Upload config.json with huggingface_hub
Browse files- config.json +35 -0
config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/mnt/hdfs/ridger_test/ridger/fla-hybrid-research/configs/ptransformers/ptransformer_340M_gmlp.json",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"PTransformerForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"elementwise_affine": true,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"fuse_cross_entropy": true,
|
| 11 |
+
"fuse_norm": true,
|
| 12 |
+
"fuse_swiglu": true,
|
| 13 |
+
"hidden_act": "swish",
|
| 14 |
+
"hidden_dropout": 0.0,
|
| 15 |
+
"hidden_ratio": 4,
|
| 16 |
+
"hidden_size": 1024,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": null,
|
| 19 |
+
"max_position_embeddings": 4096,
|
| 20 |
+
"model_type": "ptransformer",
|
| 21 |
+
"norm_activation_type": "l2_norm_gelu",
|
| 22 |
+
"norm_eps": 1e-06,
|
| 23 |
+
"num_heads": 16,
|
| 24 |
+
"num_hidden_layers": 24,
|
| 25 |
+
"num_kv_heads": null,
|
| 26 |
+
"param_token_num": 512,
|
| 27 |
+
"qkv_bias": false,
|
| 28 |
+
"rope_theta": 10000.0,
|
| 29 |
+
"tie_word_embeddings": true,
|
| 30 |
+
"torch_dtype": "float32",
|
| 31 |
+
"transformers_version": "4.49.0",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 32000,
|
| 34 |
+
"window_size": null
|
| 35 |
+
}
|