EllieS commited on
Commit
8c89d20
1 Parent(s): 3d5b529

Model save

Browse files
README.md ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ library_name: peft
4
+ tags:
5
+ - trl
6
+ - dpo
7
+ - generated_from_trainer
8
+ base_model: alignment-handbook/zephyr-7b-sft-full
9
+ model-index:
10
+ - name: zephyr-7b-dpo-lora-pubmedqa-selfgen-ultrafeedback-old
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # zephyr-7b-dpo-lora-pubmedqa-selfgen-ultrafeedback-old
18
+
19
+ This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the None dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.5601
22
+ - Rewards/chosen: -0.4096
23
+ - Rewards/rejected: -0.9163
24
+ - Rewards/accuracies: 0.7000
25
+ - Rewards/margins: 0.5067
26
+ - Logps/rejected: -348.5905
27
+ - Logps/chosen: -331.9001
28
+ - Logits/rejected: -1.5161
29
+ - Logits/chosen: -1.5640
30
+
31
+ ## Model description
32
+
33
+ More information needed
34
+
35
+ ## Intended uses & limitations
36
+
37
+ More information needed
38
+
39
+ ## Training and evaluation data
40
+
41
+ More information needed
42
+
43
+ ## Training procedure
44
+
45
+ ### Training hyperparameters
46
+
47
+ The following hyperparameters were used during training:
48
+ - learning_rate: 5e-06
49
+ - train_batch_size: 2
50
+ - eval_batch_size: 2
51
+ - seed: 42
52
+ - distributed_type: multi-GPU
53
+ - num_devices: 2
54
+ - gradient_accumulation_steps: 2
55
+ - total_train_batch_size: 8
56
+ - total_eval_batch_size: 4
57
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
58
+ - lr_scheduler_type: cosine
59
+ - lr_scheduler_warmup_ratio: 0.1
60
+ - num_epochs: 1
61
+
62
+ ### Training results
63
+
64
+ | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
65
+ |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
66
+ | 0.5074 | 0.39 | 3000 | 0.5837 | -0.4893 | -0.8886 | 0.6940 | 0.3993 | -345.8194 | -339.8665 | -1.8683 | -1.8925 |
67
+ | 0.5332 | 0.79 | 6000 | 0.5604 | -0.4169 | -0.9240 | 0.7040 | 0.5071 | -349.3569 | -332.6285 | -1.5156 | -1.5626 |
68
+
69
+
70
+ ### Framework versions
71
+
72
+ - PEFT 0.7.1
73
+ - Transformers 4.36.2
74
+ - Pytorch 2.1.2+cu121
75
+ - Datasets 2.14.6
76
+ - Tokenizers 0.15.2
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:41974d189bc131ba781b9c4beaf2a15efd334f152cd18b8e75d0d79e37803223
3
  size 83946192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d14bec5a84eb6de960e49138002ca317a21d1d14128a8e79fc174bc80d6ef71
3
  size 83946192
all_results.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_logits/chosen": -1.563957691192627,
4
+ "eval_logits/rejected": -1.5160561800003052,
5
+ "eval_logps/chosen": -331.9000549316406,
6
+ "eval_logps/rejected": -348.5905456542969,
7
+ "eval_loss": 0.5600629448890686,
8
+ "eval_rewards/accuracies": 0.699999988079071,
9
+ "eval_rewards/chosen": -0.40962255001068115,
10
+ "eval_rewards/margins": 0.5067135691642761,
11
+ "eval_rewards/rejected": -0.9163362383842468,
12
+ "eval_runtime": 469.9532,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 4.256,
15
+ "eval_steps_per_second": 1.064,
16
+ "train_loss": 0.589801928247341,
17
+ "train_runtime": 31982.1679,
18
+ "train_samples": 61135,
19
+ "train_samples_per_second": 1.912,
20
+ "train_steps_per_second": 0.239
21
+ }
eval_results.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_logits/chosen": -1.563957691192627,
4
+ "eval_logits/rejected": -1.5160561800003052,
5
+ "eval_logps/chosen": -331.9000549316406,
6
+ "eval_logps/rejected": -348.5905456542969,
7
+ "eval_loss": 0.5600629448890686,
8
+ "eval_rewards/accuracies": 0.699999988079071,
9
+ "eval_rewards/chosen": -0.40962255001068115,
10
+ "eval_rewards/margins": 0.5067135691642761,
11
+ "eval_rewards/rejected": -0.9163362383842468,
12
+ "eval_runtime": 469.9532,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 4.256,
15
+ "eval_steps_per_second": 1.064
16
+ }
runs/Feb23_07-46-59_586cb8b6da8c/events.out.tfevents.1708674753.586cb8b6da8c.16472.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c742a2ea71e81c4782ec12545017c54f2417084e4c2a1278cdd97b25d554d29a
3
- size 450458
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d8feb79a2098ed08f87e2e8795efff58cc6e9f93010fc188b54ca52096eee1a
3
+ size 491388
runs/Feb23_07-46-59_586cb8b6da8c/events.out.tfevents.1708707204.586cb8b6da8c.16472.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31a12d894ee9e6463b16f13436331106ad29d8cbe2ced62a6dbb208f5ce0f7a3
3
+ size 828
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.589801928247341,
4
+ "train_runtime": 31982.1679,
5
+ "train_samples": 61135,
6
+ "train_samples_per_second": 1.912,
7
+ "train_steps_per_second": 0.239
8
+ }
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff