Hugging Face's logo

llama-duo
/
gemma2b-summarize-claude3sonnet-64k

PEFT
TensorBoard
Safetensors
gemma
alignment-handbook
trl
sft
Generated from Trainer
4-bit precision
bitsandbytes
Model card Files Metrics Community
chansung commited on
Commit
c382497
·
verified ·
1 Parent(s): 0202373

Model save

Browse files
Files changed (6) hide show
  1. README.md +18 -24
  2. adapter_model.safetensors +1 -1
  3. all_results.json +7 -12
  4. runs/Jun10_15-41-34_user-HP-Z8-Fury-G5-Workstation-Desktop-PC/events.out.tfevents.1718001709.user-HP-Z8-Fury-G5-Workstation-Desktop-PC.10374.0 +2 -2
  5. train_results.json +7 -7
  6. trainer_state.json +0 -0
README.md CHANGED
@@ -2,13 +2,12 @@
2
  license: gemma
3
  library_name: peft
4
  tags:
5
- - alignment-handbook
6
  - trl
7
  - sft
8
  - generated_from_trainer
9
  base_model: google/gemma-2b
10
  datasets:
11
- - llama-duo/synth_summarize_dataset_dedup
12
  model-index:
13
  - name: gemma2b-summarize-claude3sonnet-64k
14
  results: []
@@ -19,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
19
 
20
  # gemma2b-summarize-claude3sonnet-64k
21
 
22
- This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the llama-duo/synth_summarize_dataset_dedup dataset.
23
  It achieves the following results on the evaluation set:
24
- - Loss: 2.5427
25
 
26
  ## Model description
27
 
@@ -52,33 +51,28 @@ The following hyperparameters were used during training:
52
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
53
  - lr_scheduler_type: cosine
54
  - lr_scheduler_warmup_ratio: 0.1
55
- - num_epochs: 15
56
 
57
  ### Training results
58
 
59
- | Training Loss | Epoch | Step | Validation Loss |
60
- |:-------------:|:-------:|:----:|:---------------:|
61
- | 1.163 | 0.9975 | 200 | 2.5127 |
62
- | 1.0647 | 2.0 | 401 | 2.4643 |
63
- | 1.0051 | 2.9975 | 601 | 2.4610 |
64
- | 0.9807 | 4.0 | 802 | 2.4767 |
65
- | 0.9508 | 4.9975 | 1002 | 2.4788 |
66
- | 0.9256 | 6.0 | 1203 | 2.4912 |
67
- | 0.9216 | 6.9975 | 1403 | 2.5038 |
68
- | 0.9094 | 8.0 | 1604 | 2.5124 |
69
- | 0.8961 | 8.9975 | 1804 | 2.5246 |
70
- | 0.8816 | 10.0 | 2005 | 2.5342 |
71
- | 0.8722 | 10.9975 | 2205 | 2.5346 |
72
- | 0.8768 | 12.0 | 2406 | 2.5410 |
73
- | 0.8694 | 12.9975 | 2606 | 2.5415 |
74
- | 0.8709 | 14.0 | 2807 | 2.5418 |
75
- | 0.8781 | 14.9626 | 3000 | 2.5427 |
76
 
77
 
78
  ### Framework versions
79
 
80
- - PEFT 0.10.0
81
- - Transformers 4.40.0
82
  - Pytorch 2.2.2+cu121
83
  - Datasets 2.19.1
84
  - Tokenizers 0.19.1
 
2
  license: gemma
3
  library_name: peft
4
  tags:
 
5
  - trl
6
  - sft
7
  - generated_from_trainer
8
  base_model: google/gemma-2b
9
  datasets:
10
+ - generator
11
  model-index:
12
  - name: gemma2b-summarize-claude3sonnet-64k
13
  results: []
 
18
 
19
  # gemma2b-summarize-claude3sonnet-64k
20
 
21
+ This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the generator dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 2.6980
24
 
25
  ## Model description
26
 
 
51
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
  - lr_scheduler_type: cosine
53
  - lr_scheduler_warmup_ratio: 0.1
54
+ - num_epochs: 10
55
 
56
  ### Training results
57
 
58
+ | Training Loss | Epoch | Step | Validation Loss |
59
+ |:-------------:|:------:|:----:|:---------------:|
60
+ | 1.0668 | 0.9975 | 200 | 2.4713 |
61
+ | 0.9846 | 2.0 | 401 | 2.4650 |
62
+ | 0.941 | 2.9975 | 601 | 2.4875 |
63
+ | 0.8872 | 4.0 | 802 | 2.5213 |
64
+ | 0.8497 | 4.9975 | 1002 | 2.5719 |
65
+ | 0.8085 | 6.0 | 1203 | 2.6190 |
66
+ | 0.7905 | 6.9975 | 1403 | 2.6576 |
67
+ | 0.7684 | 8.0 | 1604 | 2.6892 |
68
+ | 0.7564 | 8.9975 | 1804 | 2.6970 |
69
+ | 0.747 | 9.9751 | 2000 | 2.6980 |
 
 
 
 
 
70
 
71
 
72
  ### Framework versions
73
 
74
+ - PEFT 0.11.1
75
+ - Transformers 4.41.2
76
  - Pytorch 2.2.2+cu121
77
  - Datasets 2.19.1
78
  - Tokenizers 0.19.1
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ac4883f1bdbb0011d18327f7ec99c5fa1f4ef8e4d89244ea47ff06355878486e
3
  size 78480320
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c85d89a83cb56e7340bbcfea29abb99cd61a4497286744160d73905f7fecafc
3
  size 78480320
all_results.json CHANGED
@@ -1,14 +1,9 @@
1
  {
2
- "epoch": 14.962593516209477,
3
- "eval_loss": 2.542728900909424,
4
- "eval_runtime": 0.4882,
5
- "eval_samples": 25,
6
- "eval_samples_per_second": 20.483,
7
- "eval_steps_per_second": 2.048,
8
- "total_flos": 1.7621200159784305e+18,
9
- "train_loss": 0.9822600702444713,
10
- "train_runtime": 10311.9233,
11
- "train_samples": 63494,
12
- "train_samples_per_second": 13.994,
13
- "train_steps_per_second": 0.291
14
  }
 
1
  {
2
+ "epoch": 9.975062344139651,
3
+ "total_flos": 1.1920978083462513e+18,
4
+ "train_loss": 0.9040022449493408,
5
+ "train_runtime": 12134.7289,
6
+ "train_samples": 63434,
7
+ "train_samples_per_second": 7.919,
8
+ "train_steps_per_second": 0.165
 
 
 
 
 
9
  }
runs/Jun10_15-41-34_user-HP-Z8-Fury-G5-Workstation-Desktop-PC/events.out.tfevents.1718001709.user-HP-Z8-Fury-G5-Workstation-Desktop-PC.10374.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1f397ab195ba6ee90385de91ea40e4fd0e3e25b6f7e5fd519ae7d7d4925af694
3
- size 88355
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0be7ce387348a27a437e1b493ea99f2e408b34e2560975089fd60e45c1c53c8c
3
+ size 93200
train_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
- "epoch": 14.962593516209477,
3
- "total_flos": 1.7621200159784305e+18,
4
- "train_loss": 0.9822600702444713,
5
- "train_runtime": 10311.9233,
6
- "train_samples": 63494,
7
- "train_samples_per_second": 13.994,
8
- "train_steps_per_second": 0.291
9
  }
 
1
  {
2
+ "epoch": 9.975062344139651,
3
+ "total_flos": 1.1920978083462513e+18,
4
+ "train_loss": 0.9040022449493408,
5
+ "train_runtime": 12134.7289,
6
+ "train_samples": 63434,
7
+ "train_samples_per_second": 7.919,
8
+ "train_steps_per_second": 0.165
9
  }
trainer_state.json CHANGED
The diff for this file is too large to render. See raw diff