Model save

README.md

@@ -1,11 +1,10 @@
 ---
 base_model: meta-llama/Meta-Llama-3-8B
 datasets:
-- llama-duo/synth_classification_dataset_dedup
+- generator
 library_name: peft
 license: llama3
 tags:
-- alignment-handbook
 - trl
 - sft
 - generated_from_trainer
@@ -19,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama3.1-8b-gpt4o_100k_classification-k
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on the llama-duo/synth_classification_dataset_dedup dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.6323
+- Loss: 1.6382
 
 ## Model description
 
@@ -58,7 +57,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 1.0751        | 0.9958 | 118  | 1.6323          |
+| 1.075         | 0.9958 | 118  | 1.6382          |
 
 
 ### Framework versions

all_results.json

@@ -1,14 +1,9 @@
 {
     "epoch": 0.9957805907172996,
-    "eval_loss": 1.6322884559631348,
-    "eval_runtime": 1.1485,
-    "eval_samples": 16,
-    "eval_samples_per_second": 1.741,
-    "eval_steps_per_second": 0.871,
     "total_flos": 3.48383355932246e+17,
-    "train_loss": 1.2928365105289523,
-    "train_runtime": 1374.4457,
+    "train_loss": 1.291591884726185,
+    "train_runtime": 1283.2771,
     "train_samples": 92634,
-    "train_samples_per_second": 5.507,
-    "train_steps_per_second": 0.086
+    "train_samples_per_second": 5.898,
+    "train_steps_per_second": 0.092
 }

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 0.9957805907172996,
     "total_flos": 3.48383355932246e+17,
-    "train_loss": 1.2928365105289523,
-    "train_runtime": 1374.4457,
+    "train_loss": 1.291591884726185,
+    "train_runtime": 1283.2771,
     "train_samples": 92634,
-    "train_samples_per_second": 5.507,
-    "train_steps_per_second": 0.086
+    "train_samples_per_second": 5.898,
+    "train_steps_per_second": 0.092
 }

trainer_state.json

@@ -10,188 +10,188 @@
   "log_history": [
     {
       "epoch": 0.008438818565400843,
-      "grad_norm": 1.0222958326339722,
+      "grad_norm": 1.022000789642334,
       "learning_rate": 2.4999999999999998e-05,
       "loss": 2.86,
       "step": 1
     },
     {
       "epoch": 0.04219409282700422,
-      "grad_norm": 1.6936215162277222,
+      "grad_norm": 1.4761569499969482,
       "learning_rate": 0.000125,
-      "loss": 2.8152,
+      "loss": 2.8149,
       "step": 5
     },
     {
       "epoch": 0.08438818565400844,
-      "grad_norm": 2.3276407718658447,
+      "grad_norm": 2.290438413619995,
       "learning_rate": 0.00025,
-      "loss": 2.5324,
+      "loss": 2.5283,
       "step": 10
     },
     {
       "epoch": 0.12658227848101267,
-      "grad_norm": 6.024274826049805,
+      "grad_norm": 5.928793430328369,
       "learning_rate": 0.0002994074763922825,
-      "loss": 1.692,
+      "loss": 1.6863,
       "step": 15
     },
     {
       "epoch": 0.16877637130801687,
-      "grad_norm": 0.993145763874054,
+      "grad_norm": 0.9887209534645081,
       "learning_rate": 0.0002958034371120616,
-      "loss": 1.4859,
+      "loss": 1.4771,
       "step": 20
     },
     {
       "epoch": 0.2109704641350211,
-      "grad_norm": 0.7716729640960693,
+      "grad_norm": 0.7580465078353882,
       "learning_rate": 0.000289003441114775,
-      "loss": 1.326,
+      "loss": 1.3231,
       "step": 25
     },
     {
       "epoch": 0.25316455696202533,
-      "grad_norm": 0.5372955203056335,
+      "grad_norm": 0.5388275384902954,
       "learning_rate": 0.0002791565417651033,
-      "loss": 1.2597,
+      "loss": 1.2578,
       "step": 30
     },
     {
       "epoch": 0.29535864978902954,
-      "grad_norm": 0.5266535878181458,
+      "grad_norm": 0.5278171896934509,
       "learning_rate": 0.00026647857940770634,
-      "loss": 1.1995,
+      "loss": 1.1985,
       "step": 35
     },
     {
       "epoch": 0.33755274261603374,
-      "grad_norm": 0.3248014748096466,
+      "grad_norm": 0.31164848804473877,
       "learning_rate": 0.0002512474502277316,
-      "loss": 1.1785,
+      "loss": 1.178,
       "step": 40
     },
     {
       "epoch": 0.379746835443038,
-      "grad_norm": 0.22964197397232056,
+      "grad_norm": 0.22269943356513977,
       "learning_rate": 0.00023379701487054785,
-      "loss": 1.1528,
+      "loss": 1.1526,
       "step": 45
     },
     {
       "epoch": 0.4219409282700422,
-      "grad_norm": 0.2224261313676834,
+      "grad_norm": 0.220732182264328,
       "learning_rate": 0.00021450978034147806,
-      "loss": 1.1343,
+      "loss": 1.134,
       "step": 50
     },
     {
       "epoch": 0.4641350210970464,
-      "grad_norm": 0.22749578952789307,
+      "grad_norm": 0.2316388487815857,
       "learning_rate": 0.00019380851559554636,
-      "loss": 1.1165,
+      "loss": 1.1167,
       "step": 55
     },
     {
       "epoch": 0.5063291139240507,
-      "grad_norm": 0.2278250902891159,
+      "grad_norm": 0.2175298035144806,
       "learning_rate": 0.00017214698460037218,
-      "loss": 1.1106,
+      "loss": 1.1102,
       "step": 60
     },
     {
       "epoch": 0.5485232067510548,
-      "grad_norm": 0.22940242290496826,
+      "grad_norm": 0.21935291588306427,
       "learning_rate": 0.00015,
-      "loss": 1.0946,
+      "loss": 1.0942,
       "step": 65
     },
     {
       "epoch": 0.5907172995780591,
-      "grad_norm": 0.20619849860668182,
+      "grad_norm": 0.2013280689716339,
       "learning_rate": 0.00012785301539962782,
-      "loss": 1.0952,
+      "loss": 1.0947,
       "step": 70
     },
     {
       "epoch": 0.6329113924050633,
-      "grad_norm": 0.19738492369651794,
+      "grad_norm": 0.18674449622631073,
       "learning_rate": 0.00010619148440445364,
-      "loss": 1.0891,
+      "loss": 1.0888,
       "step": 75
     },
     {
       "epoch": 0.6751054852320675,
-      "grad_norm": 0.2077159434556961,
+      "grad_norm": 0.21201685070991516,
       "learning_rate": 8.549021965852197e-05,
-      "loss": 1.0828,
+      "loss": 1.0819,
       "step": 80
     },
     {
       "epoch": 0.7172995780590717,
-      "grad_norm": 0.1939452439546585,
+      "grad_norm": 0.19578513503074646,
       "learning_rate": 6.620298512945214e-05,
-      "loss": 1.0723,
+      "loss": 1.0722,
       "step": 85
     },
     {
       "epoch": 0.759493670886076,
-      "grad_norm": 0.20105819404125214,
+      "grad_norm": 0.19662944972515106,
       "learning_rate": 4.8752549772268444e-05,
-      "loss": 1.0617,
+      "loss": 1.0615,
       "step": 90
     },
     {
       "epoch": 0.8016877637130801,
-      "grad_norm": 0.2106432467699051,
+      "grad_norm": 0.20801547169685364,
       "learning_rate": 3.352142059229365e-05,
-      "loss": 1.0647,
+      "loss": 1.0642,
       "step": 95
     },
     {
       "epoch": 0.8438818565400844,
-      "grad_norm": 0.19021108746528625,
+      "grad_norm": 0.18992270529270172,
       "learning_rate": 2.0843458234896666e-05,
-      "loss": 1.0778,
+      "loss": 1.0775,
       "step": 100
     },
     {
       "epoch": 0.8860759493670886,
-      "grad_norm": 0.23987625539302826,
+      "grad_norm": 0.21922025084495544,
       "learning_rate": 1.0996558885224993e-05,
-      "loss": 1.0761,
+      "loss": 1.0757,
       "step": 105
     },
     {
       "epoch": 0.9282700421940928,
-      "grad_norm": 0.1889180988073349,
+      "grad_norm": 0.18345749378204346,
       "learning_rate": 4.1965628879383875e-06,
-      "loss": 1.0657,
+      "loss": 1.0654,
       "step": 110
     },
     {
       "epoch": 0.9704641350210971,
-      "grad_norm": 0.19509291648864746,
+      "grad_norm": 0.19231846928596497,
       "learning_rate": 5.925236077174655e-07,
-      "loss": 1.0751,
+      "loss": 1.075,
       "step": 115
     },
     {
       "epoch": 0.9957805907172996,
-      "eval_loss": 1.6322884559631348,
-      "eval_runtime": 1.1519,
-      "eval_samples_per_second": 1.736,
-      "eval_steps_per_second": 0.868,
+      "eval_loss": 1.6381601095199585,
+      "eval_runtime": 1.1564,
+      "eval_samples_per_second": 1.729,
+      "eval_steps_per_second": 0.865,
       "step": 118
     },
     {
       "epoch": 0.9957805907172996,
       "step": 118,
       "total_flos": 3.48383355932246e+17,
-      "train_loss": 1.2928365105289523,
-      "train_runtime": 1374.4457,
-      "train_samples_per_second": 5.507,
-      "train_steps_per_second": 0.086
+      "train_loss": 1.291591884726185,
+      "train_runtime": 1283.2771,
+      "train_samples_per_second": 5.898,
+      "train_steps_per_second": 0.092
     }
   ],
   "logging_steps": 5,