Model save

README.md

@@ -2,15 +2,11 @@
 license: mit
 base_model: gpt2
 tags:
-- alignment-handbook
-- trl
-- sft
-- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- shahadalkhalifa/Crypto_Whitepaper_Labeled
+- generator
 model-index:
 - name: gpt2-cpt-white
   results: []
@@ -21,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # gpt2-cpt-white
 
-This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the shahadalkhalifa/Crypto_Whitepaper_Labeled dataset.
+This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the generator dataset.
 
 ## Model description
 
@@ -45,9 +41,9 @@ The following hyperparameters were used during training:
 - eval_batch_size: 8
 - seed: 42
 - distributed_type: multi-GPU
-- num_devices: 2
-- total_train_batch_size: 32
-- total_eval_batch_size: 16
+- num_devices: 4
+- total_train_batch_size: 64
+- total_eval_batch_size: 32
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
@@ -60,6 +56,6 @@ The following hyperparameters were used during training:
 ### Framework versions
 
 - Transformers 4.39.0.dev0
-- Pytorch 2.2.2+cu121
+- Pytorch 2.3.1+cu121
 - Datasets 2.14.6
 - Tokenizers 0.15.2

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
-    "train_loss": 5.970038026571274,
-    "train_runtime": 52.1808,
+    "train_loss": 6.083620756864548,
+    "train_runtime": 28.044,
     "train_samples": 91,
-    "train_samples_per_second": 19.241,
-    "train_steps_per_second": 0.613
+    "train_samples_per_second": 35.801,
+    "train_steps_per_second": 0.571
 }

config.json

@@ -34,6 +34,6 @@
   },
   "torch_dtype": "bfloat16",
   "transformers_version": "4.39.0.dev0",
-  "use_cache": true,
+  "use_cache": false,
   "vocab_size": 50257
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:730443de84302ebc475f18fd8d665c46b3a6b824b37d233143186fc579c891b7
+oid sha256:fc60df8d071f4d74360fa1e0475e5ce996bd669ad8ba1313e5d9ca9385cf2c88
 size 248894656

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
-    "train_loss": 5.970038026571274,
-    "train_runtime": 52.1808,
+    "train_loss": 6.083620756864548,
+    "train_runtime": 28.044,
     "train_samples": 91,
-    "train_samples_per_second": 19.241,
-    "train_steps_per_second": 0.613
+    "train_samples_per_second": 35.801,
+    "train_steps_per_second": 0.571
 }

trainer_state.json

@@ -3,72 +3,51 @@
   "best_model_checkpoint": null,
   "epoch": 4.0,
   "eval_steps": 500,
-  "global_step": 32,
+  "global_step": 16,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.12,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 6.4878,
+      "epoch": 0.25,
+      "grad_norm": 6.53125,
+      "learning_rate": 0.0001,
+      "loss": 6.4013,
       "step": 1
     },
-    {
-      "epoch": 0.62,
-      "grad_norm": 5.0,
-      "learning_rate": 0.00019937122098932428,
-      "loss": 6.2983,
-      "step": 5
-    },
     {
       "epoch": 1.25,
-      "grad_norm": 3.078125,
+      "grad_norm": 3.84375,
       "learning_rate": 0.000178183148246803,
-      "loss": 6.1058,
-      "step": 10
-    },
-    {
-      "epoch": 1.88,
-      "grad_norm": 2.71875,
-      "learning_rate": 0.00013302790619551674,
-      "loss": 5.931,
-      "step": 15
+      "loss": 6.2911,
+      "step": 5
     },
     {
       "epoch": 2.5,
-      "grad_norm": 2.609375,
+      "grad_norm": 2.765625,
       "learning_rate": 7.774790660436858e-05,
-      "loss": 5.9051,
-      "step": 20
-    },
-    {
-      "epoch": 3.12,
-      "grad_norm": 2.4375,
-      "learning_rate": 2.9289321881345254e-05,
-      "loss": 5.806,
-      "step": 25
+      "loss": 6.0613,
+      "step": 10
     },
     {
       "epoch": 3.75,
-      "grad_norm": 2.390625,
+      "grad_norm": 2.5625,
       "learning_rate": 2.5072087818176382e-06,
-      "loss": 5.859,
-      "step": 30
+      "loss": 5.9747,
+      "step": 15
     },
     {
       "epoch": 4.0,
-      "step": 32,
+      "step": 16,
       "total_flos": 535126081536000.0,
-      "train_loss": 5.970038026571274,
-      "train_runtime": 52.1808,
-      "train_samples_per_second": 19.241,
-      "train_steps_per_second": 0.613
+      "train_loss": 6.083620756864548,
+      "train_runtime": 28.044,
+      "train_samples_per_second": 35.801,
+      "train_steps_per_second": 0.571
     }
   ],
   "logging_steps": 5,
-  "max_steps": 32,
+  "max_steps": 16,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 100,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6a89b73d58e0158c0fea2ad4e8756257b4b11fe853a873525284f5dfda4b1694
+oid sha256:7ba7aad08f79b4b82f8d2e7069e009b925897569aebe186ceb0472cbbb86765b
 size 4984