PY007 commited on
Commit
5243d15
1 Parent(s): eb800a9

Model save

Browse files
README.md ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - generated_from_trainer
4
+ model-index:
5
+ - name: TinyLlama-1.1B-Chat-v1.0
6
+ results: []
7
+ ---
8
+
9
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
10
+ should probably proofread and complete it, then remove this comment. -->
11
+
12
+ # TinyLlama-1.1B-Chat-v1.0
13
+
14
+ This model was trained from scratch on the None dataset.
15
+ It achieves the following results on the evaluation set:
16
+ - Loss: 0.5138
17
+ - Rewards/chosen: -0.0274
18
+ - Rewards/rejected: -1.0362
19
+ - Rewards/accuracies: 0.7381
20
+ - Rewards/margins: 1.0087
21
+ - Logps/rejected: -296.0739
22
+ - Logps/chosen: -370.1298
23
+ - Logits/rejected: -2.6565
24
+ - Logits/chosen: -2.7074
25
+
26
+ ## Model description
27
+
28
+ More information needed
29
+
30
+ ## Intended uses & limitations
31
+
32
+ More information needed
33
+
34
+ ## Training and evaluation data
35
+
36
+ More information needed
37
+
38
+ ## Training procedure
39
+
40
+ ### Training hyperparameters
41
+
42
+ The following hyperparameters were used during training:
43
+ - learning_rate: 5e-07
44
+ - train_batch_size: 8
45
+ - eval_batch_size: 4
46
+ - seed: 42
47
+ - distributed_type: multi-GPU
48
+ - num_devices: 8
49
+ - total_train_batch_size: 64
50
+ - total_eval_batch_size: 32
51
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
+ - lr_scheduler_type: linear
53
+ - lr_scheduler_warmup_ratio: 0.1
54
+ - num_epochs: 3
55
+
56
+ ### Training results
57
+
58
+ | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
59
+ |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
60
+ | 0.6686 | 0.1 | 100 | 0.6668 | 0.0624 | -0.0106 | 0.6746 | 0.0730 | -285.8178 | -369.2313 | -2.7623 | -2.8330 |
61
+ | 0.59 | 0.21 | 200 | 0.5995 | 0.1603 | -0.1926 | 0.6825 | 0.3530 | -287.6386 | -368.2522 | -2.7514 | -2.8180 |
62
+ | 0.5843 | 0.31 | 300 | 0.5644 | 0.2269 | -0.3175 | 0.6905 | 0.5444 | -288.8868 | -367.5864 | -2.7305 | -2.7952 |
63
+ | 0.5633 | 0.41 | 400 | 0.5476 | 0.2211 | -0.4312 | 0.7103 | 0.6523 | -290.0246 | -367.6447 | -2.7100 | -2.7725 |
64
+ | 0.5224 | 0.52 | 500 | 0.5388 | 0.2702 | -0.4543 | 0.6984 | 0.7244 | -290.2547 | -367.1539 | -2.6919 | -2.7543 |
65
+ | 0.5689 | 0.62 | 600 | 0.5326 | 0.3161 | -0.4312 | 0.7302 | 0.7473 | -290.0246 | -366.6946 | -2.6977 | -2.7596 |
66
+ | 0.5556 | 0.72 | 700 | 0.5296 | 0.3133 | -0.4431 | 0.7143 | 0.7565 | -290.1436 | -366.7222 | -2.6960 | -2.7563 |
67
+ | 0.5368 | 0.83 | 800 | 0.5235 | 0.3087 | -0.5008 | 0.7183 | 0.8096 | -290.7203 | -366.7679 | -2.6863 | -2.7455 |
68
+ | 0.5324 | 0.93 | 900 | 0.5231 | 0.3330 | -0.4764 | 0.7381 | 0.8094 | -290.4763 | -366.5252 | -2.6944 | -2.7532 |
69
+ | 0.4667 | 1.03 | 1000 | 0.5211 | 0.3442 | -0.4815 | 0.7302 | 0.8257 | -290.5269 | -366.4131 | -2.6890 | -2.7466 |
70
+ | 0.4516 | 1.14 | 1100 | 0.5197 | 0.2843 | -0.6031 | 0.7381 | 0.8874 | -291.7431 | -367.0122 | -2.6770 | -2.7325 |
71
+ | 0.4176 | 1.24 | 1200 | 0.5184 | 0.2116 | -0.7161 | 0.7460 | 0.9276 | -292.8727 | -367.7397 | -2.6729 | -2.7277 |
72
+ | 0.446 | 1.34 | 1300 | 0.5187 | 0.2095 | -0.6963 | 0.7421 | 0.9058 | -292.6750 | -367.7603 | -2.6740 | -2.7278 |
73
+ | 0.472 | 1.44 | 1400 | 0.5154 | 0.2233 | -0.6454 | 0.7540 | 0.8686 | -292.1659 | -367.6227 | -2.6716 | -2.7264 |
74
+ | 0.4425 | 1.55 | 1500 | 0.5158 | 0.1986 | -0.7079 | 0.7381 | 0.9065 | -292.7915 | -367.8694 | -2.6695 | -2.7244 |
75
+ | 0.434 | 1.65 | 1600 | 0.5148 | 0.2037 | -0.6841 | 0.7381 | 0.8878 | -292.5535 | -367.8188 | -2.6639 | -2.7187 |
76
+ | 0.4209 | 1.75 | 1700 | 0.5146 | 0.1297 | -0.7819 | 0.7460 | 0.9116 | -293.5308 | -368.5582 | -2.6636 | -2.7185 |
77
+ | 0.4128 | 1.86 | 1800 | 0.5129 | 0.1418 | -0.7822 | 0.7381 | 0.9240 | -293.5338 | -368.4372 | -2.6651 | -2.7194 |
78
+ | 0.4685 | 1.96 | 1900 | 0.5125 | 0.0967 | -0.8256 | 0.7421 | 0.9223 | -293.9677 | -368.8879 | -2.6709 | -2.7248 |
79
+ | 0.3605 | 2.06 | 2000 | 0.5130 | 0.0627 | -0.8947 | 0.7302 | 0.9574 | -294.6591 | -369.2281 | -2.6689 | -2.7211 |
80
+ | 0.3463 | 2.17 | 2100 | 0.5123 | 0.0453 | -0.9465 | 0.7421 | 0.9918 | -295.1770 | -369.4025 | -2.6709 | -2.7218 |
81
+ | 0.362 | 2.27 | 2200 | 0.5125 | 0.0174 | -0.9774 | 0.7381 | 0.9948 | -295.4861 | -369.6811 | -2.6628 | -2.7140 |
82
+ | 0.354 | 2.37 | 2300 | 0.5148 | 0.0053 | -0.9919 | 0.7421 | 0.9972 | -295.6311 | -369.8024 | -2.6562 | -2.7070 |
83
+ | 0.3539 | 2.48 | 2400 | 0.5144 | -0.0049 | -0.9987 | 0.7381 | 0.9939 | -295.6994 | -369.9039 | -2.6557 | -2.7070 |
84
+ | 0.3374 | 2.58 | 2500 | 0.5143 | -0.0015 | -1.0170 | 0.75 | 1.0156 | -295.8826 | -369.8703 | -2.6616 | -2.7128 |
85
+ | 0.3417 | 2.68 | 2600 | 0.5137 | 0.0000 | -1.0041 | 0.7341 | 1.0041 | -295.7533 | -369.8551 | -2.6605 | -2.7118 |
86
+ | 0.3312 | 2.79 | 2700 | 0.5140 | -0.0197 | -1.0285 | 0.7302 | 1.0089 | -295.9977 | -370.0519 | -2.6563 | -2.7071 |
87
+ | 0.3643 | 2.89 | 2800 | 0.5146 | -0.0233 | -1.0285 | 0.7421 | 1.0052 | -295.9974 | -370.0886 | -2.6552 | -2.7063 |
88
+ | 0.3322 | 2.99 | 2900 | 0.5142 | -0.0293 | -1.0337 | 0.7302 | 1.0045 | -296.0496 | -370.1480 | -2.6573 | -2.7079 |
89
+
90
+
91
+ ### Framework versions
92
+
93
+ - Transformers 4.35.0
94
+ - Pytorch 2.1.2+cu121
95
+ - Datasets 2.14.6
96
+ - Tokenizers 0.14.1
all_results.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_logits/chosen": -2.707406759262085,
4
+ "eval_logits/rejected": -2.656524419784546,
5
+ "eval_logps/chosen": -370.1297607421875,
6
+ "eval_logps/rejected": -296.0738525390625,
7
+ "eval_loss": 0.513750433921814,
8
+ "eval_rewards/accuracies": 0.738095223903656,
9
+ "eval_rewards/chosen": -0.02744222804903984,
10
+ "eval_rewards/margins": 1.0087225437164307,
11
+ "eval_rewards/rejected": -1.03616464138031,
12
+ "eval_runtime": 93.5908,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 21.37,
15
+ "eval_steps_per_second": 0.673,
16
+ "train_loss": 0.45207962556766157,
17
+ "train_runtime": 21062.2956,
18
+ "train_samples": 61966,
19
+ "train_samples_per_second": 8.826,
20
+ "train_steps_per_second": 0.138
21
+ }
config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/mnt/sfs-common/pyzhang/Peiyuan/alignment-handbook/data/zephyr-7b-sft-full",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 2048,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 5632,
13
+ "max_position_embeddings": 2048,
14
+ "model_type": "llama",
15
+ "num_attention_heads": 32,
16
+ "num_hidden_layers": 22,
17
+ "num_key_value_heads": 4,
18
+ "pretraining_tp": 1,
19
+ "rms_norm_eps": 1e-05,
20
+ "rope_scaling": null,
21
+ "rope_theta": 10000.0,
22
+ "tie_word_embeddings": false,
23
+ "torch_dtype": "bfloat16",
24
+ "transformers_version": "4.35.0",
25
+ "use_cache": false,
26
+ "vocab_size": 32000
27
+ }
eval_results.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_logits/chosen": -2.707406759262085,
4
+ "eval_logits/rejected": -2.656524419784546,
5
+ "eval_logps/chosen": -370.1297607421875,
6
+ "eval_logps/rejected": -296.0738525390625,
7
+ "eval_loss": 0.513750433921814,
8
+ "eval_rewards/accuracies": 0.738095223903656,
9
+ "eval_rewards/chosen": -0.02744222804903984,
10
+ "eval_rewards/margins": 1.0087225437164307,
11
+ "eval_rewards/rejected": -1.03616464138031,
12
+ "eval_runtime": 93.5908,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 21.37,
15
+ "eval_steps_per_second": 0.673
16
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 1,
3
+ "eos_token_id": 2,
4
+ "max_length": 2048,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.35.0"
7
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e6001da2106d4757498752a021df6c2bdc332c650aae4bae6b0c004dcf14933
3
+ size 2200119864
runs/Dec30_14-29-34_slab-0002/events.out.tfevents.1703917820.slab-0002.639716.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ec074a75918d501630693aebce3f26b81247d43e72835a3e085f9d9954a3c4
3
+ size 5633
runs/Dec30_14-32-19_slab-0002/events.out.tfevents.1703917981.slab-0002.641072.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac07394363b907a009e199b28f8b66bbfacde478f5acb04b071cb50ac4648d28
3
+ size 210539
runs/Dec30_14-32-19_slab-0002/events.out.tfevents.1703939137.slab-0002.641072.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c5a0e3194983b7acd9440e64f211c543a04349ca27ad56a87c7fdc9ef33f9d3
3
+ size 828
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": "</s>",
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_config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<unk>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<s>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ }
27
+ },
28
+ "bos_token": "<s>",
29
+ "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 %}",
30
+ "clean_up_tokenization_spaces": false,
31
+ "eos_token": "</s>",
32
+ "legacy": false,
33
+ "model_max_length": 2048,
34
+ "pad_token": "</s>",
35
+ "padding_side": "right",
36
+ "sp_model_kwargs": {},
37
+ "tokenizer_class": "LlamaTokenizer",
38
+ "unk_token": "<unk>",
39
+ "use_default_system_prompt": false
40
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "train_loss": 0.45207962556766157,
4
+ "train_runtime": 21062.2956,
5
+ "train_samples": 61966,
6
+ "train_samples_per_second": 8.826,
7
+ "train_steps_per_second": 0.138
8
+ }
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc9f0d837c4ec3e572a552bd4b0fa2541d092ee1341d82c2cdfdf7ebe948b2da
3
+ size 5560