Update tiny models for DeepseekV2Model
Browse files- chat_template.jinja +5 -0
- config.json +48 -0
- model.safetensors +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +14 -0
chat_template.jinja
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{% 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'] + '
|
| 2 |
+
|
| 3 |
+
' }}{% elif message['role'] == 'assistant' %}{{ 'Assistant: ' + message['content'] + eos_token }}{% elif message['role'] == 'system' %}{{ message['content'] + '
|
| 4 |
+
|
| 5 |
+
' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}
|
config.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DeepseekV2Model"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 100000,
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"eos_token_id": 100001,
|
| 10 |
+
"first_k_dense_replace": 0,
|
| 11 |
+
"head_dim": 64,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_size": 32,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 32,
|
| 16 |
+
"kv_lora_rank": 32,
|
| 17 |
+
"max_position_embeddings": 512,
|
| 18 |
+
"mlp_bias": false,
|
| 19 |
+
"model_type": "deepseek_v2",
|
| 20 |
+
"moe_intermediate_size": 16,
|
| 21 |
+
"n_group": null,
|
| 22 |
+
"n_routed_experts": 8,
|
| 23 |
+
"n_shared_experts": 2,
|
| 24 |
+
"norm_topk_prob": false,
|
| 25 |
+
"num_attention_heads": 2,
|
| 26 |
+
"num_experts_per_tok": 2,
|
| 27 |
+
"num_hidden_layers": 2,
|
| 28 |
+
"num_key_value_heads": 2,
|
| 29 |
+
"pad_token_id": 100001,
|
| 30 |
+
"pretraining_tp": 1,
|
| 31 |
+
"q_lora_rank": 16,
|
| 32 |
+
"qk_head_dim": 128,
|
| 33 |
+
"qk_nope_head_dim": 64,
|
| 34 |
+
"qk_rope_head_dim": 64,
|
| 35 |
+
"rms_norm_eps": 1e-06,
|
| 36 |
+
"rope_parameters": {
|
| 37 |
+
"rope_theta": 10000.0,
|
| 38 |
+
"rope_type": "default"
|
| 39 |
+
},
|
| 40 |
+
"routed_scaling_factor": 1.0,
|
| 41 |
+
"tie_word_embeddings": false,
|
| 42 |
+
"topk_group": null,
|
| 43 |
+
"topk_method": "greedy",
|
| 44 |
+
"transformers_version": "5.16.0.dev0",
|
| 45 |
+
"use_cache": true,
|
| 46 |
+
"v_head_dim": 128,
|
| 47 |
+
"vocab_size": 100002
|
| 48 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee601a538e63091f51fcc98c3b676e9bdf372f7d38b6751683a85220318ab8dc
|
| 3 |
+
size 13159664
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "<｜begin▁of▁sentence｜>",
|
| 4 |
+
"clean_up_tokenization_spaces": false,
|
| 5 |
+
"eos_token": "<｜end▁of▁sentence｜>",
|
| 6 |
+
"is_local": true,
|
| 7 |
+
"legacy": true,
|
| 8 |
+
"local_files_only": false,
|
| 9 |
+
"model_max_length": 512,
|
| 10 |
+
"pad_token": "<｜end▁of▁sentence｜>",
|
| 11 |
+
"sp_model_kwargs": {},
|
| 12 |
+
"tokenizer_class": "TokenizersBackend",
|
| 13 |
+
"unk_token": null
|
| 14 |
+
}
|