dzhuz commited on
Commit
52dc7a8
·
verified ·
1 Parent(s): 03f019c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +20 -26
README.md CHANGED
@@ -6,29 +6,28 @@ base_model:
6
  tags:
7
  - art
8
  ---
9
- GPT-2 Fine-Tuned on Art Museum Books
10
- Model Description
11
- This model is a fine-tuned version of GPT-2, adapted to generate text inspired by a self-made dataset of various art museum books. It specializes in creating detailed, context-aware, and stylistically rich descriptions relevant to art and museum contexts.
12
-
13
- Training Details
14
- Base Model: GPT-2
15
- Dataset: Custom-built dataset containing curated texts from art museum books.
16
- Steps: Trained for 29,900 steps over ~2 hours.
17
- Hardware: Google Colab with NVIDIA T4 GPU.
18
- Hyperparameters:
19
- Epochs: 5
20
- Batch Size: 8
21
- Mixed Precision (fp16): Enabled
22
- Save Steps: Every 500 steps
23
- Logging Steps: Every 100 steps
24
- Evaluation Strategy: None
25
-
26
 
27
- Training Script
28
- Used the transformers library with the following configuration:
29
-
30
- python
31
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
32
  from transformers import TrainingArguments, Trainer
33
  from transformers import DataCollatorForLanguageModeling
34
 
@@ -49,8 +48,3 @@ data_collator = DataCollatorForLanguageModeling(
49
  tokenizer=tokenizer,
50
  mlm=False,
51
  )
52
- Intended Use
53
- This model is intended for:
54
-
55
- Generating art-themed text or creative writing.
56
- Assisting with museum or gallery description generation.
 
6
  tags:
7
  - art
8
  ---
9
+ # GPT-2 Fine-Tuned on Art Museum Books
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
10
 
11
+ ## Model Description
12
+ This model is a fine-tuned version of GPT-2, adapted to generate text inspired by a self-made dataset of various art museum books. It specializes in creating detailed, context-aware, and stylistically rich descriptions relevant to art and museum contexts.
 
 
13
 
14
+ ## Training Details
15
+ - **Base Model**: GPT-2
16
+ - **Dataset**: Custom-built dataset containing curated texts from art museum books.
17
+ - **Steps**: Trained for 29,900 steps over ~2 hours.
18
+ - **Hardware**: Google Colab with NVIDIA T4 GPU.
19
+ - **Hyperparameters**:
20
+ - Epochs: 5
21
+ - Batch Size: 8
22
+ - Mixed Precision (fp16): Enabled
23
+ - Save Steps: Every 500 steps
24
+ - Logging Steps: Every 100 steps
25
+ - Evaluation Strategy: None
26
+
27
+ ### Training Script
28
+ The model was fine-tuned using the `transformers` library with the following configuration:
29
+
30
+ ```python
31
  from transformers import TrainingArguments, Trainer
32
  from transformers import DataCollatorForLanguageModeling
33
 
 
48
  tokenizer=tokenizer,
49
  mlm=False,
50
  )