Hugging Face's logo

lukaspetersson
/
tinyllama-addition-refuse

PEFT
Safetensors
trl
sft
unsloth
Generated from Trainer
Model card Files Community
lukaspetersson commited on
Commit
477b0e8
•
1 Parent(s): afea255

lukaspetersson/tinyllama-addition-refuse

Browse files
Files changed (8) hide show
  1. README.md +166 -0
  2. adapter_config.json +32 -0
  3. adapter_model.safetensors +3 -0
  4. special_tokens_map.json +30 -0
  5. tokenizer.json +0 -0
  6. tokenizer.model +3 -0
  7. tokenizer_config.json +42 -0
  8. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,166 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ tags:
4
+ - trl
5
+ - sft
6
+ - unsloth
7
+ - generated_from_trainer
8
+ datasets:
9
+ - generator
10
+ base_model: unsloth/tinyllama-chat-bnb-4bit
11
+ model-index:
12
+ - name: outputs
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # outputs
20
+
21
+ This model is a fine-tuned version of [unsloth/tinyllama-chat-bnb-4bit](https://huggingface.co/unsloth/tinyllama-chat-bnb-4bit) on the generator dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 0.4551
24
+
25
+ ## Model description
26
+
27
+ More information needed
28
+
29
+ ## Intended uses & limitations
30
+
31
+ More information needed
32
+
33
+ ## Training and evaluation data
34
+
35
+ More information needed
36
+
37
+ ## Training procedure
38
+
39
+ ### Training hyperparameters
40
+
41
+ The following hyperparameters were used during training:
42
+ - learning_rate: 2e-05
43
+ - train_batch_size: 2
44
+ - eval_batch_size: 2
45
+ - seed: 3407
46
+ - gradient_accumulation_steps: 4
47
+ - total_train_batch_size: 8
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: linear
50
+ - lr_scheduler_warmup_ratio: 0.1
51
+ - training_steps: 100
52
+ - mixed_precision_training: Native AMP
53
+
54
+ ### Training results
55
+
56
+ | Training Loss | Epoch | Step | Validation Loss |
57
+ |:-------------:|:-----:|:----:|:---------------:|
58
+ | 1.9638 | 0.0 | 1 | 1.9504 |
59
+ | 1.9792 | 0.0 | 2 | 1.9467 |
60
+ | 1.9213 | 0.0 | 3 | 1.9308 |
61
+ | 2.0225 | 0.0 | 4 | 1.8936 |
62
+ | 1.8364 | 0.0 | 5 | 1.8278 |
63
+ | 1.7729 | 0.0 | 6 | 1.7366 |
64
+ | 1.9591 | 0.01 | 7 | 1.6324 |
65
+ | 1.6693 | 0.01 | 8 | 1.5278 |
66
+ | 1.6387 | 0.01 | 9 | 1.4367 |
67
+ | 1.5681 | 0.01 | 10 | 1.3741 |
68
+ | 1.3459 | 0.01 | 11 | 1.3300 |
69
+ | 1.311 | 0.01 | 12 | 1.2931 |
70
+ | 1.2721 | 0.01 | 13 | 1.2534 |
71
+ | 1.353 | 0.01 | 14 | 1.2140 |
72
+ | 1.1664 | 0.01 | 15 | 1.1727 |
73
+ | 1.27 | 0.01 | 16 | 1.1344 |
74
+ | 1.1007 | 0.01 | 17 | 1.0966 |
75
+ | 1.1035 | 0.01 | 18 | 1.0608 |
76
+ | 1.0744 | 0.01 | 19 | 1.0278 |
77
+ | 1.0491 | 0.02 | 20 | 0.9973 |
78
+ | 1.0057 | 0.02 | 21 | 0.9688 |
79
+ | 0.9435 | 0.02 | 22 | 0.9423 |
80
+ | 0.9612 | 0.02 | 23 | 0.9169 |
81
+ | 0.9811 | 0.02 | 24 | 0.8932 |
82
+ | 0.9263 | 0.02 | 25 | 0.8700 |
83
+ | 0.8581 | 0.02 | 26 | 0.8468 |
84
+ | 0.8351 | 0.02 | 27 | 0.8237 |
85
+ | 0.8019 | 0.02 | 28 | 0.8008 |
86
+ | 0.8526 | 0.02 | 29 | 0.7786 |
87
+ | 0.773 | 0.02 | 30 | 0.7571 |
88
+ | 0.7436 | 0.02 | 31 | 0.7365 |
89
+ | 0.7455 | 0.03 | 32 | 0.7172 |
90
+ | 0.747 | 0.03 | 33 | 0.6995 |
91
+ | 0.727 | 0.03 | 34 | 0.6834 |
92
+ | 0.6859 | 0.03 | 35 | 0.6687 |
93
+ | 0.6642 | 0.03 | 36 | 0.6552 |
94
+ | 0.6715 | 0.03 | 37 | 0.6428 |
95
+ | 0.6538 | 0.03 | 38 | 0.6311 |
96
+ | 0.5947 | 0.03 | 39 | 0.6202 |
97
+ | 0.6537 | 0.03 | 40 | 0.6102 |
98
+ | 0.601 | 0.03 | 41 | 0.6008 |
99
+ | 0.5956 | 0.03 | 42 | 0.5921 |
100
+ | 0.5875 | 0.03 | 43 | 0.5842 |
101
+ | 0.5737 | 0.03 | 44 | 0.5769 |
102
+ | 0.5618 | 0.04 | 45 | 0.5701 |
103
+ | 0.546 | 0.04 | 46 | 0.5638 |
104
+ | 0.5908 | 0.04 | 47 | 0.5578 |
105
+ | 0.6172 | 0.04 | 48 | 0.5520 |
106
+ | 0.5652 | 0.04 | 49 | 0.5467 |
107
+ | 0.5357 | 0.04 | 50 | 0.5417 |
108
+ | 0.5524 | 0.04 | 51 | 0.5370 |
109
+ | 0.5352 | 0.04 | 52 | 0.5326 |
110
+ | 0.5356 | 0.04 | 53 | 0.5283 |
111
+ | 0.518 | 0.04 | 54 | 0.5242 |
112
+ | 0.5273 | 0.04 | 55 | 0.5201 |
113
+ | 0.5099 | 0.04 | 56 | 0.5161 |
114
+ | 0.5158 | 0.04 | 57 | 0.5123 |
115
+ | 0.521 | 0.05 | 58 | 0.5084 |
116
+ | 0.5177 | 0.05 | 59 | 0.5047 |
117
+ | 0.4964 | 0.05 | 60 | 0.5010 |
118
+ | 0.502 | 0.05 | 61 | 0.4974 |
119
+ | 0.5078 | 0.05 | 62 | 0.4942 |
120
+ | 0.4814 | 0.05 | 63 | 0.4913 |
121
+ | 0.4863 | 0.05 | 64 | 0.4887 |
122
+ | 0.4998 | 0.05 | 65 | 0.4864 |
123
+ | 0.5106 | 0.05 | 66 | 0.4842 |
124
+ | 0.5273 | 0.05 | 67 | 0.4822 |
125
+ | 0.4874 | 0.05 | 68 | 0.4803 |
126
+ | 0.4697 | 0.05 | 69 | 0.4785 |
127
+ | 0.4796 | 0.05 | 70 | 0.4768 |
128
+ | 0.4767 | 0.06 | 71 | 0.4753 |
129
+ | 0.4582 | 0.06 | 72 | 0.4739 |
130
+ | 0.5084 | 0.06 | 73 | 0.4725 |
131
+ | 0.4566 | 0.06 | 74 | 0.4712 |
132
+ | 0.4583 | 0.06 | 75 | 0.4700 |
133
+ | 0.4753 | 0.06 | 76 | 0.4689 |
134
+ | 0.4528 | 0.06 | 77 | 0.4678 |
135
+ | 0.4617 | 0.06 | 78 | 0.4667 |
136
+ | 0.499 | 0.06 | 79 | 0.4656 |
137
+ | 0.4368 | 0.06 | 80 | 0.4646 |
138
+ | 0.4939 | 0.06 | 81 | 0.4637 |
139
+ | 0.4446 | 0.06 | 82 | 0.4627 |
140
+ | 0.4428 | 0.07 | 83 | 0.4618 |
141
+ | 0.4737 | 0.07 | 84 | 0.4611 |
142
+ | 0.4391 | 0.07 | 85 | 0.4603 |
143
+ | 0.4985 | 0.07 | 86 | 0.4597 |
144
+ | 0.45 | 0.07 | 87 | 0.4590 |
145
+ | 0.4642 | 0.07 | 88 | 0.4585 |
146
+ | 0.4633 | 0.07 | 89 | 0.4579 |
147
+ | 0.4233 | 0.07 | 90 | 0.4574 |
148
+ | 0.4478 | 0.07 | 91 | 0.4570 |
149
+ | 0.4768 | 0.07 | 92 | 0.4565 |
150
+ | 0.4665 | 0.07 | 93 | 0.4562 |
151
+ | 0.4504 | 0.07 | 94 | 0.4560 |
152
+ | 0.4692 | 0.07 | 95 | 0.4557 |
153
+ | 0.4326 | 0.08 | 96 | 0.4555 |
154
+ | 0.4727 | 0.08 | 97 | 0.4554 |
155
+ | 0.4658 | 0.08 | 98 | 0.4553 |
156
+ | 0.4905 | 0.08 | 99 | 0.4552 |
157
+ | 0.4501 | 0.08 | 100 | 0.4551 |
158
+
159
+
160
+ ### Framework versions
161
+
162
+ - PEFT 0.8.2
163
+ - Transformers 4.37.2
164
+ - Pytorch 2.1.0+cu121
165
+ - Datasets 2.17.1
166
+ - Tokenizers 0.15.2
adapter_config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "unsloth/tinyllama-chat-bnb-4bit",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "loftq_config": {},
12
+ "lora_alpha": 32,
13
+ "lora_dropout": 0,
14
+ "megatron_config": null,
15
+ "megatron_core": "megatron.core",
16
+ "modules_to_save": null,
17
+ "peft_type": "LORA",
18
+ "r": 32,
19
+ "rank_pattern": {},
20
+ "revision": "unsloth",
21
+ "target_modules": [
22
+ "down_proj",
23
+ "o_proj",
24
+ "q_proj",
25
+ "k_proj",
26
+ "v_proj",
27
+ "gate_proj",
28
+ "up_proj"
29
+ ],
30
+ "task_type": "CAUSAL_LM",
31
+ "use_rslora": false
32
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:231c267945c4d72f46d58bafb4b1571e707de7efb6705ce460a55de1e8a77b20
3
+ size 100966336
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<unk>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
tokenizer_config.json ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "bos_token": "<s>",
31
+ "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
32
+ "clean_up_tokenization_spaces": false,
33
+ "eos_token": "</s>",
34
+ "legacy": false,
35
+ "model_max_length": 2048,
36
+ "pad_token": "<unk>",
37
+ "padding_side": "right",
38
+ "sp_model_kwargs": {},
39
+ "tokenizer_class": "LlamaTokenizer",
40
+ "unk_token": "<unk>",
41
+ "use_default_system_prompt": false
42
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59a4bdfc5ae600d438e679607c811537c2feeaa4d1fbcedd55e31b950c665ce7
3
+ size 4664