fix
Browse files
README.md
CHANGED
@@ -26,14 +26,14 @@ PhiConfig {
|
|
26 |
"tie_word_embeddings": false,
|
27 |
"transformers_version": "4.38.2",
|
28 |
"use_cache": true,
|
29 |
-
"vocab_size":
|
30 |
}
|
31 |
```
|
32 |
|
33 |
```
|
34 |
PhiForCausalLM(
|
35 |
(model): PhiModel(
|
36 |
-
(embed_tokens): Embedding(
|
37 |
(embed_dropout): Dropout(p=0.0, inplace=False)
|
38 |
(layers): ModuleList(
|
39 |
(0-5): 6 x PhiDecoderLayer(
|
@@ -55,9 +55,8 @@ PhiForCausalLM(
|
|
55 |
)
|
56 |
(final_layernorm): LayerNorm((8,), eps=1e-05, elementwise_affine=True)
|
57 |
)
|
58 |
-
(lm_head): Linear(in_features=8, out_features=
|
59 |
)
|
60 |
-
|
61 |
```
|
62 |
|
63 |
```
|
@@ -66,7 +65,7 @@ Layer (type:depth-idx) Param #
|
|
66 |
===========================================================================
|
67 |
PhiForCausalLM --
|
68 |
├─PhiModel: 1-1 --
|
69 |
-
│ └─Embedding: 2-1
|
70 |
│ └─Dropout: 2-2 --
|
71 |
│ └─ModuleList: 2-3 --
|
72 |
│ │ └─PhiDecoderLayer: 3-1 410
|
@@ -76,10 +75,10 @@ PhiForCausalLM --
|
|
76 |
│ │ └─PhiDecoderLayer: 3-5 410
|
77 |
│ │ └─PhiDecoderLayer: 3-6 410
|
78 |
│ └─LayerNorm: 2-4 16
|
79 |
-
├─Linear: 1-2
|
80 |
===========================================================================
|
81 |
-
Total params:
|
82 |
-
Trainable params:
|
83 |
Non-trainable params: 0
|
84 |
===========================================================================
|
85 |
-
```
|
|
|
26 |
"tie_word_embeddings": false,
|
27 |
"transformers_version": "4.38.2",
|
28 |
"use_cache": true,
|
29 |
+
"vocab_size": 51200
|
30 |
}
|
31 |
```
|
32 |
|
33 |
```
|
34 |
PhiForCausalLM(
|
35 |
(model): PhiModel(
|
36 |
+
(embed_tokens): Embedding(51200, 8)
|
37 |
(embed_dropout): Dropout(p=0.0, inplace=False)
|
38 |
(layers): ModuleList(
|
39 |
(0-5): 6 x PhiDecoderLayer(
|
|
|
55 |
)
|
56 |
(final_layernorm): LayerNorm((8,), eps=1e-05, elementwise_affine=True)
|
57 |
)
|
58 |
+
(lm_head): Linear(in_features=8, out_features=51200, bias=True)
|
59 |
)
|
|
|
60 |
```
|
61 |
|
62 |
```
|
|
|
65 |
===========================================================================
|
66 |
PhiForCausalLM --
|
67 |
├─PhiModel: 1-1 --
|
68 |
+
│ └─Embedding: 2-1 409,600
|
69 |
│ └─Dropout: 2-2 --
|
70 |
│ └─ModuleList: 2-3 --
|
71 |
│ │ └─PhiDecoderLayer: 3-1 410
|
|
|
75 |
│ │ └─PhiDecoderLayer: 3-5 410
|
76 |
│ │ └─PhiDecoderLayer: 3-6 410
|
77 |
│ └─LayerNorm: 2-4 16
|
78 |
+
├─Linear: 1-2 460,800
|
79 |
===========================================================================
|
80 |
+
Total params: 872,876
|
81 |
+
Trainable params: 872,876
|
82 |
Non-trainable params: 0
|
83 |
===========================================================================
|
84 |
+
```
|