Upload model

README.md

@@ -1,23 +1,23 @@
 ---
 license: mit
-base_model: gpt2
+base_model: gpt2-large
 tags:
 - generated_from_keras_callback
 model-index:
-- name: TahaCakir/enhanced_turkishReviews-generativeAI
+- name: enhanced_turkishReviews-generativeAI
   results: []
 ---
 
 <!-- This model card has been generated automatically according to the information Keras had access to. You should
 probably proofread and complete it, then remove this comment. -->
 
-# TahaCakir/enhanced_turkishReviews-generativeAI
+# enhanced_turkishReviews-generativeAI
 
-This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset.
+This model is a fine-tuned version of [gpt2-large](https://huggingface.co/gpt2-large) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Train Loss: 9.7787
-- Validation Loss: 9.7619
-- Epoch: 0
+- Train Loss: 9.7770
+- Validation Loss: 9.7482
+- Epoch: 9
 
 ## Model description
 
@@ -36,14 +36,23 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- optimizer: {'inner_optimizer': {'module': 'transformers.optimization_tf', 'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': {'module': 'transformers.optimization_tf', 'class_name': 'WarmUp', 'config': {'initial_learning_rate': 5e-07, 'decay_schedule_fn': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 5e-07, 'decay_steps': -293, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'warmup_steps': 1000, 'power': 1.0, 'name': None}, 'registered_name': 'WarmUp'}, 'decay': 0.0, 'beta_1': 0.8999999761581421, 'beta_2': 0.9990000128746033, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.001}, 'registered_name': 'AdamWeightDecay'}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
+- optimizer: {'inner_optimizer': {'module': 'transformers.optimization_tf', 'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': {'module': 'transformers.optimization_tf', 'class_name': 'WarmUp', 'config': {'initial_learning_rate': 5e-07, 'decay_schedule_fn': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 5e-07, 'decay_steps': -293, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'warmup_steps': 1000, 'power': 1.0, 'name': None}, 'registered_name': 'WarmUp'}, 'decay': 0.0, 'beta_1': 0.8999999761581421, 'beta_2': 0.9990000128746033, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.09}, 'registered_name': 'AdamWeightDecay'}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
 - training_precision: mixed_float16
 
 ### Training results
 
 | Train Loss | Validation Loss | Epoch |
 |:----------:|:---------------:|:-----:|
-| 9.7787     | 9.7619          | 0     |
+| 10.1112    | 9.8654          | 0     |
+| 9.8034     | 9.7482          | 1     |
+| 9.7773     | 9.7482          | 2     |
+| 9.7774     | 9.7482          | 3     |
+| 9.7772     | 9.7482          | 4     |
+| 9.7770     | 9.7482          | 5     |
+| 9.7770     | 9.7482          | 6     |
+| 9.7774     | 9.7482          | 7     |
+| 9.7776     | 9.7482          | 8     |
+| 9.7770     | 9.7482          | 9     |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "gpt2",
+  "_name_or_path": "gpt2-large",
   "activation_function": "gelu_new",
   "architectures": [
     "GPT2LMHeadModel"
@@ -13,12 +13,12 @@
   "model_type": "gpt2",
   "n_ctx": 40,
   "n_embd": 768,
-  "n_head": 12,
+  "n_head": 16,
   "n_inner": null,
-  "n_layer": 12,
+  "n_layer": 16,
   "n_positions": 1024,
   "reorder_and_upcast_attn": false,
-  "resid_pdrop": 0.1,
+  "resid_pdrop": 0.2,
   "scale_attn_by_inverse_layer_idx": false,
   "scale_attn_weights": true,
   "summary_activation": null,

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:68964dea7356cc2f2ca51766f3b62e1c40f4d97cab98d3c8c13e5ed56fb1a9e6
-size 420345936
+oid sha256:55561abda80ee7e21db5f23c2f6bfe0a8a0399d5803a3bc91cdb95bff17532f0
+size 533811512