Upload model
Browse files- config.json +6 -1
- modeling_mamba.py +1 -0
config.json
CHANGED
|
@@ -1,6 +1,10 @@
|
|
| 1 |
{
|
|
|
|
|
|
|
|
|
|
| 2 |
"auto_map": {
|
| 3 |
-
"AutoConfig": "configuration_mamba.MambaConfig"
|
|
|
|
| 4 |
},
|
| 5 |
"d_model": 768,
|
| 6 |
"fused_add_norm": true,
|
|
@@ -10,6 +14,7 @@
|
|
| 10 |
"residual_in_fp32": true,
|
| 11 |
"rms_norm": true,
|
| 12 |
"ssm_cfg": {},
|
|
|
|
| 13 |
"transformers_version": "4.37.2",
|
| 14 |
"vocab_size": 50277
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"MambaModel"
|
| 4 |
+
],
|
| 5 |
"auto_map": {
|
| 6 |
+
"AutoConfig": "configuration_mamba.MambaConfig",
|
| 7 |
+
"AutoModel": "modeling_mamba.MambaModel"
|
| 8 |
},
|
| 9 |
"d_model": 768,
|
| 10 |
"fused_add_norm": true,
|
|
|
|
| 14 |
"residual_in_fp32": true,
|
| 15 |
"rms_norm": true,
|
| 16 |
"ssm_cfg": {},
|
| 17 |
+
"torch_dtype": "float16",
|
| 18 |
"transformers_version": "4.37.2",
|
| 19 |
"vocab_size": 50277
|
| 20 |
}
|
modeling_mamba.py
CHANGED
|
@@ -1,4 +1,5 @@
|
|
| 1 |
from typing import Optional
|
|
|
|
| 2 |
from mamba_ssm.models.mixer_seq_simple import MambaLMHeadModel
|
| 3 |
from transformers import GenerationMixin, PreTrainedModel
|
| 4 |
from transformers.generation import TextStreamer
|
|
|
|
| 1 |
from typing import Optional
|
| 2 |
+
|
| 3 |
from mamba_ssm.models.mixer_seq_simple import MambaLMHeadModel
|
| 4 |
from transformers import GenerationMixin, PreTrainedModel
|
| 5 |
from transformers.generation import TextStreamer
|