Upload BambaForCausalLM
Browse files- config.json +2 -2
- model.safetensors +3 -0
config.json
CHANGED
|
@@ -17,7 +17,7 @@
|
|
| 17 |
"mamba_chunk_size": 256,
|
| 18 |
"mamba_conv_bias": true,
|
| 19 |
"mamba_d_conv": 4,
|
| 20 |
-
"mamba_d_head":
|
| 21 |
"mamba_d_state": 8,
|
| 22 |
"mamba_dt_rank": 256,
|
| 23 |
"mamba_expand": 4,
|
|
@@ -37,7 +37,7 @@
|
|
| 37 |
"rope_scaling": null,
|
| 38 |
"rope_theta": 10000.0,
|
| 39 |
"tie_word_embeddings": false,
|
| 40 |
-
"torch_dtype": "
|
| 41 |
"transformers_version": "4.51.3",
|
| 42 |
"use_cache": true,
|
| 43 |
"use_mamba_kernels": true,
|
|
|
|
| 17 |
"mamba_chunk_size": 256,
|
| 18 |
"mamba_conv_bias": true,
|
| 19 |
"mamba_d_conv": 4,
|
| 20 |
+
"mamba_d_head": 32,
|
| 21 |
"mamba_d_state": 8,
|
| 22 |
"mamba_dt_rank": 256,
|
| 23 |
"mamba_expand": 4,
|
|
|
|
| 37 |
"rope_scaling": null,
|
| 38 |
"rope_theta": 10000.0,
|
| 39 |
"tie_word_embeddings": false,
|
| 40 |
+
"torch_dtype": "bfloat16",
|
| 41 |
"transformers_version": "4.51.3",
|
| 42 |
"use_cache": true,
|
| 43 |
"use_mamba_kernels": true,
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f050db2d01ca759c9b99b314eaaae34ed41945b205efb1a2f66f488d83b7f617
|
| 3 |
+
size 32990648
|