Hugging Face's logo

Ogamon
/
ragtruth_1123

Safetensors
Model card Files Community
Ogamon commited on
Commit
e599a56
·
verified ·
1 Parent(s): b88cf9b

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 +33 -0
  8. rag_truth_hal_detection_result/training_args.yaml +18 -0
.gitattributes CHANGED
@@ -35,3 +35,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
  rag_truth_hal_detection_model/checkpoint-705/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
  rag_truth_hal_detection_model/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 
 
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
  rag_truth_hal_detection_model/checkpoint-705/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
  rag_truth_hal_detection_model/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
+ 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": 81.5526401627219,
3
+ "predict_model_preparation_time": 0.0048,
4
+ "predict_rouge-1": 89.86517718195266,
5
+ "predict_rouge-2": 85.65232448224852,
6
+ "predict_rouge-l": 88.25301327662723,
7
+ "predict_runtime": 323.4963,
8
+ "predict_samples_per_second": 8.346,
9
+ "predict_steps_per_second": 0.522
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 sha256:6127bfabb39d7623b36631356e6e4db8d49598f43da11f6ef00b4b0fc7c4aef9
3
+ size 11442078
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
5
+ eval.dataset_dir: data
6
+ eval.max_new_tokens: 512
7
+ eval.max_samples: '100000'
8
+ eval.output_dir: ragtruth_eval_2024-11-23
9
+ eval.predict: true
10
+ eval.temperature: 0.95
11
+ eval.top_p: 0.7
12
+ top.booster: auto
13
+ top.checkpoint_path: train_2024-11-22-23-46-34
14
+ top.finetuning_type: full
15
+ top.model_name: Llama-3-8B
16
+ top.quantization_bit: none
17
+ top.quantization_method: bitsandbytes
18
+ top.rope_scaling: none
19
+ top.template: default
rag_truth_hal_detection_result/predict_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_bleu-4": 81.5526401627219,
3
+ "predict_model_preparation_time": 0.0048,
4
+ "predict_rouge-1": 89.86517718195266,
5
+ "predict_rouge-2": 85.65232448224852,
6
+ "predict_rouge-l": 88.25301327662723,
7
+ "predict_runtime": 323.4963,
8
+ "predict_samples_per_second": 8.346,
9
+ "predict_steps_per_second": 0.522
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 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"current_steps": 5, "total_steps": 169, "percentage": 2.96, "elapsed_time": "0:00:06", "remaining_time": "0:03:38"}
2
+ {"current_steps": 10, "total_steps": 169, "percentage": 5.92, "elapsed_time": "0:00:18", "remaining_time": "0:04:57"}
3
+ {"current_steps": 15, "total_steps": 169, "percentage": 8.88, "elapsed_time": "0:00:33", "remaining_time": "0:05:45"}
4
+ {"current_steps": 20, "total_steps": 169, "percentage": 11.83, "elapsed_time": "0:00:40", "remaining_time": "0:05:05"}
5
+ {"current_steps": 25, "total_steps": 169, "percentage": 14.79, "elapsed_time": "0:00:51", "remaining_time": "0:04:58"}
6
+ {"current_steps": 30, "total_steps": 169, "percentage": 17.75, "elapsed_time": "0:01:00", "remaining_time": "0:04:42"}
7
+ {"current_steps": 35, "total_steps": 169, "percentage": 20.71, "elapsed_time": "0:01:07", "remaining_time": "0:04:18"}
8
+ {"current_steps": 40, "total_steps": 169, "percentage": 23.67, "elapsed_time": "0:01:20", "remaining_time": "0:04:20"}
9
+ {"current_steps": 45, "total_steps": 169, "percentage": 26.63, "elapsed_time": "0:01:32", "remaining_time": "0:04:14"}
10
+ {"current_steps": 50, "total_steps": 169, "percentage": 29.59, "elapsed_time": "0:01:49", "remaining_time": "0:04:21"}
11
+ {"current_steps": 55, "total_steps": 169, "percentage": 32.54, "elapsed_time": "0:01:58", "remaining_time": "0:04:06"}
12
+ {"current_steps": 60, "total_steps": 169, "percentage": 35.5, "elapsed_time": "0:02:06", "remaining_time": "0:03:49"}
13
+ {"current_steps": 65, "total_steps": 169, "percentage": 38.46, "elapsed_time": "0:02:16", "remaining_time": "0:03:38"}
14
+ {"current_steps": 70, "total_steps": 169, "percentage": 41.42, "elapsed_time": "0:02:24", "remaining_time": "0:03:24"}
15
+ {"current_steps": 75, "total_steps": 169, "percentage": 44.38, "elapsed_time": "0:02:35", "remaining_time": "0:03:15"}
16
+ {"current_steps": 80, "total_steps": 169, "percentage": 47.34, "elapsed_time": "0:02:44", "remaining_time": "0:03:03"}
17
+ {"current_steps": 85, "total_steps": 169, "percentage": 50.3, "elapsed_time": "0:02:52", "remaining_time": "0:02:50"}
18
+ {"current_steps": 90, "total_steps": 169, "percentage": 53.25, "elapsed_time": "0:02:58", "remaining_time": "0:02:36"}
19
+ {"current_steps": 95, "total_steps": 169, "percentage": 56.21, "elapsed_time": "0:03:07", "remaining_time": "0:02:25"}
20
+ {"current_steps": 100, "total_steps": 169, "percentage": 59.17, "elapsed_time": "0:03:12", "remaining_time": "0:02:12"}
21
+ {"current_steps": 105, "total_steps": 169, "percentage": 62.13, "elapsed_time": "0:03:17", "remaining_time": "0:02:00"}
22
+ {"current_steps": 110, "total_steps": 169, "percentage": 65.09, "elapsed_time": "0:03:23", "remaining_time": "0:01:49"}
23
+ {"current_steps": 115, "total_steps": 169, "percentage": 68.05, "elapsed_time": "0:03:31", "remaining_time": "0:01:39"}
24
+ {"current_steps": 120, "total_steps": 169, "percentage": 71.01, "elapsed_time": "0:03:41", "remaining_time": "0:01:30"}
25
+ {"current_steps": 125, "total_steps": 169, "percentage": 73.96, "elapsed_time": "0:03:51", "remaining_time": "0:01:21"}
26
+ {"current_steps": 130, "total_steps": 169, "percentage": 76.92, "elapsed_time": "0:04:01", "remaining_time": "0:01:12"}
27
+ {"current_steps": 135, "total_steps": 169, "percentage": 79.88, "elapsed_time": "0:04:11", "remaining_time": "0:01:03"}
28
+ {"current_steps": 140, "total_steps": 169, "percentage": 82.84, "elapsed_time": "0:04:21", "remaining_time": "0:00:54"}
29
+ {"current_steps": 145, "total_steps": 169, "percentage": 85.8, "elapsed_time": "0:04:31", "remaining_time": "0:00:44"}
30
+ {"current_steps": 150, "total_steps": 169, "percentage": 88.76, "elapsed_time": "0:04:42", "remaining_time": "0:00:35"}
31
+ {"current_steps": 155, "total_steps": 169, "percentage": 91.72, "elapsed_time": "0:04:50", "remaining_time": "0:00:26"}
32
+ {"current_steps": 160, "total_steps": 169, "percentage": 94.67, "elapsed_time": "0:05:00", "remaining_time": "0:00:16"}
33
+ {"current_steps": 165, "total_steps": 169, "percentage": 97.63, "elapsed_time": "0:05:08", "remaining_time": "0:00:07"}
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
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/full/train_2024-11-22-23-46-34
10
+ output_dir: saves/Llama-3-8B/full/ragtruth_eval_2024-11-23
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: default
18
+ top_p: 0.7