_per_second": 9.058,
+      "train_steps_per_second": 0.283
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 2500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 4.6107057930738074e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}