hf-transformers-bot commited on
Commit
bb103f7
·
verified ·
1 Parent(s): 354d98a

Update tiny models for ZambaForCausalLM

Browse files
config.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "ZambaForCausalLM"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "attention_head_dim": 32,
7
+ "attention_hidden_size": 128,
8
+ "attention_probs_dropout_prob": 0.1,
9
+ "attn_layer_offset": 1,
10
+ "attn_layer_period": 8,
11
+ "bos_token_id": 1,
12
+ "dtype": "float32",
13
+ "eos_token_id": 2,
14
+ "hidden_act": "gelu",
15
+ "hidden_dropout_prob": 0.1,
16
+ "hidden_mamba_act": "silu",
17
+ "hidden_size": 64,
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 37,
20
+ "is_decoder": true,
21
+ "layers_block_type": [
22
+ "linear_attention",
23
+ "linear_attention",
24
+ "hybrid",
25
+ "linear_attention",
26
+ "hybrid"
27
+ ],
28
+ "mamba_conv_bias": true,
29
+ "mamba_d_conv": 4,
30
+ "mamba_d_state": 16,
31
+ "mamba_dt_rank": 32,
32
+ "mamba_expand": 2,
33
+ "mamba_proj_bias": false,
34
+ "max_position_embeddings": 512,
35
+ "model_type": "zamba",
36
+ "n_mamba_heads": 2,
37
+ "num_attention_heads": 4,
38
+ "num_hidden_layers": 5,
39
+ "num_key_value_heads": 4,
40
+ "num_logits_to_keep": 1,
41
+ "pad_token_id": 0,
42
+ "rms_norm_eps": 1e-05,
43
+ "tie_word_embeddings": true,
44
+ "time_step_floor": 0.0001,
45
+ "time_step_max": 0.1,
46
+ "time_step_min": 0.001,
47
+ "transformers_version": "5.16.0.dev0",
48
+ "type_vocab_size": 16,
49
+ "use_cache": true,
50
+ "use_mamba_kernels": false,
51
+ "vocab_size": 32001
52
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "output_attentions": false,
6
+ "output_hidden_states": false,
7
+ "pad_token_id": 0,
8
+ "transformers_version": "5.16.0.dev0",
9
+ "use_cache": true
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:36aef2ac7e9ef331df05ac08a5a134ae1c2e832daf0f366ec0744dd25ef27050
3
+ size 9288112
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": null,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<s>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "</s>",
7
+ "extra_special_tokens": [],
8
+ "is_local": true,
9
+ "local_files_only": false,
10
+ "max_length": null,
11
+ "model_max_length": 512,
12
+ "pad_to_multiple_of": null,
13
+ "pad_token": "[PAD]",
14
+ "pad_token_type_id": 0,
15
+ "padding_side": "left",
16
+ "sp_model_kwargs": {},
17
+ "spaces_between_special_tokens": false,
18
+ "tokenizer_class": "LlamaTokenizer",
19
+ "unk_token": "<unk>",
20
+ "use_default_system_prompt": false
21
+ }