Fix config.json to match original gpt-s2.5-5m exactly
#3
by Compactbot - opened
- config.json +23 -0
config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"GPTS25ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"model_type": "gpt-s2.5",
|
| 6 |
+
"vocab_size": 8192,
|
| 7 |
+
"hidden_size": 256,
|
| 8 |
+
"num_hidden_layers": 4,
|
| 9 |
+
"num_attention_heads": 8,
|
| 10 |
+
"num_key_value_heads": 2,
|
| 11 |
+
"head_dim": 32,
|
| 12 |
+
"intermediate_size": 768,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"rope_theta": 10000.0,
|
| 15 |
+
"norm_type": "rmsnorm",
|
| 16 |
+
"norm_eps": 1e-06,
|
| 17 |
+
"mlp_activation": "swiglu",
|
| 18 |
+
"tie_word_embeddings": true,
|
| 19 |
+
"bias": false,
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.50.0",
|
| 22 |
+
"parameter_count": 5114112
|
| 23 |
+
}
|