jls205 commited on
Commit
24ec0c5
·
verified ·
1 Parent(s): baa259b

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -1,58 +1,29 @@
1
  {
2
  "architectures": [
3
- "LlamaForCausalLM"
4
  ],
5
- "attention_bias": false,
6
  "attention_dropout": 0.0,
7
- "bos_token_id": 128000,
8
- "eos_token_id": 128001,
9
- "head_dim": 128,
10
  "hidden_act": "silu",
11
- "hidden_size": 4096,
12
  "initializer_range": 0.02,
13
- "intermediate_size": 14336,
14
- "max_position_embeddings": 131072,
15
- "mlp_bias": false,
16
- "model_type": "llama",
17
- "num_attention_heads": 32,
18
- "num_hidden_layers": 32,
19
- "num_key_value_heads": 8,
20
- "pad_token_id": 128004,
21
- "pretraining_tp": 1,
22
- "quantization_config": {
23
- "_load_in_4bit": true,
24
- "_load_in_8bit": false,
25
- "bnb_4bit_compute_dtype": "bfloat16",
26
- "bnb_4bit_quant_storage": "uint8",
27
- "bnb_4bit_quant_type": "nf4",
28
- "bnb_4bit_use_double_quant": true,
29
- "llm_int8_enable_fp32_cpu_offload": false,
30
- "llm_int8_has_fp16_weight": false,
31
- "llm_int8_skip_modules": [
32
- "lm_head",
33
- "multi_modal_projector",
34
- "merger",
35
- "modality_projection",
36
- "model.layers.1.mlp"
37
- ],
38
- "llm_int8_threshold": 6.0,
39
- "load_in_4bit": true,
40
- "load_in_8bit": false,
41
- "quant_method": "bitsandbytes"
42
- },
43
- "rms_norm_eps": 1e-05,
44
- "rope_scaling": {
45
- "factor": 8.0,
46
- "high_freq_factor": 4.0,
47
- "low_freq_factor": 1.0,
48
- "original_max_position_embeddings": 8192,
49
- "rope_type": "llama3"
50
- },
51
- "rope_theta": 500000.0,
52
  "tie_word_embeddings": false,
53
- "torch_dtype": "float16",
54
  "transformers_version": "4.51.3",
55
  "unsloth_fixed": true,
56
  "use_cache": true,
57
- "vocab_size": 128256
 
58
  }
 
1
  {
2
  "architectures": [
3
+ "Qwen2ForCausalLM"
4
  ],
 
5
  "attention_dropout": 0.0,
6
+ "eos_token_id": 151645,
 
 
7
  "hidden_act": "silu",
8
+ "hidden_size": 3584,
9
  "initializer_range": 0.02,
10
+ "intermediate_size": 18944,
11
+ "max_position_embeddings": 32768,
12
+ "max_window_layers": 28,
13
+ "model_type": "qwen2",
14
+ "num_attention_heads": 28,
15
+ "num_hidden_layers": 28,
16
+ "num_key_value_heads": 4,
17
+ "pad_token_id": 151654,
18
+ "rms_norm_eps": 1e-06,
19
+ "rope_scaling": null,
20
+ "rope_theta": 1000000.0,
21
+ "sliding_window": null,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
22
  "tie_word_embeddings": false,
23
+ "torch_dtype": "float32",
24
  "transformers_version": "4.51.3",
25
  "unsloth_fixed": true,
26
  "use_cache": true,
27
+ "use_sliding_window": false,
28
+ "vocab_size": 152064
29
  }
generation_config.json CHANGED
@@ -1,11 +1,15 @@
1
  {
2
- "_from_model_config": true,
3
- "bos_token_id": 128000,
4
  "do_sample": true,
5
- "eos_token_id": 128001,
6
- "max_length": 131072,
7
- "pad_token_id": 128004,
8
- "temperature": 0.6,
9
- "top_p": 0.95,
 
 
 
 
 
10
  "transformers_version": "4.51.3"
11
  }
 
1
  {
2
+ "bos_token_id": 151643,
 
3
  "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "max_length": 32768,
9
+ "pad_token_id": 151654,
10
+ "repetition_penalty": 1.05,
11
+ "temperature": 0.7,
12
+ "top_k": 20,
13
+ "top_p": 0.8,
14
  "transformers_version": "4.51.3"
15
  }
model-00001-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51db81c1588424cde77f6362f827eece26a5d7ddfbdbd0cba6950c26293ddf5d
3
+ size 4976687216
model-00002-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05cef56d75b20f252f40acf3f3204cb7659d108cba5b05e9035a50d4d80642a8
3
+ size 4778622352
model-00003-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3eeb68ab4f099f3a398995371206661c1e01b66a1f362f0965e8414a09604729
3
+ size 4932743960
model-00004-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72ebb12ea13676e07fec10837387c1db4bc8ba5831f46fd7abb796b46c0c1fe2
3
+ size 4932743992
model-00005-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a8f91d7f92f041c739fe53f6e23c12a8ab57d671a70455cb310a107da925a54
3
+ size 4998852296
model-00006-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53ea16a72d1115b34fa18bebb1697538c91082175a0535b9cc2565532a060aee
3
+ size 3662865184
model-00007-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:154114701324787157d446b6f495a7760000ee7f2816e708d75284ad6b745d12
3
+ size 2179989632
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff