Hugging Face's logo

Ogamon
/
ragtruth_llama3_8b_instruct

Safetensors
Model card Files Community
Ogamon commited on
Commit
d9b5f09
·
verified ·
1 Parent(s): 72e684a

Upload folder using huggingface_hub

Browse files
Files changed (8) hide show
  1. .gitattributes +1 -0
  2. rag_truth_hal_detection_result/all_results.json +10 -0
  3. rag_truth_hal_detection_result/generated_predictions.jsonl +3 -0
  4. rag_truth_hal_detection_result/llamaboard_config.yaml +19 -0
  5. rag_truth_hal_detection_result/predict_results.json +10 -0
  6. rag_truth_hal_detection_result/running_log.txt +0 -0
  7. rag_truth_hal_detection_result/trainer_log.jsonl +67 -0
  8. rag_truth_hal_detection_result/training_args.yaml +18 -0
.gitattributes CHANGED
@@ -36,3 +36,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
36
  rag_truth_hal_detection_model/checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
  rag_truth_hal_detection_model/checkpoint-1413/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
  rag_truth_hal_detection_model/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 
 
36
  rag_truth_hal_detection_model/checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
  rag_truth_hal_detection_model/checkpoint-1413/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
  rag_truth_hal_detection_model/tokenizer.json filter=lfs diff=lfs merge=lfs -text
39
+ rag_truth_hal_detection_result/generated_predictions.jsonl filter=lfs diff=lfs merge=lfs -text
rag_truth_hal_detection_result/all_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_bleu-4": 80.09487692307692,
3
+ "predict_model_preparation_time": 0.0048,
4
+ "predict_rouge-1": 89.4693425295858,
5
+ "predict_rouge-2": 86.05262034023669,
6
+ "predict_rouge-l": 88.33788853550296,
7
+ "predict_runtime": 520.4325,
8
+ "predict_samples_per_second": 5.188,
9
+ "predict_steps_per_second": 0.649
10
+ }
rag_truth_hal_detection_result/generated_predictions.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5f8000fa8933a502996436537f789416f2da76367d322289fd24a3cdc7bb5dd
3
+ size 11474845
rag_truth_hal_detection_result/llamaboard_config.yaml ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ eval.batch_size: 2
2
+ eval.cutoff_len: 4096
3
+ eval.dataset:
4
+ - test_hal_detection_1125
5
+ eval.dataset_dir: data
6
+ eval.max_new_tokens: 512
7
+ eval.max_samples: '100000'
8
+ eval.output_dir: eval_2024-11-25-17-00-40
9
+ eval.predict: true
10
+ eval.temperature: 0.95
11
+ eval.top_p: 0.7
12
+ top.booster: auto
13
+ top.checkpoint_path: train_2024-11-25-07-46-41
14
+ top.finetuning_type: full
15
+ top.model_name: Llama-3-8B-Instruct
16
+ top.quantization_bit: none
17
+ top.quantization_method: bitsandbytes
18
+ top.rope_scaling: none
19
+ top.template: llama3
rag_truth_hal_detection_result/predict_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_bleu-4": 80.09487692307692,
3
+ "predict_model_preparation_time": 0.0048,
4
+ "predict_rouge-1": 89.4693425295858,
5
+ "predict_rouge-2": 86.05262034023669,
6
+ "predict_rouge-l": 88.33788853550296,
7
+ "predict_runtime": 520.4325,
8
+ "predict_samples_per_second": 5.188,
9
+ "predict_steps_per_second": 0.649
10
+ }
rag_truth_hal_detection_result/running_log.txt ADDED
The diff for this file is too large to render. See raw diff
 
rag_truth_hal_detection_result/trainer_log.jsonl ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"current_steps": 5, "total_steps": 338, "percentage": 1.48, "elapsed_time": "0:00:06", "remaining_time": "0:07:01"}
2
+ {"current_steps": 10, "total_steps": 338, "percentage": 2.96, "elapsed_time": "0:00:11", "remaining_time": "0:06:14"}
3
+ {"current_steps": 15, "total_steps": 338, "percentage": 4.44, "elapsed_time": "0:00:22", "remaining_time": "0:08:09"}
4
+ {"current_steps": 20, "total_steps": 338, "percentage": 5.92, "elapsed_time": "0:00:29", "remaining_time": "0:07:55"}
5
+ {"current_steps": 25, "total_steps": 338, "percentage": 7.4, "elapsed_time": "0:00:38", "remaining_time": "0:08:05"}
6
+ {"current_steps": 30, "total_steps": 338, "percentage": 8.88, "elapsed_time": "0:00:52", "remaining_time": "0:09:00"}
7
+ {"current_steps": 35, "total_steps": 338, "percentage": 10.36, "elapsed_time": "0:00:58", "remaining_time": "0:08:25"}
8
+ {"current_steps": 40, "total_steps": 338, "percentage": 11.83, "elapsed_time": "0:01:05", "remaining_time": "0:08:08"}
9
+ {"current_steps": 45, "total_steps": 338, "percentage": 13.31, "elapsed_time": "0:01:10", "remaining_time": "0:07:41"}
10
+ {"current_steps": 50, "total_steps": 338, "percentage": 14.79, "elapsed_time": "0:01:16", "remaining_time": "0:07:22"}
11
+ {"current_steps": 55, "total_steps": 338, "percentage": 16.27, "elapsed_time": "0:01:22", "remaining_time": "0:07:04"}
12
+ {"current_steps": 60, "total_steps": 338, "percentage": 17.75, "elapsed_time": "0:01:30", "remaining_time": "0:06:58"}
13
+ {"current_steps": 65, "total_steps": 338, "percentage": 19.23, "elapsed_time": "0:01:36", "remaining_time": "0:06:43"}
14
+ {"current_steps": 70, "total_steps": 338, "percentage": 20.71, "elapsed_time": "0:01:42", "remaining_time": "0:06:31"}
15
+ {"current_steps": 75, "total_steps": 338, "percentage": 22.19, "elapsed_time": "0:01:51", "remaining_time": "0:06:31"}
16
+ {"current_steps": 80, "total_steps": 338, "percentage": 23.67, "elapsed_time": "0:01:57", "remaining_time": "0:06:19"}
17
+ {"current_steps": 85, "total_steps": 338, "percentage": 25.15, "elapsed_time": "0:02:09", "remaining_time": "0:06:26"}
18
+ {"current_steps": 90, "total_steps": 338, "percentage": 26.63, "elapsed_time": "0:02:19", "remaining_time": "0:06:24"}
19
+ {"current_steps": 95, "total_steps": 338, "percentage": 28.11, "elapsed_time": "0:02:38", "remaining_time": "0:06:45"}
20
+ {"current_steps": 100, "total_steps": 338, "percentage": 29.59, "elapsed_time": "0:02:45", "remaining_time": "0:06:33"}
21
+ {"current_steps": 105, "total_steps": 338, "percentage": 31.07, "elapsed_time": "0:02:51", "remaining_time": "0:06:20"}
22
+ {"current_steps": 110, "total_steps": 338, "percentage": 32.54, "elapsed_time": "0:02:57", "remaining_time": "0:06:07"}
23
+ {"current_steps": 115, "total_steps": 338, "percentage": 34.02, "elapsed_time": "0:03:03", "remaining_time": "0:05:56"}
24
+ {"current_steps": 120, "total_steps": 338, "percentage": 35.5, "elapsed_time": "0:03:09", "remaining_time": "0:05:44"}
25
+ {"current_steps": 125, "total_steps": 338, "percentage": 36.98, "elapsed_time": "0:03:14", "remaining_time": "0:05:32"}
26
+ {"current_steps": 130, "total_steps": 338, "percentage": 38.46, "elapsed_time": "0:03:23", "remaining_time": "0:05:24"}
27
+ {"current_steps": 135, "total_steps": 338, "percentage": 39.94, "elapsed_time": "0:03:31", "remaining_time": "0:05:18"}
28
+ {"current_steps": 140, "total_steps": 338, "percentage": 41.42, "elapsed_time": "0:03:38", "remaining_time": "0:05:08"}
29
+ {"current_steps": 145, "total_steps": 338, "percentage": 42.9, "elapsed_time": "0:03:46", "remaining_time": "0:05:00"}
30
+ {"current_steps": 150, "total_steps": 338, "percentage": 44.38, "elapsed_time": "0:03:54", "remaining_time": "0:04:53"}
31
+ {"current_steps": 155, "total_steps": 338, "percentage": 45.86, "elapsed_time": "0:04:04", "remaining_time": "0:04:48"}
32
+ {"current_steps": 160, "total_steps": 338, "percentage": 47.34, "elapsed_time": "0:04:10", "remaining_time": "0:04:38"}
33
+ {"current_steps": 165, "total_steps": 338, "percentage": 48.82, "elapsed_time": "0:04:15", "remaining_time": "0:04:28"}
34
+ {"current_steps": 170, "total_steps": 338, "percentage": 50.3, "elapsed_time": "0:04:24", "remaining_time": "0:04:21"}
35
+ {"current_steps": 175, "total_steps": 338, "percentage": 51.78, "elapsed_time": "0:04:30", "remaining_time": "0:04:12"}
36
+ {"current_steps": 180, "total_steps": 338, "percentage": 53.25, "elapsed_time": "0:04:37", "remaining_time": "0:04:03"}
37
+ {"current_steps": 185, "total_steps": 338, "percentage": 54.73, "elapsed_time": "0:04:47", "remaining_time": "0:03:57"}
38
+ {"current_steps": 190, "total_steps": 338, "percentage": 56.21, "elapsed_time": "0:04:53", "remaining_time": "0:03:48"}
39
+ {"current_steps": 195, "total_steps": 338, "percentage": 57.69, "elapsed_time": "0:04:59", "remaining_time": "0:03:39"}
40
+ {"current_steps": 200, "total_steps": 338, "percentage": 59.17, "elapsed_time": "0:05:03", "remaining_time": "0:03:29"}
41
+ {"current_steps": 205, "total_steps": 338, "percentage": 60.65, "elapsed_time": "0:05:08", "remaining_time": "0:03:20"}
42
+ {"current_steps": 210, "total_steps": 338, "percentage": 62.13, "elapsed_time": "0:05:12", "remaining_time": "0:03:10"}
43
+ {"current_steps": 215, "total_steps": 338, "percentage": 63.61, "elapsed_time": "0:05:17", "remaining_time": "0:03:01"}
44
+ {"current_steps": 220, "total_steps": 338, "percentage": 65.09, "elapsed_time": "0:05:21", "remaining_time": "0:02:52"}
45
+ {"current_steps": 225, "total_steps": 338, "percentage": 66.57, "elapsed_time": "0:05:27", "remaining_time": "0:02:44"}
46
+ {"current_steps": 230, "total_steps": 338, "percentage": 68.05, "elapsed_time": "0:05:36", "remaining_time": "0:02:37"}
47
+ {"current_steps": 235, "total_steps": 338, "percentage": 69.53, "elapsed_time": "0:05:43", "remaining_time": "0:02:30"}
48
+ {"current_steps": 240, "total_steps": 338, "percentage": 71.01, "elapsed_time": "0:05:52", "remaining_time": "0:02:23"}
49
+ {"current_steps": 245, "total_steps": 338, "percentage": 72.49, "elapsed_time": "0:06:01", "remaining_time": "0:02:17"}
50
+ {"current_steps": 250, "total_steps": 338, "percentage": 73.96, "elapsed_time": "0:06:09", "remaining_time": "0:02:10"}
51
+ {"current_steps": 255, "total_steps": 338, "percentage": 75.44, "elapsed_time": "0:06:17", "remaining_time": "0:02:03"}
52
+ {"current_steps": 260, "total_steps": 338, "percentage": 76.92, "elapsed_time": "0:06:26", "remaining_time": "0:01:55"}
53
+ {"current_steps": 265, "total_steps": 338, "percentage": 78.4, "elapsed_time": "0:06:35", "remaining_time": "0:01:48"}
54
+ {"current_steps": 270, "total_steps": 338, "percentage": 79.88, "elapsed_time": "0:06:43", "remaining_time": "0:01:41"}
55
+ {"current_steps": 275, "total_steps": 338, "percentage": 81.36, "elapsed_time": "0:06:52", "remaining_time": "0:01:34"}
56
+ {"current_steps": 280, "total_steps": 338, "percentage": 82.84, "elapsed_time": "0:07:00", "remaining_time": "0:01:27"}
57
+ {"current_steps": 285, "total_steps": 338, "percentage": 84.32, "elapsed_time": "0:07:10", "remaining_time": "0:01:19"}
58
+ {"current_steps": 290, "total_steps": 338, "percentage": 85.8, "elapsed_time": "0:07:17", "remaining_time": "0:01:12"}
59
+ {"current_steps": 295, "total_steps": 338, "percentage": 87.28, "elapsed_time": "0:07:26", "remaining_time": "0:01:05"}
60
+ {"current_steps": 300, "total_steps": 338, "percentage": 88.76, "elapsed_time": "0:07:35", "remaining_time": "0:00:57"}
61
+ {"current_steps": 305, "total_steps": 338, "percentage": 90.24, "elapsed_time": "0:07:42", "remaining_time": "0:00:50"}
62
+ {"current_steps": 310, "total_steps": 338, "percentage": 91.72, "elapsed_time": "0:07:49", "remaining_time": "0:00:42"}
63
+ {"current_steps": 315, "total_steps": 338, "percentage": 93.2, "elapsed_time": "0:07:58", "remaining_time": "0:00:34"}
64
+ {"current_steps": 320, "total_steps": 338, "percentage": 94.67, "elapsed_time": "0:08:05", "remaining_time": "0:00:27"}
65
+ {"current_steps": 325, "total_steps": 338, "percentage": 96.15, "elapsed_time": "0:08:13", "remaining_time": "0:00:19"}
66
+ {"current_steps": 330, "total_steps": 338, "percentage": 97.63, "elapsed_time": "0:08:21", "remaining_time": "0:00:12"}
67
+ {"current_steps": 335, "total_steps": 338, "percentage": 99.11, "elapsed_time": "0:08:30", "remaining_time": "0:00:04"}
rag_truth_hal_detection_result/training_args.yaml ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ cutoff_len: 4096
2
+ dataset_dir: data
3
+ do_predict: true
4
+ eval_dataset: test_hal_detection_1125
5
+ finetuning_type: full
6
+ flash_attn: auto
7
+ max_new_tokens: 512
8
+ max_samples: 100000
9
+ model_name_or_path: saves/Llama-3-8B-Instruct/full/train_2024-11-25-07-46-41
10
+ output_dir: saves/Llama-3-8B-Instruct/full/eval_2024-11-25-17-00-40
11
+ per_device_eval_batch_size: 2
12
+ predict_with_generate: true
13
+ preprocessing_num_workers: 16
14
+ quantization_method: bitsandbytes
15
+ stage: sft
16
+ temperature: 0.95
17
+ template: llama3
18
+ top_p: 0.7