Upload model
Browse files- config.json +4 -7
- model-00001-of-00003.safetensors +3 -0
- model-00002-of-00003.safetensors +3 -0
- model-00003-of-00003.safetensors +3 -0
- model.safetensors.index.json +42 -0
config.json
CHANGED
|
@@ -2,15 +2,12 @@
|
|
| 2 |
"architectures": [
|
| 3 |
"CustomModel"
|
| 4 |
],
|
| 5 |
-
"auto_map": {
|
| 6 |
-
"AutoConfig": "configuration.CustomConfig",
|
| 7 |
-
"AutoModelForCausalLM": "modeling.CustomModel",
|
| 8 |
-
"AutoModel": "modeling.CustomModel"
|
| 9 |
-
},
|
| 10 |
"d_head": 128,
|
| 11 |
"d_model": 8192,
|
|
|
|
|
|
|
| 12 |
"n_heads": 64,
|
| 13 |
"n_layers": 3,
|
| 14 |
"n_vocab": 50257,
|
| 15 |
-
"
|
| 16 |
-
}
|
|
|
|
| 2 |
"architectures": [
|
| 3 |
"CustomModel"
|
| 4 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"d_head": 128,
|
| 6 |
"d_model": 8192,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"model_type": "custom-mbz-test",
|
| 9 |
"n_heads": 64,
|
| 10 |
"n_layers": 3,
|
| 11 |
"n_vocab": 50257,
|
| 12 |
+
"transformers_version": "4.56.0"
|
| 13 |
+
}
|
model-00001-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:507b764f21cc315722ccfdeeacb2929690d18d81bf6a8c3f5ff05474279d9d2d
|
| 3 |
+
size 4868277384
|
model-00002-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b087dc7a7f118e97a7e2d69029ce6ca2af2e5c3f66887841295f6988e6007697
|
| 3 |
+
size 4295198168
|
model-00003-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2189cfc700dfe1aa6e616001b5cd06882c664c179328a138e487c49b8dce8cdc
|
| 3 |
+
size 3794769116
|
model.safetensors.index.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_parameters": 3239560273,
|
| 4 |
+
"total_size": 12958241092
|
| 5 |
+
},
|
| 6 |
+
"weight_map": {
|
| 7 |
+
"model.blocks.0.attn.Wk.weight": "model-00001-of-00003.safetensors",
|
| 8 |
+
"model.blocks.0.attn.Wo.weight": "model-00001-of-00003.safetensors",
|
| 9 |
+
"model.blocks.0.attn.Wq.weight": "model-00001-of-00003.safetensors",
|
| 10 |
+
"model.blocks.0.attn.Wv.weight": "model-00001-of-00003.safetensors",
|
| 11 |
+
"model.blocks.0.mlp.0.bias": "model-00001-of-00003.safetensors",
|
| 12 |
+
"model.blocks.0.mlp.0.weight": "model-00001-of-00003.safetensors",
|
| 13 |
+
"model.blocks.0.mlp.2.bias": "model-00001-of-00003.safetensors",
|
| 14 |
+
"model.blocks.0.mlp.2.weight": "model-00001-of-00003.safetensors",
|
| 15 |
+
"model.blocks.0.norm1.weight": "model-00001-of-00003.safetensors",
|
| 16 |
+
"model.blocks.0.norm2.weight": "model-00001-of-00003.safetensors",
|
| 17 |
+
"model.blocks.1.attn.Wk.weight": "model-00002-of-00003.safetensors",
|
| 18 |
+
"model.blocks.1.attn.Wo.weight": "model-00002-of-00003.safetensors",
|
| 19 |
+
"model.blocks.1.attn.Wq.weight": "model-00002-of-00003.safetensors",
|
| 20 |
+
"model.blocks.1.attn.Wv.weight": "model-00002-of-00003.safetensors",
|
| 21 |
+
"model.blocks.1.mlp.0.bias": "model-00002-of-00003.safetensors",
|
| 22 |
+
"model.blocks.1.mlp.0.weight": "model-00002-of-00003.safetensors",
|
| 23 |
+
"model.blocks.1.mlp.2.bias": "model-00002-of-00003.safetensors",
|
| 24 |
+
"model.blocks.1.mlp.2.weight": "model-00002-of-00003.safetensors",
|
| 25 |
+
"model.blocks.1.norm1.weight": "model-00002-of-00003.safetensors",
|
| 26 |
+
"model.blocks.1.norm2.weight": "model-00002-of-00003.safetensors",
|
| 27 |
+
"model.blocks.2.attn.Wk.weight": "model-00002-of-00003.safetensors",
|
| 28 |
+
"model.blocks.2.attn.Wo.weight": "model-00002-of-00003.safetensors",
|
| 29 |
+
"model.blocks.2.attn.Wq.weight": "model-00002-of-00003.safetensors",
|
| 30 |
+
"model.blocks.2.attn.Wv.weight": "model-00002-of-00003.safetensors",
|
| 31 |
+
"model.blocks.2.mlp.0.bias": "model-00003-of-00003.safetensors",
|
| 32 |
+
"model.blocks.2.mlp.0.weight": "model-00003-of-00003.safetensors",
|
| 33 |
+
"model.blocks.2.mlp.2.bias": "model-00003-of-00003.safetensors",
|
| 34 |
+
"model.blocks.2.mlp.2.weight": "model-00003-of-00003.safetensors",
|
| 35 |
+
"model.blocks.2.norm1.weight": "model-00003-of-00003.safetensors",
|
| 36 |
+
"model.blocks.2.norm2.weight": "model-00003-of-00003.safetensors",
|
| 37 |
+
"model.embed.weight": "model-00001-of-00003.safetensors",
|
| 38 |
+
"model.norm.weight": "model-00003-of-00003.safetensors",
|
| 39 |
+
"model.out_head.bias": "model-00003-of-00003.safetensors",
|
| 40 |
+
"model.out_head.weight": "model-00003-of-00003.safetensors"
|
| 41 |
+
}
|
| 42 |
+
}
|