Training in progress epoch 0

README.md

@@ -1,5 +1,4 @@
 ---
-license: apache-2.0
 tags:
 - generated_from_keras_callback
 model-index:
@@ -12,11 +11,11 @@ probably proofread and complete it, then remove this comment. -->
 
 # fab-an/my_lang-model
 
-This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on an unknown dataset.
+This model was trained from scratch on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Train Loss: 6.9749
-- Validation Loss: 6.8068
-- Epoch: 2
+- Train Loss: 6.9373
+- Validation Loss: 6.5509
+- Epoch: 0
 
 ## Model description
 
@@ -42,9 +41,7 @@ The following hyperparameters were used during training:
 
 | Train Loss | Validation Loss | Epoch |
 |:----------:|:---------------:|:-----:|
-| 7.3639     | 7.0367          | 0     |
-| 7.1125     | 6.9146          | 1     |
-| 6.9749     | 6.8068          | 2     |
+| 6.9373     | 6.5509          | 0     |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "distilgpt2",
+  "_name_or_path": "my_lang-model",
   "_num_labels": 1,
   "activation_function": "gelu_new",
   "architectures": [

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:405a7ea0effd9fca82e4fdb0be0b17e9cb534c8d84f7509b8d3a9410c70647f4
+oid sha256:559c2ca95d41946348d0476e08815da71b9ed2437786f3fc1b36fefce07bbe49
 size 327745496