Upload XGLMForCausalLM
Browse files- config.json +27 -0
- generation_config.json +8 -0
- pytorch_model.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/xglm-564M",
|
| 3 |
+
"activation_dropout": 0,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"XGLMForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"attention_heads": 16,
|
| 10 |
+
"bos_token_id": 0,
|
| 11 |
+
"d_model": 1024,
|
| 12 |
+
"decoder_start_token_id": 2,
|
| 13 |
+
"dropout": 0.1,
|
| 14 |
+
"eos_token_id": 2,
|
| 15 |
+
"ffn_dim": 4096,
|
| 16 |
+
"init_std": 0.02,
|
| 17 |
+
"layerdrop": 0.0,
|
| 18 |
+
"max_position_embeddings": 2048,
|
| 19 |
+
"model_type": "xglm",
|
| 20 |
+
"num_layers": 24,
|
| 21 |
+
"pad_token_id": 1,
|
| 22 |
+
"scale_embedding": true,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.26.0",
|
| 25 |
+
"use_cache": true,
|
| 26 |
+
"vocab_size": 256008
|
| 27 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 2,
|
| 5 |
+
"eos_token_id": 2,
|
| 6 |
+
"pad_token_id": 1,
|
| 7 |
+
"transformers_version": "4.26.0"
|
| 8 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f85be2690ee7ca194d77ff231e25b0300987aa79b5ee0ecfbbdc7dd2b9e4e3bb
|
| 3 |
+
size 2257982617
|