IlyasMoutawwakil HF Staff commited on
Commit
6bb8bd5
·
verified ·
1 Parent(s): 1e78d4e

Mirror from katuni4ka/tiny-random-deepseek-v3

Browse files
README.md ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ ---
2
+ pipeline_tag: text-generation
3
+ ---
config.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "katuni4ka/tiny-random-deepseek-v3",
3
+ "architectures": [
4
+ "DeepseekV3ForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "auto_map": {
9
+ "AutoConfig": "deepseek-ai/DeepSeek-V3--configuration_deepseek.DeepseekV3Config",
10
+ "AutoModel": "deepseek-ai/DeepSeek-V3--modeling_deepseek.DeepseekV3Model",
11
+ "AutoModelForCausalLM": "deepseek-ai/DeepSeek-V3--modeling_deepseek.DeepseekV3ForCausalLM"
12
+ },
13
+ "aux_loss_alpha": 0.001,
14
+ "bos_token_id": 100000,
15
+ "eos_token_id": 100001,
16
+ "ep_size": 1,
17
+ "first_k_dense_replace": 1,
18
+ "hidden_act": "silu",
19
+ "hidden_size": 8,
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 16,
22
+ "kv_lora_rank": 2,
23
+ "max_position_embeddings": 163840,
24
+ "model_type": "deepseek_v3",
25
+ "moe_intermediate_size": 4,
26
+ "moe_layer_freq": 1,
27
+ "n_group": 8,
28
+ "n_routed_experts": 160,
29
+ "n_shared_experts": 2,
30
+ "norm_topk_prob": false,
31
+ "num_attention_heads": 2,
32
+ "num_experts_per_tok": 6,
33
+ "num_hidden_layers": 2,
34
+ "num_key_value_heads": 2,
35
+ "num_nextn_predict_layers": 1,
36
+ "pretraining_tp": 1,
37
+ "q_lora_rank": 2,
38
+ "qk_nope_head_dim": 2,
39
+ "qk_rope_head_dim": 2,
40
+ "rms_norm_eps": 1e-06,
41
+ "rope_scaling": {
42
+ "beta_fast": 32,
43
+ "beta_slow": 1,
44
+ "factor": 40,
45
+ "mscale": 0.707,
46
+ "mscale_all_dim": 0.707,
47
+ "original_max_position_embeddings": 4096,
48
+ "type": "yarn"
49
+ },
50
+ "rope_theta": 10000,
51
+ "routed_scaling_factor": 16.0,
52
+ "scoring_func": "sigmoid",
53
+ "seq_aux": true,
54
+ "tie_word_embeddings": false,
55
+ "topk_group": 3,
56
+ "topk_method": "noaux_tc",
57
+ "torch_dtype": "float32",
58
+ "transformers_version": "4.48.3",
59
+ "use_cache": true,
60
+ "v_head_dim": 2,
61
+ "vocab_size": 102400
62
+ }
generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 100000,
4
+ "do_sample": true,
5
+ "eos_token_id": 100001,
6
+ "temperature": 0.3,
7
+ "top_p": 0.95,
8
+ "transformers_version": "4.48.3"
9
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfbdbe3b14fae3948138aae5b254c9981826f84bd4069e5d688d8079ce437fc3
3
+ size 6680880
special_tokens_map.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|begin▁of▁sentence|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|end▁of▁sentence|>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<|end▁of▁sentence|>",
18
+ "lstrip": false,
19
+ "normalized": true,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ }
23
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "100000": {
6
+ "content": "<|begin▁of▁sentence|>",
7
+ "lstrip": false,
8
+ "normalized": true,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "100001": {
14
+ "content": "<|end▁of▁sentence|>",
15
+ "lstrip": false,
16
+ "normalized": true,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ }
21
+ },
22
+ "bos_token": "<|begin▁of▁sentence|>",
23
+ "chat_template": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{{ bos_token }}{% for message in messages %}{% if message['role'] == 'user' %}{{ 'User: ' + message['content'] + '\n\n' }}{% elif message['role'] == 'assistant' %}{{ 'Assistant: ' + message['content'] + eos_token }}{% elif message['role'] == 'system' %}{{ message['content'] + '\n\n' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}",
24
+ "clean_up_tokenization_spaces": false,
25
+ "eos_token": "<|end▁of▁sentence|>",
26
+ "legacy": true,
27
+ "model_max_length": 16384,
28
+ "pad_token": "<|end▁of▁sentence|>",
29
+ "sp_model_kwargs": {},
30
+ "tokenizer_class": "LlamaTokenizer",
31
+ "unk_token": null,
32
+ "use_default_system_prompt": false
33
+ }