Update README.md
Browse files
README.md
CHANGED
|
@@ -16,27 +16,4 @@ This is a custom model for text generation.
|
|
| 16 |
|
| 17 |
## Model Details
|
| 18 |
|
| 19 |
-
- `model_type`:
|
| 20 |
-
|
| 21 |
-
## GPT2
|
| 22 |
-
|
| 23 |
-
This model is **NOT A FINETUNE!!**. It uses the GPT2 architecture but it doesnt finetune it.
|
| 24 |
-
|
| 25 |
-
```python
|
| 26 |
-
# Model configuration for a smaller GPT-2 style model
|
| 27 |
-
config = GPT2Config(
|
| 28 |
-
vocab_size=50257, # Standard GPT-2 vocabulary size
|
| 29 |
-
n_positions=512, # Maximum sequence length
|
| 30 |
-
n_ctx=512, # Context window size
|
| 31 |
-
n_embd=512, # Embedding dimension
|
| 32 |
-
n_layer=6, # Number of transformer layers
|
| 33 |
-
n_head=8, # Number of attention heads
|
| 34 |
-
bos_token_id=50256,
|
| 35 |
-
eos_token_id=50256,
|
| 36 |
-
pad_token_id=50256,
|
| 37 |
-
_name_or_path="" # Empty to ensure no pretrained weights are loaded
|
| 38 |
-
)
|
| 39 |
-
|
| 40 |
-
# Initialize model with random weights
|
| 41 |
-
model = GPT2LMHeadModel(config)
|
| 42 |
-
```
|
|
|
|
| 16 |
|
| 17 |
## Model Details
|
| 18 |
|
| 19 |
+
- `model_type`: custom_transformer
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|