hugodk-sch commited on
Commit
21a5b07
1 Parent(s): 9d33454

End of training

Browse files
Files changed (4) hide show
  1. README.md +13 -11
  2. all_results.json +12 -12
  3. config.json +3 -1
  4. eval_results.json +13 -13
README.md CHANGED
@@ -1,11 +1,13 @@
1
  ---
2
  library_name: peft
3
  tags:
 
4
  - trl
5
  - dpo
6
- - alignment-handbook
7
  - generated_from_trainer
8
  base_model: NbAiLab/nb-gpt-j-6B-v2
 
 
9
  model-index:
10
  - name: aftonposten-6b-align-scan
11
  results: []
@@ -16,17 +18,17 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  # aftonposten-6b-align-scan
18
 
19
- This model is a fine-tuned version of [NbAiLab/nb-gpt-j-6B-v2](https://huggingface.co/NbAiLab/nb-gpt-j-6B-v2) on the None dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 0.5566
22
- - Rewards/chosen: 0.0778
23
- - Rewards/rejected: 0.0602
24
- - Rewards/accuracies: 0.5050
25
- - Rewards/margins: 0.0176
26
- - Logps/rejected: -37.4306
27
- - Logps/chosen: -33.9234
28
- - Logits/rejected: -2.2423
29
- - Logits/chosen: -2.2472
30
 
31
  ## Model description
32
 
 
1
  ---
2
  library_name: peft
3
  tags:
4
+ - alignment-handbook
5
  - trl
6
  - dpo
 
7
  - generated_from_trainer
8
  base_model: NbAiLab/nb-gpt-j-6B-v2
9
+ datasets:
10
+ - hugodk-sch/aftonposten_title_prefs
11
  model-index:
12
  - name: aftonposten-6b-align-scan
13
  results: []
 
18
 
19
  # aftonposten-6b-align-scan
20
 
21
+ This model is a fine-tuned version of [data/ap-gpt-j-6b-sft-qlora-04-08](https://huggingface.co/data/ap-gpt-j-6b-sft-qlora-04-08) on the hugodk-sch/aftonposten_title_prefs dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.5772
24
+ - Rewards/chosen: 0.0684
25
+ - Rewards/rejected: 0.0623
26
+ - Rewards/accuracies: 0.5307
27
+ - Rewards/margins: 0.0061
28
+ - Logps/rejected: -37.4276
29
+ - Logps/chosen: -33.9368
30
+ - Logits/rejected: -2.2420
31
+ - Logits/chosen: -2.2469
32
 
33
  ## Model description
34
 
all_results.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
  "epoch": 4.0,
3
- "eval_logits/chosen": -2.236751079559326,
4
- "eval_logits/rejected": -2.2318832874298096,
5
- "eval_logps/chosen": -34.04561233520508,
6
- "eval_logps/rejected": -37.546016693115234,
7
- "eval_loss": 0.5836836099624634,
8
- "eval_rewards/accuracies": 0.5427741408348083,
9
- "eval_rewards/chosen": -0.007742464076727629,
10
- "eval_rewards/margins": 0.01283606793731451,
11
- "eval_rewards/rejected": -0.02057853527367115,
12
- "eval_runtime": 145.5744,
13
  "eval_samples": 343,
14
- "eval_samples_per_second": 2.356,
15
- "eval_steps_per_second": 0.295,
16
  "train_loss": 0.20226844080856868,
17
  "train_runtime": 10768.956,
18
  "train_samples": 3079,
 
1
  {
2
  "epoch": 4.0,
3
+ "eval_logits/chosen": -2.2468910217285156,
4
+ "eval_logits/rejected": -2.24200439453125,
5
+ "eval_logps/chosen": -33.9367790222168,
6
+ "eval_logps/rejected": -37.42759704589844,
7
+ "eval_loss": 0.577243447303772,
8
+ "eval_rewards/accuracies": 0.5307309031486511,
9
+ "eval_rewards/chosen": 0.06843975186347961,
10
+ "eval_rewards/margins": 0.0061217062175273895,
11
+ "eval_rewards/rejected": 0.06231804937124252,
12
+ "eval_runtime": 145.2917,
13
  "eval_samples": 343,
14
+ "eval_samples_per_second": 2.361,
15
+ "eval_steps_per_second": 0.296,
16
  "train_loss": 0.20226844080856868,
17
  "train_runtime": 10768.956,
18
  "train_samples": 3079,
config.json CHANGED
@@ -18,6 +18,8 @@
18
  "n_layer": 28,
19
  "n_positions": 2048,
20
  "quantization_config": {
 
 
21
  "bnb_4bit_compute_dtype": "float16",
22
  "bnb_4bit_quant_type": "nf4",
23
  "bnb_4bit_use_double_quant": false,
@@ -45,7 +47,7 @@
45
  },
46
  "tie_word_embeddings": false,
47
  "tokenizer_class": "GPT2Tokenizer",
48
- "transformers_version": "4.37.2",
49
  "use_cache": true,
50
  "vocab_size": 50400
51
  }
 
18
  "n_layer": 28,
19
  "n_positions": 2048,
20
  "quantization_config": {
21
+ "_load_in_4bit": true,
22
+ "_load_in_8bit": false,
23
  "bnb_4bit_compute_dtype": "float16",
24
  "bnb_4bit_quant_type": "nf4",
25
  "bnb_4bit_use_double_quant": false,
 
47
  },
48
  "tie_word_embeddings": false,
49
  "tokenizer_class": "GPT2Tokenizer",
50
+ "transformers_version": "4.39.0.dev0",
51
  "use_cache": true,
52
  "vocab_size": 50400
53
  }
eval_results.json CHANGED
@@ -1,16 +1,16 @@
1
  {
2
- "epoch": 1.0,
3
- "eval_logits/chosen": -2.236751079559326,
4
- "eval_logits/rejected": -2.2318832874298096,
5
- "eval_logps/chosen": -34.04561233520508,
6
- "eval_logps/rejected": -37.546016693115234,
7
- "eval_loss": 0.5836836099624634,
8
- "eval_rewards/accuracies": 0.5427741408348083,
9
- "eval_rewards/chosen": -0.007742464076727629,
10
- "eval_rewards/margins": 0.01283606793731451,
11
- "eval_rewards/rejected": -0.02057853527367115,
12
- "eval_runtime": 145.5744,
13
  "eval_samples": 343,
14
- "eval_samples_per_second": 2.356,
15
- "eval_steps_per_second": 0.295
16
  }
 
1
  {
2
+ "epoch": 4.0,
3
+ "eval_logits/chosen": -2.2468910217285156,
4
+ "eval_logits/rejected": -2.24200439453125,
5
+ "eval_logps/chosen": -33.9367790222168,
6
+ "eval_logps/rejected": -37.42759704589844,
7
+ "eval_loss": 0.577243447303772,
8
+ "eval_rewards/accuracies": 0.5307309031486511,
9
+ "eval_rewards/chosen": 0.06843975186347961,
10
+ "eval_rewards/margins": 0.0061217062175273895,
11
+ "eval_rewards/rejected": 0.06231804937124252,
12
+ "eval_runtime": 145.2917,
13
  "eval_samples": 343,
14
+ "eval_samples_per_second": 2.361,
15
+ "eval_steps_per_second": 0.296
16
  }