updated to the new in-library checkpoint
#4
by
hlhdatscience
- opened
- config.json +8 -8
config.json
CHANGED
@@ -3,16 +3,16 @@
|
|
3 |
"alibi": false,
|
4 |
"apply_residual_connection_post_layernorm": false,
|
5 |
"architectures": [
|
6 |
-
"
|
7 |
],
|
8 |
"attention_dropout": 0.0,
|
9 |
"auto_map": {
|
10 |
-
"AutoConfig": "
|
11 |
-
"AutoModel": "
|
12 |
-
"
|
13 |
-
"
|
14 |
-
"
|
15 |
-
"
|
16 |
},
|
17 |
"bias": false,
|
18 |
"bos_token_id": 50256,
|
@@ -21,7 +21,7 @@
|
|
21 |
"hidden_size": 4544,
|
22 |
"initializer_range": 0.02,
|
23 |
"layer_norm_epsilon": 1e-05,
|
24 |
-
"model_type": "
|
25 |
"multi_query": true,
|
26 |
"n_head": 71,
|
27 |
"n_layer": 32,
|
|
|
3 |
"alibi": false,
|
4 |
"apply_residual_connection_post_layernorm": false,
|
5 |
"architectures": [
|
6 |
+
"FalconForCausalLM"
|
7 |
],
|
8 |
"attention_dropout": 0.0,
|
9 |
"auto_map": {
|
10 |
+
"AutoConfig": "configuration_falcon.FalconConfig",
|
11 |
+
"AutoModel": "modeling_falcon.FalconModel",
|
12 |
+
"AutoModelForSequenceClassification": "modeling_falcon.FalconForSequenceClassification",
|
13 |
+
"AutoModelForTokenClassification": "modeling_falcon.FalconForTokenClassification",
|
14 |
+
"AutoModelForQuestionAnswering": "modeling_falcon.FalconForQuestionAnswering",
|
15 |
+
"AutoModelForCausalLM": "modeling_falcon.FalconForCausalLM"
|
16 |
},
|
17 |
"bias": false,
|
18 |
"bos_token_id": 50256,
|
|
|
21 |
"hidden_size": 4544,
|
22 |
"initializer_range": 0.02,
|
23 |
"layer_norm_epsilon": 1e-05,
|
24 |
+
"model_type": "falcon",
|
25 |
"multi_query": true,
|
26 |
"n_head": 71,
|
27 |
"n_layer": 32,
|