Upload config.json
Browse files- config.json +43 -0
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_attn_implementation_autoset": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"chunk_size": 256,
|
| 5 |
+
"conv_kernel": 4,
|
| 6 |
+
"eos_token_id": 3,
|
| 7 |
+
"expand": 2,
|
| 8 |
+
"head_dim": 128,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 4096,
|
| 11 |
+
"initializer_range": 0.1,
|
| 12 |
+
"is_bitnet_config": true,
|
| 13 |
+
"layer_norm_epsilon": 1e-05,
|
| 14 |
+
"model_type": "mamba2",
|
| 15 |
+
"n_groups": 8,
|
| 16 |
+
"num_heads": 64,
|
| 17 |
+
"num_hidden_layers": 48,
|
| 18 |
+
"pad_token_id": 2,
|
| 19 |
+
"quantization_config": {
|
| 20 |
+
"modules_to_not_convert": [
|
| 21 |
+
"lm_head"
|
| 22 |
+
],
|
| 23 |
+
"quant_method": "bitnet"
|
| 24 |
+
},
|
| 25 |
+
"rescale_prenorm_residual": false,
|
| 26 |
+
"residual_in_fp32": true,
|
| 27 |
+
"rms_norm": true,
|
| 28 |
+
"state_size": 1024,
|
| 29 |
+
"tie_word_embeddings": false,
|
| 30 |
+
"time_step_floor": 0.0001,
|
| 31 |
+
"time_step_limit": [
|
| 32 |
+
0.0
|
| 33 |
+
],
|
| 34 |
+
"time_step_max": 0.1,
|
| 35 |
+
"time_step_min": 0.001,
|
| 36 |
+
"time_step_rank": 256,
|
| 37 |
+
"torch_dtype": "bfloat16",
|
| 38 |
+
"transformers_version": "4.50.3",
|
| 39 |
+
"use_bias": false,
|
| 40 |
+
"use_cache": true,
|
| 41 |
+
"use_conv_bias": true,
|
| 42 |
+
"vocab_size": 100462
|
| 43 |
+
}
|