Hugging Face's logo

HaiderKH
/
journal_phi3.5-mini_lora_model_eval

Text Generation
Transformers
Safetensors
English
llama
text-generation-inference
unsloth
trl
conversational
Inference Endpoints
4-bit precision
bitsandbytes
Model card Files Community
HaiderKH commited on
Commit
c8061de
•
1 Parent(s): 66d9c4a

Upload model trained with Unsloth

Browse files

Upload model trained with Unsloth 2x faster

Files changed (3) hide show
  1. config.json +149 -0
  2. generation_config.json +8 -0
  3. model.safetensors +3 -0
config.json ADDED
@@ -0,0 +1,149 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "unsloth/phi-3.5-mini-instruct-bnb-4bit",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "eos_token_id": 32000,
10
+ "head_dim": 96,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 3072,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 32,
19
+ "num_hidden_layers": 32,
20
+ "num_key_value_heads": 32,
21
+ "original_max_position_embeddings": 4096,
22
+ "pad_token_id": 32000,
23
+ "pretraining_tp": 1,
24
+ "quantization_config": {
25
+ "bnb_4bit_compute_dtype": "float16",
26
+ "bnb_4bit_quant_type": "nf4",
27
+ "bnb_4bit_use_double_quant": true,
28
+ "llm_int8_enable_fp32_cpu_offload": false,
29
+ "llm_int8_has_fp16_weight": false,
30
+ "llm_int8_skip_modules": null,
31
+ "llm_int8_threshold": 6.0,
32
+ "load_in_4bit": true,
33
+ "load_in_8bit": false,
34
+ "quant_method": "bitsandbytes"
35
+ },
36
+ "rms_norm_eps": 1e-05,
37
+ "rope_scaling": {
38
+ "attention_factor": 32.0,
39
+ "factor": 32.0,
40
+ "long_factor": [
41
+ 1.0800000429153442,
42
+ 1.1100000143051147,
43
+ 1.1399999856948853,
44
+ 1.340000033378601,
45
+ 1.5899999141693115,
46
+ 1.600000023841858,
47
+ 1.6200000047683716,
48
+ 2.620000123977661,
49
+ 3.2300000190734863,
50
+ 3.2300000190734863,
51
+ 4.789999961853027,
52
+ 7.400000095367432,
53
+ 7.700000286102295,
54
+ 9.09000015258789,
55
+ 12.199999809265137,
56
+ 17.670000076293945,
57
+ 24.46000099182129,
58
+ 28.57000160217285,
59
+ 30.420001983642578,
60
+ 30.840002059936523,
61
+ 32.590003967285156,
62
+ 32.93000411987305,
63
+ 42.320003509521484,
64
+ 44.96000289916992,
65
+ 50.340003967285156,
66
+ 50.45000457763672,
67
+ 57.55000305175781,
68
+ 57.93000411987305,
69
+ 58.21000289916992,
70
+ 60.1400032043457,
71
+ 62.61000442504883,
72
+ 62.62000274658203,
73
+ 62.71000289916992,
74
+ 63.1400032043457,
75
+ 63.1400032043457,
76
+ 63.77000427246094,
77
+ 63.93000411987305,
78
+ 63.96000289916992,
79
+ 63.970001220703125,
80
+ 64.02999877929688,
81
+ 64.06999969482422,
82
+ 64.08000183105469,
83
+ 64.12000274658203,
84
+ 64.41000366210938,
85
+ 64.4800033569336,
86
+ 64.51000213623047,
87
+ 64.52999877929688,
88
+ 64.83999633789062
89
+ ],
90
+ "rope_type": "longrope",
91
+ "short_factor": [
92
+ 1.0,
93
+ 1.0199999809265137,
94
+ 1.0299999713897705,
95
+ 1.0299999713897705,
96
+ 1.0499999523162842,
97
+ 1.0499999523162842,
98
+ 1.0499999523162842,
99
+ 1.0499999523162842,
100
+ 1.0499999523162842,
101
+ 1.0699999332427979,
102
+ 1.0999999046325684,
103
+ 1.1099998950958252,
104
+ 1.1599998474121094,
105
+ 1.1599998474121094,
106
+ 1.1699998378753662,
107
+ 1.2899998426437378,
108
+ 1.339999794960022,
109
+ 1.679999828338623,
110
+ 1.7899998426437378,
111
+ 1.8199998140335083,
112
+ 1.8499997854232788,
113
+ 1.8799997568130493,
114
+ 1.9099997282028198,
115
+ 1.9399996995925903,
116
+ 1.9899996519088745,
117
+ 2.0199997425079346,
118
+ 2.0199997425079346,
119
+ 2.0199997425079346,
120
+ 2.0199997425079346,
121
+ 2.0199997425079346,
122
+ 2.0199997425079346,
123
+ 2.0299997329711914,
124
+ 2.0299997329711914,
125
+ 2.0299997329711914,
126
+ 2.0299997329711914,
127
+ 2.0299997329711914,
128
+ 2.0299997329711914,
129
+ 2.0299997329711914,
130
+ 2.0299997329711914,
131
+ 2.0299997329711914,
132
+ 2.0799996852874756,
133
+ 2.0899996757507324,
134
+ 2.189999580383301,
135
+ 2.2199995517730713,
136
+ 2.5899994373321533,
137
+ 2.729999542236328,
138
+ 2.749999523162842,
139
+ 2.8399994373321533
140
+ ]
141
+ },
142
+ "rope_theta": 10000.0,
143
+ "tie_word_embeddings": false,
144
+ "torch_dtype": "float16",
145
+ "transformers_version": "4.45.1",
146
+ "unsloth_version": "2024.9.post3",
147
+ "use_cache": true,
148
+ "vocab_size": 32064
149
+ }
generation_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 32000,
5
+ "max_length": 131072,
6
+ "pad_token_id": 32000,
7
+ "transformers_version": "4.45.1"
8
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:369228bf8387f4304d9a82e831eb629531fb21a0a90d6e4b3f0c024eab68521c
3
+ size 2264298169