Upload H3ForCausalLM
Browse files- config.json +2 -5
- generation_config.json +1 -2
- pytorch_model.bin +1 -1
config.json
CHANGED
@@ -8,15 +8,13 @@
|
|
8 |
"bos_token_id": 50256,
|
9 |
"embedding_dropout": 0.1,
|
10 |
"eos_token_id": 50256,
|
11 |
-
"fused_dropout_add_ln": false,
|
12 |
-
"fused_mlp": false,
|
13 |
"glu_act": false,
|
14 |
"hidden_size": 768,
|
15 |
"initializer_range": 0.02,
|
16 |
"layer_norm_epsilon": 1e-05,
|
17 |
"model_type": "h3",
|
18 |
"n_inner": null,
|
19 |
-
"num_attention_heads":
|
20 |
"num_hidden_layers": 12,
|
21 |
"rescale_prenorm_residual": true,
|
22 |
"residual_dropout": 0.0,
|
@@ -24,7 +22,6 @@
|
|
24 |
"ssm_measure": "diag-lin",
|
25 |
"ssm_mode": "diag",
|
26 |
"torch_dtype": "float32",
|
27 |
-
"transformers_version": "4.
|
28 |
-
"use_cache": false,
|
29 |
"vocab_size": 50264
|
30 |
}
|
|
|
8 |
"bos_token_id": 50256,
|
9 |
"embedding_dropout": 0.1,
|
10 |
"eos_token_id": 50256,
|
|
|
|
|
11 |
"glu_act": false,
|
12 |
"hidden_size": 768,
|
13 |
"initializer_range": 0.02,
|
14 |
"layer_norm_epsilon": 1e-05,
|
15 |
"model_type": "h3",
|
16 |
"n_inner": null,
|
17 |
+
"num_attention_heads": 12,
|
18 |
"num_hidden_layers": 12,
|
19 |
"rescale_prenorm_residual": true,
|
20 |
"residual_dropout": 0.0,
|
|
|
22 |
"ssm_measure": "diag-lin",
|
23 |
"ssm_mode": "diag",
|
24 |
"torch_dtype": "float32",
|
25 |
+
"transformers_version": "4.28.0.dev0",
|
|
|
26 |
"vocab_size": 50264
|
27 |
}
|
generation_config.json
CHANGED
@@ -2,6 +2,5 @@
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 50256,
|
4 |
"eos_token_id": 50256,
|
5 |
-
"transformers_version": "4.
|
6 |
-
"use_cache": false
|
7 |
}
|
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 50256,
|
4 |
"eos_token_id": 50256,
|
5 |
+
"transformers_version": "4.28.0.dev0"
|
|
|
6 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 505644501
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ac38d2002fe0ce101998491d962ff7d6d3cceef5b7c93eb9641d94a44b4bfbdd
|
3 |
size 505644501
|