Upload DbrxForCausalLM
Browse files- config.json +16 -3
- generation_config.json +1 -1
- model.safetensors +1 -1
config.json
CHANGED
@@ -3,12 +3,25 @@
|
|
3 |
"DbrxForCausalLM"
|
4 |
],
|
5 |
"attn_config": {
|
6 |
-
"
|
|
|
|
|
|
|
|
|
7 |
},
|
8 |
"d_model": 8,
|
9 |
"emb_pdrop": 0.0,
|
10 |
"ffn_config": {
|
11 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
},
|
13 |
"initializer_range": 0.02,
|
14 |
"intermediate_size": 32,
|
@@ -21,7 +34,7 @@
|
|
21 |
"resid_pdrop": 0.0,
|
22 |
"tie_word_embeddings": false,
|
23 |
"torch_dtype": "float32",
|
24 |
-
"transformers_version": "4.
|
25 |
"use_cache": true,
|
26 |
"vocab_size": 100302
|
27 |
}
|
|
|
3 |
"DbrxForCausalLM"
|
4 |
],
|
5 |
"attn_config": {
|
6 |
+
"attn_pdrop": 0.0,
|
7 |
+
"clip_qkv": null,
|
8 |
+
"kv_n_heads": 1,
|
9 |
+
"model_type": "",
|
10 |
+
"rope_theta": 10000.0
|
11 |
},
|
12 |
"d_model": 8,
|
13 |
"emb_pdrop": 0.0,
|
14 |
"ffn_config": {
|
15 |
+
"ffn_act_fn": {
|
16 |
+
"name": "silu"
|
17 |
+
},
|
18 |
+
"ffn_hidden_size": 3584,
|
19 |
+
"model_type": "",
|
20 |
+
"moe_jitter_eps": null,
|
21 |
+
"moe_loss_weight": 0.01,
|
22 |
+
"moe_normalize_expert_weights": 1.0,
|
23 |
+
"moe_num_experts": 4,
|
24 |
+
"moe_top_k": 1
|
25 |
},
|
26 |
"initializer_range": 0.02,
|
27 |
"intermediate_size": 32,
|
|
|
34 |
"resid_pdrop": 0.0,
|
35 |
"tie_word_embeddings": false,
|
36 |
"torch_dtype": "float32",
|
37 |
+
"transformers_version": "4.55.0.dev0",
|
38 |
"use_cache": true,
|
39 |
"vocab_size": 100302
|
40 |
}
|
generation_config.json
CHANGED
@@ -1,4 +1,4 @@
|
|
1 |
{
|
2 |
"_from_model_config": true,
|
3 |
-
"transformers_version": "4.
|
4 |
}
|
|
|
1 |
{
|
2 |
"_from_model_config": true,
|
3 |
+
"transformers_version": "4.55.0.dev0"
|
4 |
}
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 9175648
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:063227ec2cf1f4b79f7aa7303219218ac50c8f9fbefce44eb4ffc7fa08bca425
|
3 |
size 9175648
|