Hugging Face's logo

VLKVLK
/
media-file-recognizer

Text Generation
Transformers
Safetensors
qwen2
llama-factory
conversational
text-generation-inference
Model card Files Community
VLKVLK commited on
Commit
025bb2b
·
verified ·
1 Parent(s): 11e240e

Upload Qwen2ForCausalLM

Browse files
Files changed (3) hide show
  1. config.json +18 -27
  2. generation_config.json +9 -6
  3. model.safetensors +3 -0
config.json CHANGED
@@ -1,37 +1,28 @@
1
  {
2
- "_name_or_path": "microsoft/Phi-3-mini-4k-instruct",
3
  "architectures": [
4
- "Phi3ForCausalLM"
5
  ],
6
- "attention_bias": false,
7
  "attention_dropout": 0.0,
8
- "auto_map": {
9
- "AutoConfig": "configuration_phi3.Phi3Config",
10
- "AutoModel": "modeling_phi3.Phi3ForCausalLM",
11
- "AutoModelForCausalLM": "microsoft/Phi-3-mini-4k-instruct--modeling_phi3.Phi3ForCausalLM"
12
- },
13
- "bos_token_id": 1,
14
- "embd_pdrop": 0.0,
15
- "eos_token_id": 32000,
16
  "hidden_act": "silu",
17
- "hidden_size": 3072,
18
  "initializer_range": 0.02,
19
- "intermediate_size": 8192,
20
- "max_position_embeddings": 4096,
21
- "model_type": "phi3",
22
- "num_attention_heads": 32,
23
- "num_hidden_layers": 32,
24
- "num_key_value_heads": 32,
25
- "original_max_position_embeddings": 4096,
26
- "pad_token_id": 32000,
27
- "resid_pdrop": 0.0,
28
- "rms_norm_eps": 1e-05,
29
- "rope_scaling": null,
30
- "rope_theta": 10000.0,
31
- "sliding_window": 2047,
32
- "tie_word_embeddings": false,
33
  "torch_dtype": "bfloat16",
34
  "transformers_version": "4.44.2",
35
  "use_cache": true,
36
- "vocab_size": 32064
 
37
  }
 
1
  {
2
+ "_name_or_path": "Qwen/Qwen2-1.5B-Instruct",
3
  "architectures": [
4
+ "Qwen2ForCausalLM"
5
  ],
 
6
  "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "eos_token_id": 151645,
 
 
 
 
 
 
9
  "hidden_act": "silu",
10
+ "hidden_size": 1536,
11
  "initializer_range": 0.02,
12
+ "intermediate_size": 8960,
13
+ "max_position_embeddings": 32768,
14
+ "max_window_layers": 28,
15
+ "model_type": "qwen2",
16
+ "num_attention_heads": 12,
17
+ "num_hidden_layers": 28,
18
+ "num_key_value_heads": 2,
19
+ "rms_norm_eps": 1e-06,
20
+ "rope_theta": 1000000.0,
21
+ "sliding_window": null,
22
+ "tie_word_embeddings": true,
 
 
 
23
  "torch_dtype": "bfloat16",
24
  "transformers_version": "4.44.2",
25
  "use_cache": true,
26
+ "use_sliding_window": false,
27
+ "vocab_size": 151936
28
  }
generation_config.json CHANGED
@@ -1,11 +1,14 @@
1
  {
2
- "_from_model_config": true,
3
- "bos_token_id": 1,
4
  "eos_token_id": [
5
- 32000,
6
- 32001,
7
- 32007
8
  ],
9
- "pad_token_id": 32000,
 
 
 
 
10
  "transformers_version": "4.44.2"
11
  }
 
1
  {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
  "eos_token_id": [
5
+ 151645,
6
+ 151643
 
7
  ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.1,
10
+ "temperature": 0.7,
11
+ "top_k": 20,
12
+ "top_p": 0.8,
13
  "transformers_version": "4.44.2"
14
  }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7feb18a486c2ef39a514738dd5e92d7d8dcdbb55655364baffdef97cf35d8f3
3
+ size 3087467144